Method for predicting a temperature-induced or heat-induced deformation of a workpiece during laser cutting
A predictive method for laser cutting processes identifies and addresses potential deformations before manufacturing, enhancing reliability and reducing costs by avoiding real-time system failures.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- TRUMPF WERKZEUGMASCHINEN GMBH & CO KG
- Filing Date
- 2025-11-20
- Publication Date
- 2026-06-18
AI Technical Summary
Existing methods for real-time adjustment of cutting speed during laser cutting are prone to system failures, leading to potential damage and reduced process reliability, necessitating costly and complex fail-safe systems.
A predictive method that identifies potential process safety-critical events before the manufacturing process, using geometry and manufacturing data to determine deformations and adjust parameters to prevent critical events, eliminating the need for real-time systems.
Enhances process reliability by preventing critical events through proactive measures, reducing downtime and costs associated with real-time system failures.
Smart Images
Figure EP2025083625_18062026_PF_FP_ABST
Abstract
Description
[0001] Methods for predicting a critical event and methods for carrying out a manufacturing process
[0002] Background of the invention
[0003] The invention relates to a method for predicting at least one process safety-critical event and a method for carrying out a manufacturing process.
[0004] CN 115329631 A discloses a method for compensating for errors caused by thermal deformation of a workpiece during laser cutting. For this purpose, the thermal deformation of the workpiece during laser cutting is determined. Depending on the thermal deformation, the cutting speed is adjusted in real time during laser cutting to reduce the thermal deformation at the cutting gap.
[0005] A problem with CN 115329631 A is that the cutting speed is adjusted in real time during laser cutting. Adjusting the cutting speed in real time during laser cutting can, in the event of a faulty or failed real-time system, impair or disrupt the laser cutting machine and / or the processing process. For example, if the real-time system malfunctions, the resulting high heat input into the workpiece and any residual grid due to thermal expansion can initially cause the workpiece and any residual grid to tilt, followed by warping. The resulting protrusion can create an obstacle against which a laser nozzle can collide and be damaged. This reduces process reliability.Therefore, the real-time system for the laser cutting machine must be designed to be particularly reliable and fail-safe, which makes the development of the real-time system more difficult and increases the costs for the real-time system.
[0006] The object of the invention is to provide a method that is improved in terms of process reliability while being simple and cost-effective to implement.
[0007] Description of the invention
[0008] This problem is solved according to the invention by a method having the features of claim 1 and a further method having the features of claim 15. The dependent claims describe preferred embodiments of the invention.
[0009] According to the invention, a method is provided for predicting at least one process safety-critical event of a manufacturing process of a workpiece to be manufactured from a starting material, in particular using a laser processing device.
[0010] A method for predicting a process safety-critical event in a manufacturing process typically involves carrying out the method, and in particular its steps, before the actual execution of the manufacturing process. This allows the process safety-critical event to be predicted before the manufacturing process is carried out.
[0011] A process safety-critical event can be an event that results in at least a temporary standstill of the manufacturing process.
[0012] The manufacturing process can include a laser processing process and subsequent process steps, such as picking up and transferring the workpiece to another location for further processing immediately after the laser processing. The laser processing process can be carried out using a laser processing device. The laser processing process is preferably a laser cutting process and / or the laser processing device is preferably a laser cutting device. It is also conceivable that the laser processing process is a laser surface treatment or a laser welding process and / or that the laser processing device is a laser surface treatment or laser welding device.
[0013] A laser cutting process is a separation process in which the starting material is cut using a laser beam. Within a defined process zone on the starting material, the laser beam causes temporary, localized vaporization and melting. Moving the laser beam along a predetermined cutting path creates a kerf. The resulting molten material is preferably expelled from the kerf by gas pressure, preventing solidification and enabling the cutting process.
[0014] The starting material and / or the workpiece can be plate-like. Plate-like means that the starting material and / or the workpiece extends in a plane and has a low material thickness compared to flat surfaces.
[0015] The process includes at least the following process steps, which are carried out before the manufacturing process of the workpiece:
[0016] In a first process step, geometry data of the workpiece and manufacturing data for the manufacturing process of the workpiece are provided.
[0017] Providing the data can mean that the data is made available to a computing unit via a human-machine interface for the execution of the process. A human-machine interface can be a touchscreen or a keyboard. Alternatively or additionally, the geometry data can be provided by supplying a production plan that includes the geometry data. The production plan can, for example, be transmitted digitally to the computing unit by an order management unit.
