Method and device for producing a workpiece from a workpiece blank and computer program with program code

Abstract

A method and a device for producing a workpiece from a workpiece blank (26) use a first workpiece machining machine (12) having a first workpiece holder (22) which holds the workpiece blank (26). A first machining tool (28) can be moved relative to the first workpiece holder (22). A machine controller (14) controls the first machining tool (28) along a first trajectory (30). During machining of the workpiece blank (26), a multiplicity of sensors (36) pick up a multiplicity of sensor signals (38) which form a signal signature characteristic of the machining of the workpiece blank (26). A data set is provided and represents at least one signal signature of earlier machining operations using the first workpiece machining machine (12). The workpiece blank (26) is machined at at least one second machining position on the basis of a first workpiece coordinate which is determined using the current signal signature and the data set.

Description

Method and apparatus for manufacturing a workpiece from a workpiece blank and computer program with program code

[0001] The present invention relates to a method and a device for producing a workpiece from a workpiece blank using an automatically controlled workpiece processing machine and a corresponding computer program.

[0002] Many industries make intensive efforts to increase production output and product quality. Product quality is often defined by whether the manufactured workpieces conform to predefined workpiece specifications, particularly regarding the shape, position, and dimensions of defined workpiece features, surface roughness, structure, and / or gloss level. Workpieces that do not meet the predefined specifications must either be reworked or rejected, which negatively impacts production output and efficiency. To monitor product quality, it is common practice to integrate one or more measuring devices into the production plant and sometimes even directly into a workpiece processing machine, especially a machine tool.However, adjusting a manufacturing process in real time when product quality declines is difficult because acquiring and evaluating measurement results on workpieces during or shortly after production takes time, and the time lag is often too great to efficiently use the measurement results for production control. For many measurement methods, the time until a result is available is often significantly longer than the production cycle times or the period of fluctuations in the production process.

[0003] Particular challenges arise when a complex workpiece is manufactured in a process chain with multiple machining steps that require reclamping the workpiece blank. Reclamping in this case means that the workpiece blank, or the partially manufactured workpiece, must be transferred from one fixture to a second and, if necessary, secured. This is especially true when a workpiece blank is to be machined with tight tolerances on six sides ("all around"). Reclamping typically results in the loss of a clear reference between the workpiece surfaces machined in different setups.

[0004] In some cases, it is possible to integrate quality inspections into the manufacturing process by, for example, measuring a machined surface on a workpiece blank in the machining area and in the fixture used for machining, before the partially machined workpiece blank is further processed. Measuring probes, such as those offered by companies like Renishaw or Blum-Novotest, can be used in the machining area for this purpose. However, the machining process must be interrupted for such in-process measurements, which impacts the machine's production output.

[0005] Nevertheless, quality control in industrial manufacturing processes has been of great interest for years in order to achieve cost-efficient production and high product acceptance among customers. Many concepts and approaches exist for establishing quality assurance processes in the industrial manufacturing of workpieces.

[0006] US 11,249,458 B2, for example, discloses a device with a controller that manages the machining of a workpiece and with sensors that capture an image of the workpiece during machining. The controller generates a three-dimensional model of the workpiece to be machined based on the captured image, compares the generated model with a three-dimensional model created by a machining simulation, and determines the presence or absence of a machining defect based on this comparison. If a If an editing error exists and reprocessing is possible, a setting is changed depending on the cause of the editing error, and additional processing is carried out based on the changed setting.

[0007] US 11 049236 B2 discloses a system and method for real-time quality inspection of objects. An optical scanning unit captures images of the objects to be inspected. The images are compared with CAD models of the objects, and the matched CAD model is extracted. A laser with a wavelength in the violet or ultraviolet range scans the objects and generates three-dimensional point clouds. The point clouds are compared with the extracted CAD models for each object, and the object is classified as acceptable or defective based on the discrepancy between the point cloud and the CAD model.

[0008] US 2021 / 0208568A discloses a manufacturing system comprising a communication module for receiving a three-dimensional model and for receiving control commands containing manufacturing instructions for a manufacturing machine with respective reference and tolerance values; a manufacturing module; a calculation module that uses the three-dimensional model and the manufacturing instructions to calculate control commands for the manufacturing module; and a measuring device that measures a manufactured object using sensors, wherein a deviation of the measured values ​​from applicable manufacturing reference values ​​and an exceedance of the associated manufacturing tolerance values ​​results in a control signal.

[0009] US 9,383,742 B2 discloses a system and method for error compensation in the positioning of a complexly shaped gas turbine engine component during its manufacture. First, theoretical measurements are retrieved for a multitude of control points on the component. Then, actual measurements for the control points are acquired in a coordinate system of the machine. If an error between the actual and theoretical measurements is outside a tolerance, a transformation matrix is ​​calculated. The transformation matrix represents a transformation to be applied to the coordinate system to adjust its position to compensate for the error. The transformation matrix can be calculated and applied iteratively to the coordinate system until the actual measurements are within the tolerance. The tolerances are within acceptable limits. A machining program for manufacturing the part can then be created.

[0010] EP 3 045 992 A1 discloses a method for compensating for errors occurring in a production process using a feedback loop. The method comprises generating actual property data for at least one prototype object manufactured in a production assembly according to a production model, performing a target-actual comparison, thereby generating deviation data, and automatically generating an adapted production model based on the target property data and the deviation data. The adapted production model can be used in an adapted production process to manufacture an adapted object in the production assembly and differs from the target property data, so that the errors occurring in the production process are at least partially compensated in the adapted production process.

[0011] US 6,975,918 B2 discloses a production system for the serial production of products, comprising a machining unit that, depending on control commands, actuates a tool for machining one of the products, a measuring unit for automatically measuring an actual geometric dimension on one of the machined products, and a correction unit coupled to the machining unit and the measuring unit that compares the actual dimension with a predetermined target dimension that lies within a tolerance interval. The correction unit intervenes in the tool's control commands to correct the actual dimension if it lies outside an intervention interval that is within the tolerance interval.

