Method for carrying out a forming process, computer program and electronically readable data carrier

Abstract

The invention relates to a method for carrying out a forming process in which a drawn part (3) is formed from a semi-finished product (1) by means of a deep-drawing device (2), comprising the steps: - Detection of a property value at at least one measuring point of the drawn part (3) by a sensor device; (S1) - Determining the current material flow during the forming process from the semi-finished product (1) to the drawn part (3) based on the property value of at least one measuring point of the drawn part; (S2) - Determining a deviation of the current material flow from a reference material flow, which depends on at least one semi-finished product property and / or at least one process parameter and for which a drawn part quality is known; (S3) - When a threshold value is exceeded, which indicates an expected drawn part quality based on the current material flow, by a certain deviation: output a signal and / or specify at least one change to process parameters for at least one upcoming forming process. (S4) Furthermore, the invention relates to a computer program and a data carrier.

Description

[0001] The invention relates to a method for carrying out a forming process in which a drawn part is formed from a semi-finished product by means of a deep-drawing device, in particular a press, according to patent claim 1. Furthermore, the invention relates to a computer program which can be executed, for example, directly in a memory of an electronic computing device of the deep-drawing device, as well as to an electronically readable data carrier according to patent claims 10 and 11.

[0002] A component of a motor vehicle, such as a body component, is usually formed from a drawn part, which in turn is formed from a semi-finished product by means of a forming process, such as deep drawing in a deep-drawing device. Such a deep-drawing device, which can in particular be designed as a press or pressing shop, processes several semi-finished products or thin sheets one after the other for series production of the drawn parts. Fluctuations in the semi-finished products or thin sheets can occur; for example, a sheet thickness, a lubricant quantity, a roughness and / or an elasto-plastic material property can fluctuate. Depending on the extent of this fluctuation, it may be necessary to adapt process parameters of the production process or forming process in order to achieve the required quality of the produced drawn parts and thus of the components formed from them.

[0003] In the current state of the art, process parameters are usually adjusted during a plant downtime and based on experience by a plant operator. Depending on how long it takes to find a suitable combination of process parameters that can achieve the required quality, considerable costs can arise from the downtime of the production or forming process.

[0004] The object of the present invention is to provide a method for carrying out a forming process, a computer program and a corresponding data carrier by means of which the quality of a drawn part can be determined in a simple manner and, in the event of deviations from the desired quality, appropriate countermeasures for further forming processes can be carried out.

[0005] This object is achieved according to the invention by the subject matter of the independent patent claims. Advantageous embodiments and further developments of the invention are specified in the dependent patent claims as well as in the description and the drawings.

[0006] A first aspect of the invention relates to a method for carrying out a forming process in which a drawn part is formed from a semi-finished product which has at least one semi-finished product property and is in particular a thin sheet, by means of a deep-drawing device, in particular a press or a pressing plant.

[0007] Deep drawing is typically performed using tools such as a die, a blank holder, and a punch mounted in the press. The shape of the die and punch determines the geometry of the drawn part, which represents the result of the deep drawing process and can be an intermediate product on the way to a component, such as a car body part.

[0008] The method according to the invention comprises the following steps: In a first step, a property value or a point property is recorded at at least one measuring point or measuring position of the drawn part by a sensor device.

[0009] In a second step, a current material flow during the forming process from the semi-finished product to the drawn part is determined based on the property value of at least one measuring point of the drawn part.

[0010] In a third step, a deviation of the currently determined material flow from a reference material flow, in particular stored in a database, is determined, which reference material flow characterizes in particular previous forming processes, depends on at least one semi-finished product property and / or on at least one process parameter and / or at least one state variable and for which a drawn part quality is known or which describes a drawn part quality.

[0011] In a fourth step, when a threshold value is exceeded, which indicates or describes an expected drawn part quality based on the current material flow, the determined deviation results in the output of a signal and / or the specification or initiation of at least one change to process parameters, which characterize at least one upcoming forming process of a deep-drawing device, for the upcoming forming process.

[0012] The measuring point of the drawn part or on the drawn part is a measuring point at which the property value or at least one property value is determined. The property value can, for example, be a position of the measuring point and thus describe, in particular, a drawn part edge and / or, for example, a lead edge and / or a near-running edge. The property value thus describes a property of the drawn part, which can be used to describe the material flow during forming.