[0018] In a second process step, at least one deformation of the workpiece and / or the remaining starting material in relation to a predetermined geometry is determined during the manufacturing process according to geometry data and depending on the manufacturing data.
[0019] Deformation of the workpiece and / or the remaining starting material can be understood as an increase or decrease in volume and / or a change in shape of the workpiece and / or the remaining starting material.
[0020] The workpiece can be manufactured from the starting material, for example, a sheet, producing a residual material, such as a residual grid. This residual material can be the starting material minus the workpiece material. In other words, the residual material can be the material left over after at least partial processing of the starting material (laser cutting). It is also conceivable that the starting material is in multiple parts and is joined together to form the workpiece through processing (laser welding). Furthermore, it is conceivable that the workpiece is created from the starting material through processing (laser surface treatment). In the latter two variants, no residual material may remain.
[0021] In a third process step, the respective process safety-critical event of the manufacturing process is predicted depending on the respective deformation of the workpiece and / or the remaining starting material.
[0022] In a fourth process step, the respective process safety-critical event is output and / or the manufacturing process is adapted to the respective process safety-critical event by implementing at least one measure to deal with this event.
[0023] The output of the respective process safety-critical event can be done via a human-machine interface. This interface can be configured as a display to visualize the event. The display can be a screen. Alternatively or additionally, the event can be output in a machine-readable format for further processing.
[0024] The adjustment of the manufacturing process is essentially aimed either at reducing the heat input into a predicted process safety-critical area of the workpiece and / or the remaining starting material during the actual laser processing, and / or at allowing a process safety-critical area of the workpiece and / or the remaining starting material that has already been heated by heat input to cool down.
[0025] The process is carried out using at least one computing unit. The computing unit is typically configured to perform at least one process step.
[0026] By predicting the relevant process safety-critical event before the workpiece manufacturing process begins, advantageous measures can be taken to address the event even before production starts. This allows the manufacturing process to be adapted to the specific event, either preventing it from occurring altogether or finding a way to manage it without negatively impacting the process. This avoids downtime, thus improving process safety. Because the procedural steps are performed before, rather than during, the workpiece manufacturing process, a real-time system is not required. This makes the method simple and cost-effective.
[0027] In a preferred embodiment, the temperature change of the workpiece and / or the remaining raw material during the manufacturing process is determined based on the geometric and manufacturing data to ascertain the respective deformation. This temperature change can be determined using a lookup table that correlates manufacturing data and temperature change. The temperature change can be spatially and / or temporally resolved. The spatial and / or temporal temperature change can be presented as a heat map, which depicts the heating of the workpiece and / or the remaining raw material.
[0028] In a further preferred embodiment, the temperature change of the workpiece and / or the remaining raw material is determined using a self-learning algorithm. Determining the temperature change as a function of the geometric data and the manufacturing data using a self-learning algorithm is advantageous in that the algorithm delivers a result relatively quickly to the input data, in particular within a few seconds. The training of the self-learning algorithm with respect to the relationship between the geometric data, the manufacturing data, and the temperature change of the workpiece and / or the remaining raw material is described in more detail below.
[0029] In a further preferred embodiment, the respective deformation is determined as a function of the temperature change of the workpiece and / or the remaining starting material, and in particular, the manufacturing data. The respective deformation can be determined using a conversion table that assigns temperature changes of the workpiece and / or the remaining starting material to deformations of the workpiece and / or the remaining starting material. Preferably, however, the respective deformation is calculated from the temperature change and the manufacturing data. In a further preferred embodiment, a criticality of the respective deformation is determined to identify the respective process safety-critical event. Preferably, the criticality can be defined by at least one conversion table (look-up table) that assigns a criticality to the respective deformation.The criticality can at least include the assessments of (process safety-)critical and (process safety-)non-critical.
[0030] In a further preferred embodiment, it is provided that critical points of the workpiece and / or the remaining starting material are identified depending on the criticality of the respective deformation. For example, a deformation at a point where the workpiece and the remaining starting material are in positive and / or non-positive contact with each other after laser cutting is identified as critical to process safety. A deformation at a point where the workpiece and / or the remaining starting material bulges after laser cutting, welding, or laser surface treatment is also identified as critical to process safety. A deformation at a point where the workpiece is spaced apart from the remaining starting material is identified as not critical to process safety.