[0012] US 10 180667 B2 discloses a measurement technology integrated into a manufacturing machine, in which the measurement results are interpreted by a trained artificial intelligence (AI). The AI ​​determines new target control data based on the measurement results.

[0013] DE 10 2023 109245 A1 discloses a method for setting up a machine tool for the automatic machining of a workpiece by a machine-controlled tool, comprising the steps of: producing a workpiece with a plurality of target machining points, wherein production includes approaching the target machining points and machining the workpiece at the target machining points with at least one tool; recording a load-dependent machine operating parameter during production; adapting a production speed profile with which a sequence of approach steps and machining steps is carried out during production, depending on the determined machine operating parameter for producing another workpiece.

[0014] Some prior art approaches aim to make corrections even before a workpiece is actually manufactured. In other words, they attempt to preemptively correct potential production errors by incorporating insights gained from a previously produced workpiece into the production process of a later one. Such proactive error correction appears very promising for optimizing the efficiency and output of a real, albeit non-ideal, manufacturing plant. Unfortunately, industrial manufacturing processes and plants can be highly complex, and it is often difficult to definitively identify all the causes and effects that can lead to undesirable production errors and product defects. Therefore, it is common practice to operate a real manufacturing plant with carefully selected process parameters based on numerous test runs, which are generally chosen conservatively."Conservative" here means that the production plant's capacity is not fully utilized in the hope that the desired product characteristics will still be achieved even if the actual production run is unpredictably disrupted. In other words, it is accepted practice not to push a manufacturing plant to its limits when high product quality is a key objective.

[0015] Predictive error correction is particularly difficult when the number of workpieces to be produced, i.e., the batch size, is small. The causes and effects leading to production errors are hard to predict when only a small number of samples are available. On the other hand, producing workpieces in small batch sizes (down to a single piece) is becoming increasingly desirable to allow for customer-specific variations.

[0016] Against this background, it is an object of the present invention to provide a high-performance method and apparatus for manufacturing workpieces of high product quality. In particular, it is an object of the invention to provide a method and apparatus that enable the efficient production of workpieces with tight tolerances on spaced-apart and, in particular, mutually opposite workpiece faces. A further object is to provide a manufacturing method and apparatus that efficiently utilize knowledge from previous production runs to achieve high product quality.

[0017] According to one aspect of the invention, a method according to claim 1 for producing a workpiece from a workpiece blank is provided, comprising the following steps: - Providing a first workpiece machining machine with a first workpiece fixture configured to hold the workpiece blank, with a first machining tool movable relative to the first workpiece fixture, with at least one machine control configured to control the first machining tool along a first path relative to the first workpiece fixture, and with a plurality of sensors providing a plurality of sensor signals during machining of the workpiece blank with the machining tool, wherein the plurality of sensor signals form a signal signature characteristic of the machining of the workpiece blank, - Obtaining an initial data set representing the desired workpiece properties, - Obtaining a second data set representing signal signatures from previous machining operations using the first workpiece machining machine, - Machining the workpiece blank at at least one first machining position, wherein the machine control controls the first machining tool depending on the first data set relative to the first workpiece holder, - Recording a current signal signature during the machining of the workpiece blank at at least one initial machining position using the multitude of sensors, - Determining at least one first workpiece coordinate representing the at least one first machining position, using the current signal signature and the second data set, and - Machining the workpiece blank at at least one second machining position depending on at least one first workpiece coordinate.

[0018] According to another aspect, a device according to claim 14 is provided for producing a workpiece from a workpiece blank, with - a first workpiece machining machine with a first workpiece fixture designed to hold the workpiece blank, and with a first machining tool that is movable relative to the first workpiece fixture, - with a multitude of sensors that provide a multitude of sensor signals during machining of the workpiece blank with the machining tool, wherein the multitude of sensor signals form a signal signature characteristic of the machining of the workpiece blank, and - at least one machine control, wherein the at least one machine control has a first memory configured to receive a first data record, wherein the first data record The at least one machine control has a second memory configured to receive a second data set, wherein the second data set represents signal signatures from previous machining operations using the first workpiece machining machine, wherein the at least one machine control is configured to control the first machining tool along a first path relative to the first workpiece fixture as a function of the first data set in order to machine the workpiece blank at at least one first machining position, wherein the at least one machine control is further configured to record a current signal signature using the plurality of sensors during the machining of the workpiece blank at the at least one first machining position, wherein the at least one machine control is further configured toto determine at least one first workpiece coordinate, representing the at least one first machining position, using the current signal signature and the second data set, and wherein the at least one machine control is further configured to determine a further path as a function of the at least one first workpiece coordinate and to control further machining of the workpiece blank at at least one second machining position using the further path.

[0019] Furthermore, a computer program according to claim 15 is provided with program code configured to implement the following method steps when the program code is executed on at least one machine control of a device of the aforementioned type: - Obtaining an initial data set representing the desired workpiece properties of a workpiece to be manufactured, - Obtaining a second data set representing signal signatures from previous machining operations using the first workpiece machining machine, - Machining a workpiece blank at at least one first machining position, wherein the at least one machine control controls the first machining tool depending on the first data set relative to the first workpiece fixture, - Recording a current signal signature during the machining of the workpiece blank at at least one initial machining position using the multitude of sensors, - Determining at least one first workpiece coordinate representing the at least one first machining position, using the current signal signature and using the second data set, and - Machining the workpiece blank at at least one second machining position depending on at least one first workpiece coordinate.

[0020] The first data set can be a CAD data set, a CAM data set, and / or a set of control commands for at least one machine control system. It defines and / or represents nominal workpiece data, in particular dimensional properties relating to the shape, size, and location of features that the workpiece to be manufactured should possess. Features can include, for example, the diameter, depth, and location of a hole, radii, distances, and / or lengths of edges, among others. The industrial production of workpieces based on or dependent upon such a data set is common practice today.