[0013] The at least one state variable or a process state variable describes a process state, such as a temperature of tool surfaces that are in contact with the sheet metal during forming.

[0014] Material flow occurs when the punch and die, between which the sheet metal or semi-finished product is located during the forming process, move towards or into each other, whereby a relative movement occurs between the semi-finished product and the surface of the die, the punch and a sheet metal holder, whereby the relative movement of the material represents the material flow. In the area of ​​the sheet metal holder, a relative movement can occur between the sheet metal and the sheet metal holder, since there is a contact pressure between the die, semi-finished product and sheet metal holder, which allows a relative movement between the semi-finished product and the tool surfaces. The material flow can therefore be measured by measuring the position of the edge of the intermediate product or drawn part after it has been formed. Due to the fact that the drawn part deforms after forming, which is known as elastic springback or bounce, the material flow should be measured relative.For example, the distance between the edge of the drawn part and the beginning of the die entry radius in the direction of material flow is considered and used as a measure of material flow. This measure is invariant with respect to elastic springback. Similarly, the positions of the leading and trailing edges should also be determined relatively.

[0015] The lead-in edge is the beginning of a mark in the contact area between the tool and semi-finished product surface due to relative movement. If the material flow changes, the position of the lead-in edge on the drawn part also changes. When the lead-in edge is formed, a trailing edge is also induced on the opposite side of the semi-finished product or sheet. The material flow can be derived from the position of the lead-in edge or near-running edge as well as the edge of the drawn part. Advantageously, more than one measuring point is used; in particular, measuring points can be provided at advantageous points, such as the edge area of ​​the semi-finished product or drawn part and at corresponding edges.

[0016] Material flow can be considered a key parameter for controlling the quality of the deep drawing or forming process. A material flow analysis can be used automatically or manually to select process parameters. For example, excessive material flow in the area of ​​the blank holder can lead to the formation of wrinkles. Conversely, insufficient material flow may lead to cracking. This relationship allows material flow to be considered a proxy parameter for the quality of the drawn part and thus for the quality of the drawn part and thus for the quality of the final component.

[0017] Thus, the threshold value or a deviation from a reference value of the material flow, which may have been determined in particular in previous forming processes or may be based on longer series of measurements, allows the quality to be directly estimated and, in the event of a deterioration in the quality of the drawn part, to easily warn the user by means of the signal or to take countermeasures.

[0018] In other words, after the semi-finished product has been formed into a drawn part, the material flow occurring during the forming process is recorded by a sensor device, for example a camera, and compared with a reference material flow. This reference material flow is based on empirical values ​​and formed processes from previous production processes, in which a desired or required drawn part quality is achieved, in particular with the same process parameters and / or state variables with which the deep-drawing device is operated, and with the same or similar semi-finished product properties of the semi-finished product currently being formed. The reference material flow can, for example, represent an average of several forming processes or describe a specific interval.

[0019] The quality of the formed drawn part can be easily estimated by comparing the current material flow with the reference material flow. If this estimation shows that the drawn part quality is not within the desired range and the deviation thus exceeds a threshold value characterizing the range, appropriate countermeasures or warnings are issued. The reference material flow depends on at least one semi-finished product property and / or at least one process parameter and / or at least one state variable. Thus, for example, different reference material flows can be maintained for different values ​​of the at least one semi-finished product property.

[0020] The reference material flow can be provided, in particular, by analyzing data and determining the deviation. The reference material flow can also be derived, for example, from previously recorded data or the data sets created for each drawn part formed on the deep-drawing device. The respective data set can include semi-finished product properties of the underlying semi-finished product and / or process parameters used during the associated forming process and / or during the production of the semi-finished product in a coiling system, and / or state variables and the material flow.

[0021] This results in the advantage that the method can be used to determine the quality of the part formed by the deep-drawing device, the drawn part or the resulting component, in a particularly advantageous and simple manner and, if necessary, countermeasures can be carried out for future forming processes in the form of adjustment of the process parameters in order to be within the range of the desired drawn part quality in the following forming processes.

[0022] An advantageous embodiment of the invention maintains a data record for each drawn part already formed, in particular stored and / or maintained in a database. Additionally or alternatively, data records characterizing the reference material flow are assigned, which were used in particular during its creation. Additionally or furthermore, the reference material flow used to determine the deviation is selected from the stored data records by filtering, so that, for example, the reference material flow is formed only from data records or is based on data records whose semi-finished product properties meet certain criteria. The data record can be assigned to the respective semi-finished product even after the forming process, in particular via a serial number of the semi-finished product.