[0031] In a further preferred embodiment, it is provided that when a process safety-critical event is output, at least one measure for dealing with this event is suggested. This measure can be suggested by the self-learning algorithm. The self-learning algorithm is preferably trained with regard to outputting measures for dealing with a process safety-critical event by means of a large number of known process safety-critical events and their relationships to each other and to known measures. To determine a relationship between process safety-critical events and / or measures, for example, a manual assignment of the process safety-critical event to a single measure or a set of measures can be carried out by one or more experts.For example, a manual evaluation of process safety-critical events can be provided, which are then prepared and stored for the purpose of training the self-learning algorithm. Preferably, the process safety-critical events and possible measures are evaluated by an expert after the process safety-critical event has been handled and automatically made available to the self-learning algorithm for further development and improvement. Preferably, a large number of evaluated process safety-critical events and / or measures are transmitted via known data transmission to a central data storage system for the execution of the procedure, which serves as the basis for training the self-learning algorithm.
[0032] In a further preferred embodiment, it is provided that, as a measure to deal with the respective process safety-critical event, at least one parameter of the laser processing device and / or a processing sequence of manufacturing process steps is changed. With regard to the parameter change of the laser processing device, the processing head traverse speed, in particular the cutting speed, the laser power of the laser processing device, the focus position of the laser processing device, the composition of a process gas, in particular cutting gas, and / or the fluid pressure of the process gas, in particular cutting gas, can be changed. In particular, it can be provided that the parameter change of the laser processing device is carried out automatically.
[0033] In a further preferred embodiment, the respective process safety-critical event is at least one of the following: - Tilting of the workpiece and the remaining starting material. During laser cutting, the workpiece and / or the remaining starting material expand due to heating, such that a positive and / or force-fit connection is created between the workpiece and the remaining starting material at several critical points, causing them to tilt; - Warping of the workpiece and / or the remaining starting material. Starting from the positive and / or force-fit connection between the workpiece and the remaining starting material, the workpiece and / or the remaining starting material can warp during laser cutting due to further thermal expansion. It is also conceivable that the workpiece warps due to heat input during laser welding or laser surface treatment.
[0034] In a further preferred embodiment, the following steps are provided before the geometry data and the manufacturing data are made available for training the self-learning algorithm with regard to the relationship between the geometry data, the manufacturing data and the temperature change of the workpiece and / or the remaining starting material:
[0035] - Providing further geometry data of another workpiece and further manufacturing data for a further manufacturing process of the further workpiece;
[0036] - Determining further temperature changes of the workpiece and / or the remaining raw material during the manufacturing process, depending on the geometry data and manufacturing data;
[0037] - Training the self-learning algorithm using further geometry data, further manufacturing data and further temperature changes of the further workpiece and / or the remaining raw material.
[0038] The additional geometry data and / or manufacturing data can be real data based on which workpieces have already been manufactured. Alternatively, the additional geometry data and / or manufacturing data can be taken from a database containing data on actually manufactured workpieces.
[0039] The self-learning algorithm is trained on the relationship between further geometry data, further manufacturing data, and further temperature changes of the workpiece and / or the remaining raw material, preferably using a large number of known geometry and manufacturing data points and their relationships to each other and to known temperature changes. To determine a relationship between geometry and manufacturing data and / or a temperature change, for example, one or more experts can manually assign the geometry and manufacturing data to a single temperature change or a set of temperature changes. For example, a manual evaluation of geometry and manufacturing data can be provided, which is then prepared and stored accordingly for the purpose of training the self-learning algorithm.Preferably, the geometry and manufacturing data, as well as any temperature changes following a temperature change, are evaluated by an expert, and this evaluation is automatically provided to the self-learning algorithm for further development and improvement. Preferably, a large number of evaluated geometry and manufacturing data and / or temperature change data are transmitted via known data transmission to the central data storage for the execution of the process, which serves as the basis for training the self-learning algorithm.
[0040] In a further preferred embodiment, a finite element method is performed to determine the further temperature change of the further workpiece and / or the remaining starting material, using the further geometric data and the further manufacturing data. The finite element method can be a thermally transient finite element method.