[0021] The new method and device also utilize a second data set representing signal signatures from previous machining operations using the first workpiece machining machine. These signal signatures result from a multitude of sensor signals recorded during the actual machining of workpieces, thus representing the real-world behavior of the machining machine in the form of process-accompanying sensor signals. Preferably, the sensor signals are acquired at a sampling rate that is... The processing of a defined workpiece feature yields a large number of sensor signal values, thus providing a temporal profile of sensor signal values ​​for that feature. The time intervals between successive sensor signal samples are therefore small compared to the processing time of the defined workpiece feature. Advantageously, the temporal profiles of the sensor signals in the second data set are correlated and / or contextualized with the workpiece properties of the manufactured features.

[0022] In particular, the sensor signals can include the time course of a motor current that powers a drive motor, which in turn moves the machining tool relative to the workpiece blank. Typically, the motor current increases significantly during machining when the machining tool first makes contact with the workpiece blank after approaching it. Analyzing the signal curve makes it possible to precisely determine the time of initial contact between the machining tool and the workpiece blank. This information, along with other sensor signals representing the position of the machining tool relative to the workpiece blank and / or relative to the tool holder, can be advantageously used to determine the actual machining area on the workpiece blank relative to a reference frame that can be freely chosen.Advantageously, the position can be determined relative to a predefined workpiece coordinate system and / or relative to a machine coordinate system whose origin uses a reference point of the machining tool, for example, the so-called tool center point. At least one initial workpiece coordinate represents the actual machining location on the workpiece blank in the form of a data value, enabling error-correcting, modified control of the machining tool in the subsequent manufacturing process.

[0023] The new method and device use the second data set and a current signal signature, acquired during machining of the workpiece blank at a first machining position, to identify the actual first machining position. This can vary depending on wear of the machining tool, current environmental parameters (such as temperature, humidity, etc.), and the individual material properties of the workpiece. The actual first machining position of a workpiece blank may deviate from the assumed first machining position. For example, the actual first machining position of a workpiece blank may differ from the assumed first machining position due to a void on or immediately below the workpiece surface and / or due to a worn cutting edge.

[0024] The new method and device implement a virtual measurement technique using the second data set and the current signal signature. This means that actual workpiece properties are determined based on the signal signatures during machining, thus being "virtually measured." This virtual measurement based on signal signatures makes it possible to measure a partially machined workpiece in situ and concurrently during the manufacturing process. This is achieved by using signal signatures and workpiece properties known from previous machining operations to determine the actual machining activity on the workpiece based on the current signal signature.By evaluating the signal signatures, the machining tool effectively becomes a virtual measuring tool. The new method and device use the measured values ​​obtained in this way during the ongoing manufacturing process to optimize the further machining of the workpiece blank. Advantageously, a second machining position can be selected based on the actual first machining position obtained in the described manner by moving the machining tool to the second machining position with a modified control command.

[0025] The new process and device enable the production of a workpiece with very precisely arranged features. Advantageously, it eliminates the need for the interruption of the manufacturing process that was required in previous methods to measure a partially machined workpiece within the machine's work area. The new process and device thus enable higher production output when manufacturing workpieces with tightly toleranced features. The aforementioned problem is therefore completely solved.

[0026] In a preferred embodiment of the invention, the multitude of sensor signals represent at least one and preferably at least two of the following signal profiles: force profile, torque profile, time profile of a motor current, structure-borne sound profile, temperature profile, each at the first machining tool and / or at the first workpiece holder.

[0027] In addition to the aforementioned time profiles, the sensor signals in preferred embodiments also include one or more current profiles representing the respective time profiles of the drive currents of the machine tool before and during machining, and preferably also afterward. Furthermore, the sensor signals can represent current operating parameters of the machining tool, such as rotational speed and / or operating hours, as well as encoder signals representing the current positions of the workpiece holder and / or the machining tool within the machining tool's work area. These signal profiles enable a detailed and realistic modeling of the machining process and advantageously contribute to determining at least one workpiece coordinate with high accuracy and reliability.It is conceivable to generate further information through conventional measurement on the partially machined workpiece, for example, using probes to measure selected points on the workpiece. However, in preferred embodiments, at least one workpiece coordinate is determined solely on the basis of virtual measurement technology, i.e., based on the aforementioned sensor signals and known signal signatures from the second data set. This design has the advantage that the aforementioned sensor signals can be recorded at a high data acquisition rate without affecting the ongoing machining process.

[0028] In a further embodiment, determining at least one workpiece coordinate includes a sub-step in which the initial physical contact between the first machining tool and the workpiece blank is determined. Advantageously, the at least one workpiece coordinate can be determined using a computational step that includes interpolation and / or back-extrapolation using the sensor signals.

[0029] In this configuration, the machining tool is used like a tactile measuring probe. Vibrations and other disturbances can be effectively compensated for using algorithms known from tactile coordinate measuring technology. This configuration enables very high measurement resolution and is particularly advantageous when the first and second machining positions must be manufactured with very tight tolerances relative to each other.

[0030] In a further embodiment, determining at least one workpiece coordinate includes a temporally preceding sub-step in which the first machining position is selected from a multitude of defined working positions depending on the first data record.

[0031] In this configuration, the determination of at least one initial workpiece coordinate using virtual measurement is limited to previously selected machining positions and / or workpiece features. In principle, it is conceivable to measure the workpiece over a large area or even its entire surface using in-process virtual measurement technology. In contrast, this configuration focuses on selected measurement positions, and consequently, no initial workpiece coordinates are determined in the manner described for unselected workpiece areas. In other words, this configuration involves determining initial workpiece coordinates for some workpiece features and deliberately omitting them for others. Advantageously, this selection is based on the first data set.Preferably, the determination of at least one first workpiece coordinate is limited to workpiece features that are of particular importance for product quality and / or must meet particularly tight tolerances. In preferred embodiments, the tightly toleranced workpiece features can be determined based on the first data set. This design contributes to high production output and, moreover, enables the flexible manufacturing of a large number of workpieces with individual, product-specific accuracy requirements.

[0032] In a further embodiment, the machining of the workpiece blank at at least one second machining position includes a temporally prior step. Partial step in which the workpiece blank is removed from the first workpiece holder and inserted into a second workpiece holder.