[0023] The respective data set contains, in particular, semi-finished product properties, such as a sheet thickness, a lubricant quantity, and / or elastoplastic material properties. Furthermore, the data set can include, in particular, process parameters, such as a pressure of the displacement cylinders of the sheet holder, a contact pressure, and so on. Additionally or alternatively, the data set contains the property value of at least one measuring point and / or state variables, or at least in particular one state variable, such as an ambient temperature of the deep-drawing device and / or a tool temperature of the deep-drawing device, air pressure, or the like.

[0024] The specific criterion of the semi-finished product property is in particular a similarity of the semi-finished product properties or advantageously an identity of the semi-finished product properties of the semi-finished products which form the basis or foundation of the reference material flow to the semi-finished product properties of the currently formed drawn part.

[0025] This provides the advantage of the most comprehensive data collection possible, thus creating a database that allows for advantageous analyses of the material flow and / or other conditions during the forming process. Another advantage, particularly of filtering, is that the material flow can be compared with particularly relevant data from previous forming processes.

[0026] In a further advantageous embodiment of the invention, the reference material flow is determined by specifying process parameters or a process parameter combination for at least one forming process. At least one semi-finished product is formed using the deep-drawing device using these specified process parameters. The material flow determined in this way is used as the reference material flow once the drawn part quality has been achieved. In other words, particularly if there is not yet enough data available to create a reference material flow, which, for example, describes an average value of several deep-drawing processes with the same semi-finished product properties and / or process parameters, a process parameter combination is specified. This combination is then used to form a semi-finished product into a drawn part. If this semi-finished product achieves the desired drawn part quality with the specified process parameters, the material flow can be specified as the reference material flow.This has the advantage that the process can be applied, for example, shortly after starting a forming process.

[0027] In a further advantageous embodiment of the invention, the reference material flow is determined by means of a simulation, in particular using the finite element method. In other words, for example, if a sufficient amount of data is not yet available or a new forming process is to be commissioned, the forming process is modeled using a computer simulation, and the material flow in the drawn part virtually formed by the simulation is determined. This offers the advantage that the method can be used immediately even when commissioning a new forming process.

[0028] In a further advantageous embodiment of the invention, the reference material flow is determined by selecting a plurality of data sets based on semi-finished product properties, process parameters, and / or state variables, and averaging the respective property values ​​at at least one mutually corresponding measuring point of the data sets before the reference material flow is determined, such that the reference material flow is determined from the averaged point properties. As an alternative to averaging and thus to the mean value, other statistical key figures can be used. In other words, the reference material flow is a set of data sets, each of which contains a reference to a measuring position as well as reference values. For one measuring point, a plurality of data sets can be determined, which describe the dependence of the reference values ​​on semi-finished product properties, process parameters, state variables of the press line, or the forming device.Subsequently, a confidence level describing the accuracy of the calculated reference material flow can be determined for each data set. This provides the advantage of allowing the reference material flow to be determined particularly effectively.

[0029] In a further advantageous embodiment of the invention, the process parameters are specified when the threshold value is exceeded and / or the process parameters for determining the reference material flow are specified using a machine learning method. In other words, artificial intelligence in the form of a machine learning algorithm and / or a neural network is used to determine, based on boundary conditions such as semi-finished product properties and / or drawn part quality, the process parameters with which the deep-drawing device can or should be operated. The machine learning method can be trained using training data, which, for example, is obtained from data sets of previous forming processes and / or using simulation data, such as, for example, by simulating the aforementioned finite element method.This has the advantage that process parameters can be output in a particularly simple manner without any user intervention.

[0030] In a further advantageous embodiment of the invention, the data sets are weighted during filtering, in particular based on a time of the respective associated forming process. In other words, during filtering, which specifies which data sets are used to generate the reference material flow, the individual data sets are weighted, in particular based on the time of the forming process. For example, it is advantageous to assign a higher weight to data sets that were recently generated during a forming process, as this makes it easier to counteract tool wear due to frequent forming processes.For example, tool wear may already have been compensated for by adjusting process parameters, and this information is at least inherently contained in the data sets, although, for example, a gradual change in the process parameters may also occur, which can be advantageously compensated for by weighting or filtering.