[0041] In a further preferred embodiment, an evaluation of the additional temperature change of the workpiece determined by the finite element method is performed. This evaluation can be based on the use of temperature measuring devices, particularly tactile ones, which measure actual manufactured workpieces based on the additional geometric data and manufacturing data. The measurement results are then compared with the results of the finite element method. If a threshold for the agreement between the measured and the temperature change determined by the finite element method is exceeded, the temperature change determined by the finite element method is considered sufficiently accurate.
[0042] In a further preferred embodiment, it is provided that the respective manufacturing process includes a laser processing process, during which heat is introduced into the respective workpiece and / or the respective remaining starting material, so that the respective deformation is in the form of a thermal expansion or change in shape of the respective workpiece and / or the respective remaining starting material due to the heat input.
[0043] In a further preferred embodiment, the respective manufacturing data includes at least one member of the following group: parameters of the laser processing device, characteristics of the respective workpiece and / or the respective starting material, information on the minimum distance of each point of the respective workpiece and / or the respective remaining starting material from a contour thereof, and parameters of the manufacturing process. Parameters of the laser processing device may include the trajectory of the laser processing device, the activation time and / or the deactivation time of the laser processing device, the focus position of the laser processing device, the speed of the processing head, in particular the cutting speed, and / or the laser power of the laser processing device.The workpiece and / or the remaining raw material may be characterized by its material thickness and / or material type. Manufacturing process parameters may include the process time and / or the sequence of the process steps.
[0044] In a further preferred embodiment, it is provided that the preceding process steps for training the self-learning algorithm are carried out multiple times with different additional geometry data and different additional manufacturing data for different additional workpieces.
[0045] In a further preferred embodiment, it is provided that at least one process step is carried out automatically.
[0046] The invention also relates to a method for carrying out a manufacturing process, comprising the following process steps: carrying out the above method for predicting at least one process safety-critical event; and carrying out the manufacturing process of the workpiece.
[0047] Further advantages of the invention will become apparent from the claims, the description, and the drawings. Likewise, the features mentioned above and those described in more detail below can each be used individually or in combination in any suitable way according to the invention. The embodiments shown and described are not to be understood as an exhaustive list, but rather serve as examples illustrating the invention.
[0048] Detailed description of the invention and drawing
[0049] Fig. 1 shows, schematically, a method according to the invention for predicting at least one process safety-critical event.
[0050] Fig. 2 shows, schematically, a method according to the invention for carrying out a manufacturing process, which includes the prediction method according to Fig. 1.
[0051] Fig. 3 shows, schematically, abbreviated and not to scale, a process safety-critical event in which a workpiece and a remaining starting material are jammed together due to expansion caused by heat input.
[0052] Fig. 4 shows, schematically, abbreviated and not to scale, the workpiece and the remaining starting material after carrying out a measure to deal with the process safety-critical event from Fig. 3.
[0053] Manufacturing processes that include a laser processing process, preferably a laser cutting process, are state of the art. In these processes, a plate-like workpiece 14 (Figs. 3, 4) can be produced from a plate-like starting material, a so-called sheet, generating a residual starting material 16 (Figs. 3, 4), a so-called residual grid. During the laser processing process, heat is introduced into the starting material, particularly into the workpiece 14. This heat input can cause the workpiece 14 and / or the residual starting material 16 to deform, for example, by expanding and / or changing shape. If the workpiece 14 and / or the residual starting material 16 expands, it can become misaligned with the residual starting material 16 (Fig. 3). It is also conceivable that the workpiece 14 and / or the residual starting material 16 may additionally warp after misalignment (not shown).Furthermore, the manufacturing process includes subsequent process steps following the laser processing, such as picking up and transferring the workpiece 14 to another location for further processing immediately after the laser processing. A workpiece 14 that is jammed with the remaining starting material 16 (Fig. 3) cannot be picked up and subsequently processed further, resulting in a process standstill. Such a process standstill is avoided by the present predictive method according to the invention, the steps of which are carried out before the actual manufacturing of the workpiece 14.
[0054] Fig. 1 shows a method 10 according to the invention for predicting a process safety-critical event 12 (Fig. 3) of a manufacturing process of a workpiece 14 (see Figs. 3, 4).
[0055] The inventive method 10 comprises the following process steps, which are carried out before the actual manufacturing of the workpiece 14: In a first process step 18, geometry data of the workpiece 14 and manufacturing data for the manufacturing process of the workpiece 14 are provided.