[0033] In this embodiment, workpiece production involves reclamping the partially machined workpiece blank, with the first machining position being machined before reclamping and the second machining position being machined afterward. Advantageously, the workpiece blank is machined at at least one second machining position depending on at least one first workpiece coordinate. Machining at the second machining position is thus performed with reference to the virtually measured first machining position. This embodiment enables very high product quality with very tight tolerances regarding the relationship between the first and second machining positions. This embodiment is particularly advantageous when tightly toleranced features are located on workpiece surfaces that cannot be machined in a single setup, such as opposing and diverging surfaces.This design particularly benefits from the efficiency advantages resulting from virtual process-integrated measurement technology.

[0034] In a further embodiment, the second workpiece holder is arranged in the first workpiece processing machine.

[0035] This design enables very fast 6-sided machining of a workpiece with very high product quality because the environmental conditions during machining are largely or even completely identical at the first and second machining positions. The at least one first workpiece coordinate can therefore be incorporated into the machining of the workpiece at the second machining position with particularly high accuracy. In some preferred embodiments, the first workpiece machining machine has a first and a second rotary spindle arranged opposite each other. In these cases, turning with the second rotary spindle is performed depending on at least one workpiece feature produced with the first rotary spindle, measured virtually, i.e., using signal signatures.In further embodiments, the first workpiece processing machine can have one or more rotary spindles and. Furthermore, they must have at least one milling spindle. In these cases, too, the production of a workpiece feature at the second machining position can advantageously be made dependent on a previously virtually measured workpiece feature at the first machining position, for example by modifying a path for machining the workpiece at the second machining position depending on the at least one first workpiece coordinate.

[0036] In a further embodiment, the second workpiece holder is arranged in a second workpiece processing machine.

[0037] In this embodiment, the partially machined workpiece blank is removed from the work area of ​​the first machine tool and transported to a second machine tool. This design allows for great flexibility and a wide range of machining options because the workpiece blank can be machined on a variety of different machines. Furthermore, the presence of at least one workpiece coordinate makes it easier than before to continue machining on the second machine tool with high accuracy and tight tolerances relative to previously manufactured workpiece features. In preferred embodiments, the partially machined workpiece blank is removed from the work area of ​​the first machine tool and placed into the work area of ​​a second machine tool using an automated handling device, particularly a robot. This also contributes advantageously to high dimensional accuracy.

[0038] In a further embodiment, a third data set is obtained, which represents signal signatures of previous machining operations using the second workpiece machining machine, wherein the at least one second machining position is determined depending on the at least one first workpiece coordinate and depending on the third data set.

[0039] In this configuration, characteristic signal signatures are also available for the second workpiece machining machine, making it easier to determine an optimal machining coordinate system for further processing. The workpiece blank is determined on the second machining center. This design enables very fast and highly precise subsequent machining of the workpiece blank on the second machining center. The number of test runs and sample productions for optimizing the process chain can be reduced.

[0040] In a further embodiment, machining the workpiece blank at the at least one second machining position includes a preceding sub-step in which a modified path is determined relative to the second workpiece fixture, depending on the at least one first workpiece coordinate and the first data set. The workpiece blank is then machined at the at least one second machining position using this modified path. The modified path differs from a path determined without considering the at least one first workpiece coordinate in at least one machining parameter.At least one machining parameter can be, for example, a changed feed path of the machining tool relative to the workpiece blank, a changed machining speed (tool or workpiece speed), a changed path speed of the tool and / or workpiece blank, and / or a changed cutting angle. The latter can involve a changed tool.

[0041] This design enables preemptive compensation for workpiece-dependent shape changes that can be caused by reclamping the workpiece blank, shrinkage due to temperature changes, and / or gravitational bending. Such shape changes can be modeled using the first data set. For example, a workpiece that is thin compared to its length can easily deform when it is reclamped from a horizontal position during the first machining operation, such as turning, to a vertical position for further machining at the second machining position. The present design advantageously helps to minimize resulting errors, with the at least one first workpiece coordinate providing an actual reference point for the modified path.

[0042] In a further embodiment, obtaining the second data set includes a sub-step in which a sample workpiece, held in the first workpiece holder, is machined with the first machining tool.

[0043] In this embodiment, one or more signal signatures for the second data set are determined in a temporally prior step. In some preferred embodiments, one or more signal signatures are determined in close temporal proximity to the subsequent manufacturing process for the actual workpiece required, for example, within one day or less than 12 hours before the workpiece is manufactured. Preferably, a plurality of sensor signals are acquired while the prototype workpiece is being machined with the first machining tool, and subsequently, the features produced on the prototype workpiece are measured with a separate, physical (non-virtual) measuring tool, particularly with regard to dimension, shape, and / or position of prototype features.The recorded sensor signals and the measured values ​​on the prototype workpiece are correlated, so that the second data set makes it possible to predict the properties of a subsequently manufactured workpiece based on a comparison of the respective recorded signal signatures. In some embodiments, specific features can be produced on the prototype workpiece that require particularly tight tolerances in workpieces of subsequent series production. It is advantageous if the prototype workpiece and one or more subsequently manufactured production workpieces are of the same type.In some embodiments, only selected manufacturing steps are performed on the prototype workpiece. These steps are characteristic of subsequent series production and / or of particular importance with regard to the tolerances to be maintained, resulting in the prototype workpiece and the subsequently manufactured production workpiece being different types. This design has the advantage of preserving current and characteristic signal signatures for subsequent series production, which contributes to high dimensional accuracy and product quality in production.

[0044] In a further embodiment, obtaining the second data set includes a sub-step in which a sample workpiece, held in the first workpiece holder, is machined at several widely separated machining positions and / or with The workpiece is machined using different machining tools, with the tool engagement points of the machining tools being determined relative to each other based on the sample workpiece. Advantageously, the features produced on the sample workpiece are measured with a separate physical measuring tool, particularly with regard to dimension, form and / or position, and the measurement results obtained are represented in the second data set.

[0045] This configuration offers an efficient way to determine the geometric position of different tool engagement points on a workpiece relative to each other. The individual machine drives and guides, and the resulting workpiece properties, are represented in the second data set. In preferred embodiments of this configuration, the actual shape of the first machining tool can be determined from the obtained data, enabling an even more precise determination of at least one first workpiece coordinate. In some embodiments, a tool wear indicator is determined using the current signal signature. This can be advantageously used to replace the first machining tool in a timely manner, before scrap products result due to the worn tool.