[0031] In an advantageous initial situation of the invention, the property value of the respective measuring point is characterized by the position of a drawing edge, a lead-in edge, and / or a trailing edge. In other words, the property value includes information from which the material flow can be easily deduced, such as the position of the drawing edge, the positions and / or shapes and sizes of the lead-in edge and / or the trailing edge. This results in the advantage that the method can be used particularly efficiently.

[0032] In a further advantageous embodiment of the invention, the property value of the respective measuring point is recorded using an optical sensor and / or a mechanical contact sensor. In other words, a camera or another optical sensor is used, for example, through which the drawn part can be imaged and which detects corresponding drawn edges and / or contact edges and / or near-running edges using computer vision. Additionally or alternatively, the drawn part edge, for example, can be detected using a mechanical contact sensor. This offers the advantage that the property value(s) can be recorded advantageously using simple, fail-safe sensors.

[0033] In a further advantageous embodiment of the invention, the current material flow is determined by determining a position of the edge of the drawn part, which thus represents, for example, the property value stored for the respective measuring point and / or relative to two edges of the drawn part, wherein one of the edges is invariant with respect to the material flow. In other words, the material flow can be determined in particular at critical points or measuring positions, wherein a measurement of the material flow is carried out relative to the drawn part between two edges. The position of one of the two edges, which serves as a reference for the material flow measurement, should be inherent in relation to the material flow. Thus, the distance between two movable and / or stationary edges or edges that are invariant with respect to the material flow serves as a measure of the material flow.This results in the advantage that the method can be used particularly advantageously for determining the material flow and thus for determining the quality of the drawn part.

[0034] For use cases or application situations that may arise during this procedure and are not explicitly described here, it is possible to display an error message and / or prompt the user for feedback. Furthermore, a default setting or a predefined initial state can be set.

[0035] A second aspect of the invention comprises a computer program. The computer program can, for example, be loaded into a memory of the electronic computing device, in particular a deep-drawing device that performs the forming process, and comprises program means for executing the steps of the method when the computer program is executed in the electronic computing device or a control device.

[0036] Advantages and advantageous embodiments of the first aspect of the invention are to be regarded as advantages and advantageous embodiments of the second aspect of the invention and vice versa.

[0037] A third aspect of the invention relates to an electronically readable data carrier. The electronically readable data carrier comprises electronically readable control information stored thereon, which comprises at least one computer program as just presented and is configured such that, when the data carrier is used in an electronic computing device, it can execute a method presented here according to the first aspect of the invention.

[0038] Advantages and advantageous embodiments of the third aspect of the invention are to be regarded as advantages and advantageous embodiments of both the second and the first aspect of the invention and vice versa.

[0039] Further features of the invention emerge from the claims, the figures, and the description of the figures. The features and combinations of features mentioned above in the description, as well as the features and combinations of features mentioned below in the description of the figures and / or shown alone in the figures, can be used not only in the respective specified combination, but also in other combinations or on their own.

[0040] The invention will now be explained in more detail using a preferred embodiment and with reference to the drawings. It shows: Fig. 1 is a schematic flow diagram for a method for carrying out a conversion process in which a drawn part is formed from a semi-finished product by means of a deep-drawing device; and Fig. 2 in a schematic view the deep-drawing device in a first state before the conversion process and in a second state after the conversion process of the drawn part.

[0041] Fig. Figure 1 shows a schematic flow diagram for a method for carrying out a forming process in which a drawn part 3 is formed from a semi-finished product 1 by means of a deep-drawing device 2. The method presented comprises the steps S1 to S4: In step S1, a property value is detected at at least one measuring point of the drawn part 3 by a sensor device. In a step S2, a current material flow during the forming process from the semi-finished product 1 to the drawn part 3 is determined based on the property value of at least one measuring point. In step S3, a deviation of the current material flow from a reference material flow is determined, which depends on at least one semi-finished product property and / or at least one process parameter and / or at least one state variable and for which a drawn part quality is known. In a fourth step S4, when a threshold value, which indicates an expected drawn part quality based on the current material flow, is exceeded by the deviation, a signal is output and / or at least one change to process parameters for upcoming forming processes is specified for at least one upcoming forming process.