[0056] The manufacturing data can include parameters of a laser processing device, characteristics of the respective workpiece 14 and / or the respective starting material, and / or parameters of the manufacturing process. Parameters of the laser processing device can include the activation time and / or the deactivation time of the laser processing device, the focus position of the laser processing device, the speed of the processing head, in particular the cutting speed, and / or the laser power of the laser processing device. Characteristics of the workpiece 14 and / or the remaining starting material 16 (see Figs. 3, 4) can include the material thickness of the workpiece 14 and / or the remaining starting material 16 and / or the material type of the workpiece 14 and / or the remaining starting material 16.Parameters of the manufacturing process can include the process time and / or the processing sequence of the process steps of the manufacturing process.
[0057] The geometry data and the manufacturing data can be provided in such a way that a user of procedure 10 makes this data available to a computing unit (not shown) for the execution of procedure 10. For example, this provision can be done via a human-machine interface (not shown) in the form of a touchscreen or a keyboard. Alternatively or additionally, the geometry data can be provided by supplying a manufacturing plan that includes the geometry data. The manufacturing plan can, for example, be transmitted digitally to the computing unit by an order management unit.
[0058] In a second process step 20, at least one deformation 22 (Fig. 3) of the workpiece 14 and / or the remaining starting material 16 is determined in relation to a predetermined geometry according to geometric data during the manufacturing process, depending on the manufacturing data. The respective deformation 22 can be caused by heat input into the workpiece 14 and / or the remaining starting material 16 by the laser processing method.
[0059] To determine the respective deformation 22, a temperature change of the workpiece 14 and / or the remaining starting material 16 during the manufacturing process is determined based on the geometry data and the manufacturing data. The temperature change can be spatially and / or temporally resolved. The spatial and / or temporal temperature change can be represented as a so-called heat map, which depicts the heating of the workpiece 14 and / or the remaining starting material 16. The temperature change of the workpiece 14 and / or the remaining starting material 16 can be determined using a self-learning algorithm.
[0060] Depending on the temperature change of the workpiece 14 and / or the remaining starting material 16, the respective deformation 22 is calculated.
[0061] In a third process step 28, the respective process safety-critical event 12 of the manufacturing process is predicted depending on the respective deformation 22 of the workpiece 14 and / or the remaining starting material 16.
[0062] To predict the respective process safety-critical event 12, a criticality of the respective deformation 22 is determined. The criticality can be defined by at least one lookup table that assigns a criticality to the respective deformation 22. The criticality can include at least the ratings (process safety-)critical and (process safety-)non-critical. Depending on the criticality of the respective deformation 22, critical points 24 (see Fig. 3) of the workpiece 14 and / or the remaining starting material 16 are identified. For example, a deformation 22 at a point 24 where the workpiece 14 and the remaining starting material 16 are in form-fit and / or force-fit contact with each other after laser processing is identified as process safety-critical (Fig. 3).For example, a deformation at a point where the workpiece 14 and / or the remaining starting material 16 bulge after laser processing (not shown) is identified as critical to process safety. Conversely, a deformation 22 at a point 25 (see Fig. 3) is identified as not critical to process safety if the deformation 22 is present but the workpiece 14 is spaced apart from the remaining starting material 16.
[0063] As a result, the respective process safety-critical event 12 can be a tilting of the workpiece 14 and the remaining starting material 16. Figure 3 shows such a tilting of the workpiece 14 with the remaining starting material 16 due to heat input into the workpiece 14 and / or the remaining starting material 16. The workpiece 14 is in positive and / or force-fit contact with the remaining starting material 16 at several critical points 24, so that no cutting gap 26 remains free at these points 24.
[0064] In a fourth process step 30, the respective process safety-critical event 12 is output. Output of the respective process safety-critical event 12 can be achieved via a human-machine interface. The human-machine interface can be configured as a display to visualize the process safety-critical event 12. The display can be a screen. When the respective process safety-critical event 12 is output, at least one measure for dealing with this event 12 is suggested. The measure can be suggested by a self-learning algorithm.
[0065] Alternatively or in addition to issuing the process safety-critical event 12, the manufacturing process is adapted to the respective process safety-critical event 12 by implementing at least one of the measures for dealing with this event.