[0046] In a further embodiment, machining the workpiece blank involves removing material from the workpiece blank. Preferably, the first machining tool is a cutting tool with a geometrically defined cutting edge, in particular a turning tool, a drill, a tap, or a milling head.

[0047] The new method and device can be implemented particularly efficiently and cost-effectively using a machining tool that maintains physical contact with the workpiece. Furthermore, machining of workpiece blanks is widespread and well-established. The physical contact between the machining tool and the workpiece blank closely resembles tactile coordinate measurement in terms of virtual measurement. Therefore, established algorithms from tactile manufacturing metrology can be applied for signal evaluation.

[0048] It is understood that the aforementioned features and those to be explained below can be used not only in the combination specified, but also in other combinations or alone, without leaving the scope of the present invention.

[0049] Exemplary embodiments of the invention are shown in the drawing and are explained in more detail in the following description, wherein Fig. 1 shows a schematic representation of an embodiment of the device, Fig. 2 shows a schematic representation of a further embodiment of the device, wherein the details shown here may also be included in the embodiment according to Fig. 1, and Fig. 3 shows a flowchart illustrating an embodiment of a method for manufacturing a workpiece.

[0050] In Fig. 1, an embodiment of the new device is designated in its entirety by reference numeral 10. Individual details of such a device are shown in Fig. 2, to which further reference is made below.

[0051] In this embodiment, the device 10 comprises three workpiece processing machines 12.1, 12.2, 12.3, each controlled by an associated machine control 14.1, 14.2, 14.3. Fig. 2 shows a workpiece processing machine 12 with a machine control 14. The workpiece processing machine 12 from Fig. 2 can, in principle, be any one of the workpiece processing machines 12.1, 12.2, 12.3 from Fig. 1. Furthermore, the workpiece processing machine 12 from Fig. 2 shows a separate embodiment.

[0052] The workpiece machining machines are shown here as multi-axis machine tools capable of performing machining processes such as cutting, milling, drilling, turning, grinding, etc. on a workpiece. Suitable machine tools are commercially available from a variety of suppliers such as Nakamura, DMG Mori Seiki, Chiron, Heller, and many others. In principle, the setup can include 10 workpiece processing machines capable of welding, bending, pressing, additive manufacturing, and / or chemical processing a workpiece. Without any general restrictions, any type and brand of automated, controlled workpiece processing machine capable of manufacturing or machining a workpiece from raw material based on a CAD dataset or comparable dataset can be used.

[0053] Workpiece machining machines have at least one movable machine element, which is controlled by the machine control 14. In Fig. 2, the workpiece machining machine 12 has a first turning spindle 16, a second turning spindle 18 opposite the first turning spindle 16, and a drilling and milling spindle 20. The first turning spindle 16 has a first mounting 22, typically in the form of a chuck. The second turning spindle 18 has a second mounting 24. A workpiece blank in the form of a bar 26 is clamped in the first mounting 22. The turning spindle 16 rotates the bar 26 about its longitudinal axis in a known manner, thus enabling turning operations with a turning tool 28, which can be moved along a path 30 relative to the workpiece mounting 22 and the workpiece blank held there. In Fig.Figure 2 shows the turning tool 28 at a second position 28', which leads to a different workpiece engagement point than the position at which the turning tool 28 is shown. The workpiece blank 26 can also be machined with several different turning tools, for example, with different cutting angles. In some embodiments, the workpiece machining machine 12 has a tool turret equipped with a multitude of tools (not shown here).

[0054] The second turning spindle 18 enables turning of the workpiece blank 26 with another turning tool 32 after the workpiece blank 26 has been reclamped from the first fixture 22 to the second fixture 24. The milling spindle 20 carries a drilling or milling tool 34 in a known manner and enables further machining of the workpiece blank 26.

[0055] In the embodiment shown in Fig. 2, sensors 36 are shown in the area of ​​each spindle. Furthermore, additional sensors can be arranged in the working area of ​​the device 12 and / or in the area of ​​the tools 28, 32, 34. The sensors provide a multitude of sensor signals 38, which in the illustrated embodiment are supplied to the machine control 14. In other embodiments, the sensor signals 38 can alternatively or additionally be supplied to other evaluation and processing units, as explained below with reference to Fig. 1.

[0056] The workpiece processing machines 12.1, 12.2, 12.3 with their associated machine controls 14.1, 14.2, 14.3 can advantageously be arranged at a common factory location, for example, in a common factory hall. In some embodiments, however, the workpiece processing machines 12.1, 12.2, 12.3 and their respective associated machine controls 14.1, 14.2, 14.3 can be located remotely in different buildings, regions, cities, or even countries, but nevertheless together form the device 10.

[0057] Each workpiece machining center 12 and its associated machine control 14 exchange control data. This control data includes, in particular, control commands and numerical control parameters, as known to those skilled in the art. A control command, for example, can correspond to the well-known G-code or M-code and activate a spindle motor, while a numerical control parameter quantifies the spindle speed or the travel distance of a tool head.

[0058] In the embodiment shown here, a partially machined workpiece is removed after machining in the workpiece machining machine 12.1, preferably by an automated handling system (not shown here), and transported to the workpiece machining machine 12.2 for further machining, as indicated here by a block arrow 13.1. Likewise, the partially machined workpiece can be transported to the workpiece machining machine 12.3 for final machining after machining in the workpiece machining machine 12.2.