[0042] During the forming process or deep drawing operation, as in Fig. 2, tools are typically used which comprise at least a die 4, a sheet holder 5 and a punch 6, which are mounted in a press or the deep-drawing device 2. The die 4 is firmly connected to a ram 7. The punch 6 can be mounted on a press table, and the sheet holder 5 can be connected to displacement cylinders 8 which move under the action of force. At the start of the forming or forming process, a flat sheet, the semi-finished product 1, is placed on the sheet holder 5. The ram 7 then moves downwards with the die 4. As soon as the sheet holder 5, the semi-finished product 1 and the die 4 are in contact, the sheet holder 5 is displaced downwards by the displacement cylinders 8.The force required to move the displacement cylinders 8, which determines the contact pressure between the die 4, the semi-finished product 1, and the blank holder 5, can be adjusted, for example, by means of or as a process parameter. During the displacement of the blank holder 5, the semi-finished product 1 is pulled over the punch 6 and thereby formed into the drawn part 3.

[0043] The shape of the die 4 and the punch 6 determines the geometry of the drawn part 3, which represents the result of the forming process and can be viewed as an intermediate product for a component to be formed from the drawn part 3. During the forming or the forming process, the semi-finished product 1 or sheet metal is plastically stretched. In addition, a relative movement can also occur between the semi-finished product 1 or sheet metal and the surface of the die 4, the punch 6 and / or the sheet metal holder 5. This is referred to as the material flow. The material flow can be determined by measuring the position of the edge of the drawn part 3 after forming or can be detected by the sensor device. This edge can be referred to as the drawn part edge.

[0044] If the material flow increases, the distance, in particular the orthogonal distance, between the edge of the drawn part 3 and the outline of the punch 6 decreases. However, the material flow can occur not only in the area of ​​the sheet holder 5. In principle, on sufficiently slightly curved surfaces in the contact area between the punch 6, the die 4 and / or the sheet holder 5 and the sheet surface or the surface of the semi-finished product 1, relative movements occur in the contact area, which can in particular result in visible marks on the sheet surface. These marks are in particular marks in the form of scratches, which can arise due to the relative movement between the tool surfaces or the surface of the punch 6 and the die 4 and the sheet surface.

[0045] This can result in a lead-in edge, which is the beginning of the aforementioned mark. If the material flow changes, the position of the lead-in edge on the intermediate product also changes. When the lead-in edge is formed, a trailing edge is always induced on the opposite side of the semi-finished product 1 or drawn part 3. The position of this lead-in edge and / or trailing edge can also generally be used to assess the material flow.

[0046] It is therefore advantageous for the method if the property value of the respective measuring point characterizes a position of the drawn edge, a formed starting edge and / or a formed near-running edge of the drawn part 3. It is also advantageous if the property value is measured or determined by means of at least one optical sensor and / or a mechanical contact sensor of the sensor device. In principle, various methods are available for determining the material flow. For example, optical sensors can be used to determine the relative position of an edge to a spatially fixed edge on the drawn part 3. This method is therefore suitable for determining the position of the drawn edge as well as starting and near-running edges. Furthermore, there are methods for determining the position of the edge of the drawn part that require mechanical contact between the sensor and the surface of the material.

[0047] The procedure according to Fig. 1 is based on the realization that material flow can be a key variable for controlling the quality of the deep-drawing process. It should also be emphasized that material flow analysis can also be used with manual selection of process parameters. For example, excessive material flow in the area of ​​the blank holder 5 can lead to the formation of wrinkles. Conversely, insufficient material flow may lead to cracking. This relationship makes it possible to consider material flow as a proxy for the quality of the component and thus also estimate the drawn part quality of the drawn part 3, or vice versa.

[0048] For example, the material flow can be influenced by changing the contact pressure distribution between the blank holder 5 and the die 4. The contact pressure distribution represents, for example, a value or a state under which the deep-drawing device 2 is operated. Thus, the contact pressure can be specified by selecting the forces of the displacement cylinders 8, which can represent a process parameter acting on the blank holder 5, and / or by adjusting the drawing aids.

[0049] The respective process parameter combination, which summarizes at least all recorded process parameters for a single forming process, can be combined into a vector, whereby each component of the vector can represent a process parameter of the process parameter combination as a value. If the values ​​of at least one component of two vectors differ, these are two different process parameter combinations, whereby the respective process parameter combination describes the process parameters used for the respective forming process or summarizes them into the vector. Thus, with the additional use of a lubricant application, the number of components of this respective vector or of the respective process parameter combination can comprise more than 1000 entries.