[0066] As a measure to address the process safety-critical event 12, at least one parameter of the laser processing device is modified. This parameter modification can include the processing head traverse speed, in particular the cutting speed, the laser power of the laser processing device, the focus position of the laser processing device, the composition of a process gas, in particular cutting gas, and / or the fluid pressure of the process gas, in particular cutting gas. This modification of at least one parameter of the laser processing device is intended to reduce the heat input into a predicted process safety-critical location 24 of the workpiece 14 and / or the remaining starting material 16 during the actual laser processing.This prevents such deformation 22 of the workpiece 14 and / or the remaining starting material 16 at the predicted process safety-critical point 24 from occurring in the first place, such that the workpiece 14 becomes jammed with the remaining starting material 16.
[0067] As a measure to address the predicted critical process safety event 12, the processing sequence of the manufacturing process can be changed, either alternatively or additionally. This change aims to allow a critical process safety point 24 of the workpiece 14 and / or the remaining raw material 16, which is heated by heat input during the actual laser processing process, to cool down by prioritizing other process steps of the manufacturing process before picking up the workpiece 14. This allows any tilting of the workpiece 14 with the remaining raw material 16 to resolve itself through cooling. Figure 4 shows a workpiece 14 and the remaining raw material 16 after cooling as a measure to circumvent the critical event 24. Here, the workpiece 14 is completely separated from the remaining raw material 16, forming a cutting gap 26.This allows workpiece 14 to be picked up and taken to another location for further processing.
[0068] Before providing the geometry data and the manufacturing data, the following steps are carried out to train the self-learning algorithm with regard to the relationship between the geometry data, the manufacturing data and the temperature change of the workpiece 14 and / or the remaining starting material 16:
[0069] In a first step, 32 additional geometry data for another workpiece and further manufacturing data for a further manufacturing process of the further workpiece are provided.
[0070] In addition to the manufacturing data, further manufacturing data can include: information about the minimum distance of each point of the workpiece and / or the remaining raw material from a contour of the workpiece. The trajectory of the laser processing device can also be included as an additional parameter in the training of the self-learning algorithm.
[0071] Subsequently, in a second step 34, a further temperature change of the further workpiece and / or the remaining raw material during the further manufacturing process is determined as a function of the further geometric data and the further manufacturing data. To determine the further temperature change of the further workpiece and / or the remaining raw material, a finite element method is performed using the further geometric data and the further manufacturing data.
[0072] Subsequently, an evaluation of the further temperature change of the remaining workpiece and / or the remaining raw material, as determined by the finite element method, can be performed. This evaluation can be based on the use of temperature measuring devices (not shown), particularly tactile ones, which measure actual manufactured workpieces. The measurement results are then compared with the results of the finite element method. If a threshold for the agreement between the measured and the finite element-determined temperature change is exceeded, the temperature change determined by the finite element method is considered sufficiently accurate. The temperature change can be spatially and / or temporally resolved. The spatial and / or temporal temperature change can be presented in the form of a heat map.
[0073] Finally, in a third step 36, the self-learning algorithm is trained using the additional geometry data, the additional manufacturing data and the additional temperature change of the additional workpiece and / or the additional remaining starting material.
[0074] The first to third steps 32, 34, 36 are preferably carried out multiple times with different additional geometry data and different additional manufacturing data for different additional workpieces.
[0075] Fig. 2 shows a method 38 for carrying out a manufacturing process. The method 38 for carrying out the manufacturing process comprises the following steps: In a first process step 40, the method for predicting at least one process safety-critical event 12 (Fig. 3) is carried out. In a second process step 42, the manufacturing process of the workpiece 14 (Figs. 3, 4) is carried out from the starting material 16 (Figs. 3, 4).