[0059] As explained below with reference to Fig. 3, the device 10 uses a respective data set 40.1, 40.2, 40.3 for at least one of the workpiece machining machines 12.1, 12.2, 12.3, which represents at least one signal signature 42 from previous machining operations using the respective workpiece machining machine. Without loss of generality, the data sets 40.1, 40.2, 40.3 are hereinafter referred to collectively by the reference numeral 40. In some embodiments, the data sets 40 are generated in temporal relation to the production of a new workpiece using sample workpieces. For this purpose, one or more sample workpieces (not shown here) are machined with a respective workpiece machining machine, and simultaneously the respective sensor signals 38 are recorded using the sensors 36, which together form a signal signature 42 characteristic of the sample machining on the respective machine.Subsequently, features of the respective machined sample workpiece are measured. Preferably, a sample workpiece can be transferred to a measuring device 44 using an automated robot (not shown here). In preferred embodiments, the measuring device 44 can be a coordinate measuring machine (CMM) with a contact (tactile) and / or non-contact probe, a computed tomography device, an industrial microscope, a handheld optical scanning device, and / or any other measuring device suitable for measuring the properties of a machined workpiece. Optical measurement technology with cameras and / or laser scanners is particularly relevant in the fields of machining, additive manufacturing (3D printing), and various forming processes. However, the measurement is not limited to dimensional measurement techniques. Other methods of defect detection are also conceivable, e.g.,Deflectometry, eddy current analysis, surface roughness profilometer, acoustic measurements, etc.

[0060] The measuring device 44 is preferably located near the workpiece processing machine 12 and is preferably configured to automatically measure the workpiece or prototype according to a predefined inspection plan. In some embodiments, the measuring device 44 can be configured to generate 3D point cloud data of workpieces in order to determine dimensional and / or geometric features according to the predefined inspection plan. In preferred embodiments, the measuring device 44 is commercially available from Carl Zeiss Industrial Metrology GmbH, Germany. Available software tools are used, such as Calypso (for regular geometries), Caligo (for freeform surfaces), Gear Pro (especially for measuring gears), GOM Inspect, and / or GOM Volume Inspect. The measured workpiece properties are correlated with the sensor signals and stored in the data sets 40. The correlated sensor signals and workpiece properties form characteristic signal signatures of the respective workpiece processing machine 12, which can be used to virtually measure the actual workpiece properties of a workpiece subsequently manufactured on it during production, namely based on a current signal signature recorded simultaneously with the production process.

[0061] In some preferred embodiments, the device 10 includes a cleaning station 45 in which a manufactured or partially machined workpiece is cleaned before being measured with the measuring device 44. In particular, residues of a cooling lubricant (CML), which is often used in machining processes, are completely removed before the workpiece is measured with the measuring device 44. Preferably, the cleaning process is automated. In particular, the manufactured or partially machined workpiece can be automatically removed from the respective workpiece processing machine 12 by a robot and fed to the cleaning station before the workpiece is transported to the measuring device 44 by another robot or the same robot. In some embodiments, the robot can not only feed the workpiece to the measuring station ornot take over from the measuring device, but be integrated into the measuring process as controlled kinematics, i.e., adjust the relative orientation and / or position of the workpiece to the measuring device depending on the test plan.

[0062] The device 10 includes, for each of the workpiece processing machines 12.1, 12.2, 12.3, a correction control 46.1, 46.2, 46.3, each configured to perform the correction steps described below for the respective associated machine control. In preferred embodiments, the respective correction control 46 is implemented in the form of one or more software components, which are executed on one or more hardware processors in a manner known per se. The one or more hardware processors can be commercially available microprocessors from Intel, AMD, Apple, IBM, Fairchild, ARM, or others. In some embodiments, the software components implementing the correction control 46 can be executed on commercially available computer hardware running one or more commercially available operating systems, such as Windows, Unix, Linux, or MacOS. Advantageously, the software components can be executed on one or more virtual machines, such as virtual machines based on Hyper-V, PowerShell, and / or Cybernetes clusters. Furthermore, the software components implementing the correction control 46 can be installed on hardware already present in a workpiece machining device, such as the hardware implementing the machine controls 14, 14.1, 14.2, or 14.3.For example, there are programmable logic controllers (PLCs) that function as machine controllers and are implemented on hardware similar to that of a conventional personal computer running an operating system such as Windows or Unix / Linux. The functionality of the correction control 46 can be implemented on such a hardware platform. In further embodiments, the software components implementing the functionality of the correction control 46 can be executed on cloud computers and / or edge computers of a computer network.

[0063] The correction controls 46 are specifically designed to determine modified control commands and / or control parameters for the respective assigned machine controls using the signal signatures 42 and to transmit them to the assigned machine controls. For example, a cutting tool can be moved slightly further into the material during the machining of a workpiece blank than would have been the case when using the nominal control parameters, in order to compensate for the increasing wear of the cutting tool over several production runs.

[0064] In the embodiment shown here, the device 10 optionally has a raw data processor and sensor controller 48, which functions as a measuring device adapter and generates formatted point cloud data in a predefined format from raw measurement values ​​obtained from the measuring device 44. The formatted point cloud data represent a measured workpiece by means of a multitude of 3D points relative to a predefined coordinate system in a standardized form, so that various Measuring devices 44 can be used for communication with the correction controllers 46. The correction controllers 46 and the raw data processor & sensor controller 48 can exchange data with each other and with the respective machine control and the measuring device 44. Preferably, the correction controllers 46 each include a machine adapter configured to translate error correction commands into the modified control commands and parameters for a specific type and brand of machine control 14.

[0065] In the embodiment shown here, the device 10 also has an optional higher-level control instance 50, which can be implemented on a commercially available processor from Intel, AMD, Apple, IBM, Analog Devices, Fairchild, ARM, or others, and in particular on virtual machines based on Hyper-V, PowerShell, and / or Cybernetes clusters, on cloud computing devices, and / or edge computing devices. Regardless of the implementation method, the control instance 50 is configured to dynamically assign production programs, including control commands, control parameters, corrections, etc., to the various workpiece processing machines. The control instance 50 can further be configured to determine higher-level error correction commands for each of the connected machine controllers 14.1, 14.2, 14.3 by evaluating signal signatures from multiple workpiece processing machines.In some embodiments, the control instance 50 can use machine learning, in particular deep learning techniques and / or artificial intelligence, to analyze the cause-and-effect relationships in the workpiece processing machines 12.1, 12.2, 12.3 on the basis of the data from each correction control 46 and to determine preemptive corrections for the workpiece processing machines 12.1, 12.2, 12.3.