[0050] It is advantageous for the process that after the intermediate product is removed from the drawing stage or the deep-drawing device 2, the material flow is measured, and the results are recorded and stored in a database. Thus, the measurement data regarding the material flow can be stored for the individually manufactured drawn part 3.

[0051] Advantageously, the material flow is measured at discrete positions, the measuring points. These positions can be determined based on experience and / or finite element simulations. It is also advantageous, especially when only a few data sets are available for determining the reference material flow, to determine it by simulation, particularly using the finite element method. Measurements of previous real forming processes are generally preferable.

[0052] For example, a data set can be stored for each forming process and / or kept in a database, which contains, in particular, semi-finished product properties, process parameters, state variables, the property value of the respective material flow and / or the material flow itself and / or other information. The semi-finished product properties and direct or indirect data on the component quality and the associated intermediate products, such as the drawn part 3, as well as process parameters, can be recorded in a coiling system and / or the deep-drawing device 2 and assigned to a semi-finished product 1 and a drawn part 3, respectively. The assignment can be made, for example, via a serial number that is applied to the material during cutting of the semi-finished product 1 or the blank of the coiling system. Furthermore, measurement results that characterize the component quality or drawn part quality can also be saved.This data can also be used for an assistance system. For example, it can be advantageous to use machine learning methods to specify the process parameters when the threshold is exceeded and / or to specify the process parameters for determining the reference material flow.

[0053] For example, in order to benefit from measuring the material flow during the forming process, the measurement can be compared with the reference measurement or the reference material flow. In other words, the measurement data is interpreted by comparing the instantaneous or current material flow, for example, in the form of the current position of the drawing edge of the drawn part 3 or its leading and / or near-running edges, with a reference material flow.

[0054] A key challenge here is determining the reference material flow, as the material flow depends on semi-finished product properties and the selected process parameters and / or state variables. The semi-finished product properties within a coil are not necessarily constant. For example, due to usual delivery conditions, several different cold-rolled strips may be joined by welding to form a coil from a supplier. Different cold-rolled strips often have different semi-finished product properties. For example, within a blank stack of blanks or semi-finished products 1 made from different cold-rolled strips, different semi-finished product properties may exist. Furthermore, changes in state variables, such as the surface temperature of tools, e.g., die 4, affect the material flow.

[0055] For example, certain specifications for drawn part quality cannot be met even with constant process parameters and the same semi-finished product properties. Therefore, it is advantageous to determine the reference material flow, taking the variability of these parameters into account. This way, both the desired material flow and a tolerance to be maintained are taken into account as the reference for the material flow or reference material flow.

[0056] When determining the material flow, it is found that there are usually only a small number of process parameter combinations, since a change in process parameters leads to downtimes in the deep-drawing device 2 and these are only changed in practice if quality problems in the production of the drawn part 3 are adjusted, so a change in the process parameters in a critical process in series production of motor vehicle body parts usually rarely occurs.

[0057] In contrast, as already mentioned above, there is a relatively large number of process parameters, which can easily reach four digits. This results in a particularly small number of points in a space of process parameters, which significantly complicates the analysis of the relationship between process parameters and material flow. Furthermore, it may turn out that semi-finished products 1 are ready for processing whose properties have not yet even occurred. In such a case, a reference material flow cannot be specified directly and, accordingly, upon the occurrence of such a state, cannot be prepared by the process without first specifying or determining a reference material flow.

[0058] In principle, it is possible to determine the reference material flow through simulations. However, simulation models often do not take into account the elastic behavior of the entire system, consisting of tools and press. Thus, for example, effects resulting from adjusting drawing aids cannot be derived. Furthermore, the variation in the elasto-plastic behavior of semi-finished products from different batches can only be modeled with considerable effort.

[0059] Therefore, a measurement data-based determination of the reference material flow is preferred. For example, the material flow can simply be averaged across many produced components. However, due to the above-mentioned dependencies of the material flow on semi-finished product properties, process parameters and / or state variables, this can lead to an unfavorable reference. Therefore, for example, filtering is advantageously carried out when determining the reference material flow so that the reference material flow is only formed from data sets whose semi-finished product properties are similar to the criteria or the semi-finished product properties currently in use. Furthermore, the filtering can involve weighting the data sets, in particular based on a time point in the respective forming process, so that tool wear, for example, can be taken into account.