[0076]
[0077] 10 methods for prediction
[0078] 12. Process safety-critical event
[0079] 14 workpieces
[0080] 16 remaining starting material
[0081] 18 First procedural step: Prediction procedure
[0082] 20 Second procedural step: Prediction procedure
[0083] 22 Deformation
[0084] 24 process safety-critical points
[0085] 25 non-process safety-critical points
[0086] 26 cutting gap
[0087] 28 Third procedural step: Prediction procedure
[0088] 30 Fourth procedural step Prediction procedure
[0089] 32 First step: Prediction procedure
[0090] 34 Second step: Prediction procedure
[0091] 36 Third step: Prediction procedure
[0092] 38 methods for carrying out a manufacturing process
[0093] 40 First procedural step Procedure for carrying out
[0094] 42 Second procedural step Procedure for carrying out
Claims
Patent claims 1. Method (10) for predicting at least one process safety-critical event (12) of a manufacturing process of a workpiece (14) to be manufactured from a starting material, in particular using a laser processing device, comprising the following steps carried out before the manufacturing process of the workpiece (14): - Providing (18) geometry data of the workpiece (14) and manufacturing data for the manufacturing process of the workpiece (14); - Determining (20) at least one deformation (22) of the workpiece (14) and / or the remaining starting material (16) in relation to a predetermined geometry according to geometry data during the manufacturing process depending on the manufacturing data; - Predictions (28) of the respective process safety-critical event (12) of the manufacturing process depending on the respective deformation (22) of the workpiece (14) and / or the remaining starting material (16); - Output (30) of the respective process safety-critical event (12) and / or Adapting the manufacturing process to the respective process safety-critical event (12) by implementing at least one measure to deal with this event (12).
2. Method (10) according to claim 1, characterized in that, in order to determine (20) the respective deformation (22) based on the geometry data and the manufacturing data, a temperature change of the workpiece (14) and / or the remaining starting material (16) is determined during the manufacturing process.
3. Method (10) according to claim 2, characterized in that the temperature change of the workpiece (14) and / or the remaining starting material (16) is determined by means of a self-learning algorithm.
4. Method (10) according to claim 2 or 3, characterized in that the respective deformation (22) is determined, in particular calculated, as a function of the temperature change of the workpiece (14) and / or the remaining starting material (16).
5. Method (10) according to one of the preceding claims, characterized in that a criticality of the respective deformation (22) is determined in order to predict (28) the respective process safety-critical event (12).
6. Method (10) according to claim 5, characterized in that critical points of the workpiece (14) and / or the remaining starting material (16) are identified depending on the criticality of the respective deformation (22).
7. Method (10) according to one of the preceding claims, characterized in that when the respective process safety-critical event (12) is output (30), at least one measure for dealing with this event (12) is proposed.
8. Method (10) according to one of the preceding claims, characterized in that, as a measure to deal with the respective process safety-critical event (12), at least one parameter of the laser processing device and / or a processing sequence of manufacturing process steps is changed.
9. Method (10) according to one of the preceding claims, characterized in that the respective process safety-critical event (12) is at least one of the following events: tilting of the workpiece (14) and the remaining starting material (16); warping of the workpiece (14) and / or the remaining source material (16).
10. Method (10) according to one of the preceding claims, characterized in that, prior to providing the geometry data and the manufacturing data for training the self-learning algorithm with regard to the relationship between the geometry data, the manufacturing data and the temperature change of the workpiece (14) and / or the remaining starting material (16), the following steps are carried out: - Providing further geometry data of another workpiece and further manufacturing data for a further manufacturing process of the further workpiece; - Determining further temperature changes of the workpiece and / or the remaining raw material during the manufacturing process, depending on the geometry data and manufacturing data; - Training the self-learning algorithm using further geometry data, further manufacturing data and further temperature changes of the further workpiece and / or the remaining raw material.
11. Method (10) according to claim 10, characterized in that a finite element method is carried out to determine the further temperature change of the further workpiece and / or the further remaining starting material using the further geometry data and the further manufacturing data.
12. Method (10) according to claim 11, characterized in that an evaluation of the further temperature change of the further workpiece and / or the further remaining starting material determined by the finite element method is carried out.
13. Method (10) according to one of the preceding claims, characterized in that the manufacturing process and / or the further manufacturing process comprises a laser processing process in the course of which heat is introduced into the respective workpiece (14) and / or the respective remaining starting material (16), so that the respective deformation (22) is in the form of a thermal expansion or change in shape of the respective workpiece (14) and / or the respective remaining starting material (16) due to the heat input.
14. Method (10) according to one of the preceding claims, characterized in that the manufacturing data and / or the further manufacturing data comprise at least one member of the following group: parameters of the laser processing device, characteristic values of the respective workpiece and / or the respective starting material, information about a minimum distance of each point of the respective workpiece (14) and / or the respective remaining starting material (16) from a contour thereof, parameters of the manufacturing process.
15. Method (38) for carrying out a manufacturing process, comprising the following process steps: - Performing (40) the method for predicting at least one process safety-critical event (12) according to any one of the preceding claims; and - Carrying out (42) the manufacturing process of the workpiece (14).