[0066] Fig. 3 shows an embodiment of the method for manufacturing a workpiece. The same reference numerals denote the same elements as before.

[0067] According to step 60, a sample workpiece is first machined on a workpiece machining center 12 at a first machining position. Without loss of generality, this could, for example, be a turning operation with the machining tool 28 according to Fig. 2. According to step 62, sensor signals are simultaneously The sample workpiece is recorded with sensors 36. According to step 64, the sample workpiece is machined at at least one second machining position, and according to step 66, sensor signals are again simultaneously recorded with sensors 36. According to step 68, the sample workpiece is measured at least at the first and second machining positions. Furthermore, according to step 70, the respective tool engagement points at the first and second machining positions are determined. The information obtained is evaluated together with the sensor signals and forms a signal signature 42, which is representative of the machining at the respective machining position.

[0068] According to step 72, a data set is obtained that represents the desired workpiece properties of a newly manufactured workpiece. This can be a CAD data set. Based on the CAD data, a first machining position on a workpiece blank is selected according to step 74. This position should have a very tight tolerance relative to a subsequent machining position on the workpiece blank. Subsequently, according to step 76, the workpiece blank is machined at the selected first machining position using a tool. Simultaneously, according to step 78, the corresponding sensor signals are recorded, which—preferably—were also recorded in steps 62 and 66.

[0069] Based on the sensor signals currently recorded in step 78 and signal signature 42, a first workpiece coordinate is determined in step 80, representing the first actually machined position on the workpiece blank. This means that the first workpiece coordinate is a virtually measured workpiece coordinate based on the aforementioned sensor signals and signal signatures, representing the actual machining of the workpiece blank at the selected machining position. The first workpiece coordinate therefore varies depending on the actual variations of the machining process, in particular the current wear level of the machining tool, variations due to current environmental parameters such as temperature, humidity, etc., and variations due to the individual workpiece blank used with its individual material properties.

[0070] According to step 82, the partially machined workpiece blank is removed from the first fixture 22 and transferred to a second fixture 24. According to step 84, Depending on the previously determined first workpiece coordinate and further depending on the CAD data of the workpiece to be manufactured, a modified path is determined for machining the workpiece blank at a second selected machining position. Optionally, in some embodiments, the effective machining coordinate system or a modified second machining position can be determined depending on a further data set 86, which represents signal signatures from previous machining operations with the second workpiece fixture. According to step 88, the workpiece blank is machined at the second machining position.

[0071] It is understood that the described process can be extended to a large number of first and second machining positions on a workpiece blank, so that the workpiece ultimately produced largely meets the desired specifications.

[0072] The following advantageous steps for manufacturing a workpiece on a machining center 12, as shown in Fig. 2, are summarized again in chronological order:

[0073] Step 1: Machining trials are conducted on the first turning spindle 16, during which the parameters relevant to the machining result are varied. Parameter space exploration can preferably be experimental, but can also be performed virtually / model-based or in a hybrid manner, at least as a supplement. During these trials, data from the machine control and any appropriately selected additional sensors (force sensors, torque sensors, structure-borne sound sensors, etc. in the spindle and / or tool holder) are recorded at a suitably selected readout rate. The machining results, such as the form, position, and surface finish of the features on the turned part, are determined using appropriately selected physical measurement technology. The results from the physical measurement enable the calibration of the physical "process twin" for the turning process, i.e.,A quantitative process outcome forecast, translated from input, environmental, control and additional sensor signals into tolerance deviations, is possible from this point on.

[0074] Step 2: By comparing data traces from machining operations with different tools or different tool engagement points and physical measurements of the resulting features, two pieces of information are preferably generated. First, the geometric position of the tools or tool engagement points relative to each other is determined (possibly also the actual tool shape). Second, control or auxiliary sensor signal signatures can be identified, which can be used to probe (measure) with the tool itself or with which the tool engagement can be detected. Preferably, the tool engagement point can be "extrapolated back" when the tool is positioned.

[0075] Step 3: Steps 1 and 2 are repeated for all machining operations required for tightly toleranced features, using the spindles required for this purpose.

[0076] Step 4: The process is implemented using the interpretations for signal signatures developed in steps 1 to 3. In particular, the approaches described above allow the transfer of the coordinate system in which the machining operations of tools 1 to n on the first turning spindle 16 took place to the counter spindle 18. If control or additional sensor signal signatures are also available for the tools required for machining on the counter spindle, from which the actual tool engagement can be reliably inferred, the effective machining coordinate system on the counter spindle can be determined directly. This is then executed unchanged after appropriate translation transformation. In this way, any shifts in the machining coordinate systems due to transfer / reclamping deviations, machine deformations, or tool changes (such as wear, built-up edges, etc.) are compensated for.

[0077] Step 5: Step 4 can be performed analogously for milling operations with spindle 20. The transformations of the machining program may require a translation and a rotation.

[0078] Step 6: In some cases, such as large and / or soft and / or long parts and / or parts that do not easily stretch or shrink, or not For simple / isotropic component deformations, in addition to translational and rotational transformations, a new, adapted machining program may be necessary. This means that the actual toolpath in the transformed machining coordinate system must also be changed. In this case, the workpiece pre-formations, as a function of the workpiece surface or machining program interval, can again be calculated either from the control or auxiliary sensor signals or detected using physical measurement technology. Another reason for this necessity could be a tool shape deviation, as identified in step 2.

[0079] Step 7: During ongoing series production, continuous refinements and updates are made to the control and auxiliary sensor signal signature interpretation using physical measurement technology. This may also involve automated anomaly detection in both signal processing and physical measurement technology, as well as statistical, model-driven, or evolutionary approaches for parameter space exploration.

[0080] Step 8: The possibility of performing or supporting parts of the measurements, such as transfer errors, clamping errors, component deformations, with the additional aid of, for example, Renishaw or Novotest probes, always remains.