[0060] When determining reference values ​​that form the reference material flow, methods from statistics can be used, for example, since it makes little sense to derive the reference material flow based on a single drawn part 3. For example, it can be a mean value that directly describes the desired material flow and / or quantiles can be used to characterize a scatter band of the permissible material flow. A reference material flow is therefore in particular a set of data records, each of which contains a reference to a measuring position and also reference values. For a measuring point or measuring location, several data records can be determined that describe the dependence of the reference values ​​on semi-finished product properties, process parameters and / or state variables of the deep-drawing device 2.Finally, a confidence value can be determined for each data set, which describes the accuracy of the calculated reference material flow. The respective reference value has, in particular, the dimension of the respective property value.

[0061] When starting a new production order and / or in the event of significant changes in the semi-finished product properties in a coil and / or within a stack of blanks or semi-finished products 1 to be processed in the upcoming forming processes, it is necessary to identify process parameters to achieve the required drawn part quality. In this context, the process parameters can be determined in addition to experience-based methods and / or with the aid of a model from the field of machine learning, which can be trained, for example, using the aforementioned simulation data and / or previous measurements.

[0062] Provided that a sufficient amount of data or data sets is available, the serial numbers of the semi-finished products 1 that have previously been processed using the identified process parameter combination can first be determined. Using the serial number, the associated recorded semi-finished product properties, the process states, and the measured material flow can be accessed. The determined data can initially be filtered. For the further process, only data sets resulting from semi-finished products 1 that were processed during a stable process sequence or forming process are considered. Furthermore, only semi-finished products 1 whose semi-finished product properties are identical to the properties of the semi-finished product 1 currently to be processed are considered. A certain deviation in the properties can be permitted.The determination of a reasonable tolerance can be made depending on the respective production process, for example based on statistical analyses and / or experience.

[0063] Due to tool wear and / or tool modifications, data records from the recent past should be given greater weighting, if necessary. Based on the data or data records remaining after filtering, the reference material flow can now be derived particularly advantageously, since each semi-finished product 1 is also assigned to the measured material flow via the serial number.

[0064] Using the filtered data, the statistical distribution of the material flow can be determined at each measuring point or location. Based on this, a target value for the material flow and a standard deviation multiplied by a factor can be determined to determine an acceptable scatter band. Alternatively or additionally, models from the field of machine learning can be used to determine the reference values ​​for the reference material flow. Additionally or alternatively, a confidence can also be calculated, which allows a statement to be made about the quality of the determined key figures, such as the median. Furthermore, the determined material flow and the current material flow measured during production can be visualized and compared. A representation of the confidence of the calculated reference material flows can also be shown.

[0065] In addition, a drawn part from the past can be used that shows the smallest deviation from the mean value of several drawn parts with regard to material flow. This offers the advantage of considering a physically occurring case as the reference material flow. If only an average of all measured material flow variables is used, this can describe a reference material flow that cannot physically occur.

[0066] Using such a visualization, the plant operator can, for example, decide to what extent they should adjust the process parameters. Alternatively or additionally, the comparison between the reference material flow and the current material flow can also be carried out mathematically using the values ​​introduced above, and a system-protected warning can be issued to the plant operator. The plant operator can then make the changes to the process parameters based on experience. In addition, an algorithm for recommending process parameters can also be used. In this context, it would be conceivable, for example, to search for data whose associated semi-finished products 1 have properties that are as similar as possible to those currently being processed in order to determine a suggestion for a suitable process parameter combination.Thus, the method is also advantageously suitable for integration into a process control or process regulation, so that the process parameters can generally be determined automatically.

[0067] If no semi-finished products 1 have yet been formed using this process parameter combination, the procedure described in more detail below can be used to successfully carry out the presented process: If, particularly at the beginning of a series of forming processes, there is an insufficient amount of data available to determine a reference material flow, for example because a tool for a deep-drawing operation is being put into operation or changes are being made to the production process, and in particular to the tools for the deep-drawing operation, which have a significant impact on the material flow. Another example would be a change of supplier for the processed semi-finished products. If, for example, the expected values ​​for the sheet thicknesses of the new supplier differ significantly from those of the previous supplier, it may not be possible to identify blanks from the past that are comparable to the blanks currently to be processed. This makes calculating or determining the reference material flow considerably more difficult.Furthermore, it may happen that semi-finished product properties of the blanks or semi-finished products 1 currently awaiting processing have not yet occurred in the past and that none of the process parameter combinations previously used are suitable for their processing.