Claims

Patent claims 1. Method for producing a workpiece from a workpiece blank (26), comprising the steps: - Providing a first workpiece machining machine (12; 12.1) with a first workpiece holder (22) configured to hold the workpiece blank (26), with a first machining tool (28) movable relative to the first workpiece holder (22), with a machine control (14) configured to control the first machining tool (28) along a first path (30) relative to the first workpiece holder (22), and with a plurality of sensors (36) which provide a plurality of sensor signals (38) during machining of the workpiece blank (26) with the machining tool (28), wherein the plurality of sensor signals (38) form a signal signature characteristic of the machining of the workpiece blank (26), - Obtaining (72) a first data set representing desired workpiece properties of the workpiece, - Obtaining a second data set representing at least one signal signature (42) from previous machining operations using the first workpiece machining machine (12), - Machining (76) of the workpiece blank (26) at at least one first machining position, wherein the machine control (14) controls the first machining tool (28) depending on the first data set relative to the first workpiece holder (22), Recording (78) a current signal signature during the machining of the workpiece blank (26) at the at least one first machining position using the plurality of sensors (36), - Determining (80) at least one first workpiece coordinate representing the at least one first machining position, using the current signal signature and the second data set, and - Machining (88) of the workpiece blank (26) at at least a second machining position depending on the at least one first workpiece coordinate.

2. Method according to claim 1, wherein the plurality of sensor signals (38) represent at least one of the following signal profiles: force profile, torque profile, time profile of a motor current, structure-borne sound profile, temperature profile, each at the first machining tool (28) and / or at the first workpiece holder (22).

3. Method according to claim 1 or 2, wherein the determination (80) of the at least one workpiece coordinate includes a sub-step in which an initial contact between the first machining tool (28) and the workpiece blank (26) is determined.

4. Method according to one of claims 1 to 3, wherein the determination (80) of the at least one workpiece coordinate includes a temporally preceding sub-step (74) in which the first machining position is selected from a plurality of defined working positions depending on the first data set.

5. Method according to one of claims 1 to 4, wherein the machining (88) of the workpiece blank (26) at the at least one second machining position includes a temporally prior partial step (82) in which the workpiece blank (26) is removed from the first workpiece holder (22) and inserted into a second workpiece holder (24).

6. Method according to claim 5, wherein the second workpiece holder (24) is arranged in the first workpiece processing machine (12; 12.1).

7. Method according to claim 5, wherein the second workpiece holder (24) is arranged in a second workpiece processing machine (12.2).

8. Method according to claim 7, wherein a third data set (86) is obtained representing signal signatures of previous machining operations using the second workpiece machining machine (12.2), and wherein the at least one second machining position is determined depending on the at least one first workpiece coordinate and depending on the third data set.

9. Method according to one of claims 5 to 8, wherein the machining (88) of the workpiece blank (26) at the at least one second machining position included a further temporally prior sub-step (84) in which, depending on the at least one first workpiece coordinate and depending on the first data set, a modified path curve relative to the second workpiece fixture (24) is determined, wherein the workpiece blank (26) is machined at the at least one second machining position using the modified path curve.

10. Method according to any one of claims 1 to 9, wherein obtaining the second data set includes a partial step (60) in which a sample workpiece held in the first workpiece holder (22) is machined with the first machining tool (28).

11. Method according to any one of claims 1 to 10, wherein obtaining the second data set includes a sub-step (60, 64) in which a sample workpiece held in the first workpiece holder (22) is machined at several machining positions (28, 28') that are separated from each other and / or with different machining tools, wherein tool engagement points of the machining tools relative to each other are determined on the basis of the sample workpiece (70).

12. Method according to any one of claims 1 to 11, wherein the machining of the workpiece blank (26) includes material removal from the workpiece blank (26).

13. Method according to any one of claims 1 to 12, wherein a tool wear measure is determined using the current signal signature.

14. Device for producing a workpiece from a workpiece blank (26), comprising: - a first workpiece machining machine (12, 12.1) with a first workpiece holder (22) designed to hold the workpiece blank (26) and with a first machining tool (28) that is movable relative to the first workpiece holder (22), - with a plurality of sensors (36) which provide a plurality of sensor signals (38) during machining of the workpiece blank (26) with the machining tool (28), wherein the plurality of sensor signals (38) form a signal signature characteristic of the machining of the workpiece blank (26), and - at least one machine control (14, 14.1), wherein the at least one machine control (14, 14.1) has a first memory configured to receive a first data record (72), wherein the first data record represents desired workpiece properties of the workpiece, wherein the at least one machine control (14, 14.1) has a second memory configured to receive a second data record, wherein the second data record represents at least one signal signature (42) of previous machining operations using the first workpiece machining machine (12, 12.1), wherein the at least one machine control (14, 14.1) is configured to control the first machining tool (28) as a function of the first data set along a path curve (30) relative to the first workpiece holder (22) in order to machine the workpiece blank (26) at at least one first machining position (76), and wherein the at least one machine control (14, 14.1) is further configured to record a current signal signature using the plurality of sensors (36) during the machining of the workpiece blank (26) at the at least one first machining position (78), wherein the at least one machine control (14, 14.1) is further configured to determine at least one first workpiece coordinate representing the at least one first machining position using the current signal signature and using the second data set (80), and wherein the at least one machine control (14, 14.1) is further designed to determine a further path curve depending on the at least one first workpiece coordinate (84) and to control further machining of the workpiece blank (26) at at least a second machining position using the further path curve (88).

15. Computer program with program code configured to implement the following process steps when the program code is executed on at least one machine control (14, 14.1) of a device according to claim 14: - Obtaining (72) a first data set representing desired workpiece properties of a workpiece to be manufactured, Obtaining a second data set representing at least one signal signature (42) from previous machining operations using the first workpiece machining machine (12, 12.1), Machining the workpiece blank (26) at at least one first machining position, wherein the at least one machine control (14, 14.1) controls the first machining tool (28) depending on the first data set relative to the first workpiece holder (22), Recording (78) a current signal signature during the machining of the workpiece blank (26) at the at least one first machining position using the plurality of sensors (36), Determining (80) at least one first workpiece coordinate representing the at least one first machining position using the current signal signature and the second data set, and Machining (88) of the workpiece blank (26) at at least a second machining position depending on the at least one first workpiece coordinate.

Images