[0068] Under these circumstances, it may be necessary for the plant operator and / or a machine learning model or an alternative algorithm to determine a suitable process parameter combination that can or at least probably should fulfill the condition of the required drawn part quality of the drawn parts 3. In this way, an estimate of the reference material flow can be made even after fewer components have been produced. The reference material flow should, for example, be updated in short cycles. Such a cycle can, for example, comprise the forming of 20 drawn parts 3 if these are car body parts. When making such an estimate, the user should be signaled by specifying the confidence value that a lower accuracy of the specific reference material flow should be expected, which specifies a rapid data transfer of data records to the plant control system.If no process conditions are taken into account in the process line, the reference material flow can also be calculated immediately after the sheets have been cut in the coil line.

[0069] Based on the presented method, a corresponding computer program can be provided that can carry out the steps of the method, particularly if this is carried out, for example, in the control device of the deep-drawing device 2. This results in the advantage that a flexible algorithm can be presented that is suitable for both manual and automated control of production processes in press shops or deep-drawing devices 2. The method and the computer program can make a significant contribution to reducing the effort required for material flow analysis. In addition, an overview of the material flow of many produced parts, including statistical evaluation, is provided. List of reference symbols 1 semi-finished product 2 deep drawing device 3 Drawn part 4 die 5 sheet holders 6 stamps 7 plungers 8 displacement cylinders S1 first step S2 second step S3 third step S4 fourth step

Claims

[1] Method for carrying out a forming process in which a drawn part (3) is formed from a semi-finished product (1) having at least one semi-finished product property by means of a deep-drawing device (2), comprising the steps: - detecting a property value at at least one measuring point of the drawn part (3) by a sensor device; (S1) - Determining a current material flow during the forming process from the semi-finished product (1) to the drawn part (3) based on the property value of at least one measuring point of the drawn part; (S2) - Determining a deviation of the current material flow from a reference material flow, which depends on at least one semi-finished product property and / or at least one process parameter and / or at least one state variable and for which a drawn part quality is known; (S3) - When a threshold value, which indicates the expected drawn part quality based on the current material flow, is exceeded by the determined deviation: Output of a signal and / or specification of at least one change to process parameters for at least one upcoming forming process. (S4) [2] Method according to claim 1, characterized by that a data set is stored and / or kept for each formed drawn part (3) and / or data sets characterizing the reference material flow are assigned and / or the reference material flow used in determining the deviation is selected by means of a filter, so that only data sets are included whose semi-finished product properties and / or at least one state variable meet certain criteria. [3] Method according to claim 1 or 2, characterized bythat the reference material flow is determined by specifying process parameters for at least one forming process and using these to form at least one semi-finished product and the material flow determined in this way is used as the reference material flow when the drawn part quality is achieved. [4] Method according to one of the preceding claims, characterized by that the reference material flow is determined by means of a simulation, in particular using the finite element method. [5] Method according to one of the preceding claims, characterized by that the reference material flow is determined by selecting several data sets based on semi-finished product properties, process parameters and / or state variables and averaging or statistically considering the respective property values ​​at at least one measuring point and / or determining the reference material flow from the averaged property values. [6] Method according to one of the preceding claims / according to claim 3, characterized by that the specification of the process parameters when the threshold value is exceeded and / or the specification of the process parameters for determining the reference material flow are carried out by a machine learning method. [7] Method according to any one of claims 2 to 6, characterized by that during filtering, the data sets are weighted, in particular based on the time of the respective forming process. [8] Method according to one of the preceding claims, characterized by that the property value of the respective measuring point characterizes a position of a drawing edge, a starting edge and / or a close-running edge. [9] Method according to one of the preceding claims, characterized by that the property value of the respective measuring point is recorded by means of optical sensors and / or by means of a mechanical contact sensor of the sensor device. [10] Method according to one of the preceding claims, characterized by that the determination of the current material flow is carried out by determining a position of the drawn part edge and / or relative to two edges of the drawn part (3), wherein one of the edges is invariant with respect to the material flow. [11] A computer program which can be loaded directly into a memory of an electronic computing device, comprising program means for carrying out the steps of the method according to any one of claims 1 to 9 when the computer program is executed. [12] Electronically readable data carrier with electronically readable control information stored thereon, which comprises at least one computer program according to claim 10 and is designed such that, when the data carrier is used in an electronic computing device, it carries out a method according to one of claims 1 to 9.

Images