Method for producing a light metal component, and device therefor

Abstract

The invention relates to a method for producing a light metal component, in particular from an aluminium-based alloy, wherein, in order to form the light metal component, a light metal material is subjected to thermal, in particular thermomechanical, treatment, the method comprising the following steps: a) determining electrical conductivity data for the light metal material during the thermal treatment of the light metal material; b) predicting material properties, in particular a microstructure, of the light metal component using a computer-implemented machine learning procedure (5), wherein input data for the machine learning procedure (5) can be generated using the conductivity data; c) comparing the predicted material properties with reference material properties and preferably adapting production parameters of the production of the light metal component on the basis of the predicted material properties. The invention also relates to a device for data processing, to an eddy current sensor and to a device for producing the light metal component.

Description

[0001] Method for producing a light metal component and device therefor

[0002] The invention relates to a method for producing a light metal component, in particular from an aluminum-based alloy, wherein a light metal material is subjected to a thermal, in particular thermomechanical, treatment for the formation of the light metal component.

[0003] The invention further relates to a device for data processing.

[0004] Furthermore, the invention relates to an eddy current sensor for determining electrical conductivity data of a metal material.

[0005] Furthermore, the invention relates to a use of an eddy current sensor.

[0006] Furthermore, the invention relates to a device for producing a light metal component, comprising a treatment device for the thermal, in particular thermomechanical, treatment of a light metal material.

[0007] In the context of the production of light metal components, wherein a light metal material is subjected to a thermal, in particular thermomechanical, treatment, it would often be desirable in practice to make predictions about material properties, in particular a microstructure, of the light metal component based on a measurement of temperature-dependent physical characteristics of the light metal material carried out during the thermal treatment of the light metal material and, in particular, to adapt manufacturing parameters of the light metal component on the basis of the predictions.However, this has proven difficult to implement in practice, as appropriate measurement methods are generally lacking, particularly for use at the higher temperatures of conventional thermal treatment of light metal materials. Due to the complexity of microstructural metal structure processes, the physical characteristics of the light metal material determined during thermal treatment generally do not allow sufficient interpretation for reliable prediction of material properties. Attempts to conduct X-ray-based examinations of the light metal material during thermal treatment have been reported, but these are associated with safety concerns, as have attempts to determine the elastic mechanical properties of the light metal material using laser ultrasound.

[0008] This is where the invention comes in. The object of the invention is to provide a method of the type mentioned above with which the production of a light metal component can be improved. In particular, it should enable a prediction of material properties based on a measurement of physical light metal material characteristics carried out during thermal treatment. In particular, it should enable a prediction of production deviations based on a measurement of physical light metal material characteristics carried out during thermal treatment in order to detect process anomalies.

[0009] A further aim is to provide a data processing device of the type mentioned above, with which the production of a light metal component can be improved. In particular, it should enable a prediction of material properties based on a measurement of physical light metal material characteristics carried out during thermal treatment.

[0010] Furthermore, it is an object to provide an eddy current sensor of the type mentioned above, with which the production of a light metal component can be improved. In particular, it is intended to enable a prediction of material properties based on a measurement of physical light metal material characteristics carried out with the eddy current sensor during thermal treatment. Furthermore, it is an object to provide a use of such an eddy current sensor of the type mentioned above, with which this can be achieved.

[0011] Furthermore, it is an object to provide a device of the type mentioned above for producing a light metal component, with which device the production of a light metal component can be improved. In particular, it should enable a prediction of material properties based on a measurement of physical light metal material characteristics carried out during the thermal treatment.

[0012] The object is achieved according to the invention by a method of the type mentioned above, if the method comprises the following steps: a) determining electrical conductivity data of the light metal material during the thermal treatment of the light metal material; b) predicting material properties, in particular a microstructure, of the light metal component using a computer-implemented machine learning procedure, wherein input data of the machine learning procedure are formed using the conductivity data; c) comparing the predicted material properties with comparative material properties and preferably adapting manufacturing parameters for the production of the light metal component based on the predicted material properties.

[0013] The invention is based on the finding that electrical conductivity data of the light metal material determined during thermal treatment are suitable for predicting material properties, in particular a microstructure, of the light metal component based on the electrical conductivity data. The electrical conductivity data can be represented by an electrical conductivity curve, often as a function of the temperature of the light metal material and / or as a function of time. As a rule, the conductivity data are subject to systematic influences, such as temperature influences and / or the influence of microstructural processes, so that a direct interpretation of the conductivity data for deriving material properties proves to be unreliable and / or difficult.This is especially true if the light metal material is made of an alloy, especially an aluminum-based alloy, formed with at least three alloying elements. This is particularly problematic if phase transformations and / or precipitation processes and / or dissolution processes of phases of the alloy, especially in the solid state of the alloy, occur during the thermal treatment or during the determination of the electrical conductivity data. It has been shown that a prediction of material properties, in particular a microstructure, of the light metal component can be implemented using a computer-implemented machine learning procedure, with the input data of the machine learning procedure being formed using the conductivity data.In addition to the conductivity data, the input data typically includes other data, such as the temperature of the light metal material and / or process parameters, which describe the thermal treatment and / or a mechanical deformation of the light metal material, as described in particular below in this document. The temperature of the light metal material is typically implemented using time-varying temperature data of the light metal material, which is determined in parallel with the determination of the conductivity data. A combination of determining or ascertaining electrical conductivity data during the thermal treatment of the light metal material and processing the conductivity data with the machine learning procedure, by forming the input data with the conductivity data, enables a prediction of material properties, in particular a microstructure, of the light metal component.Advantageously, a control and / or adjustment, in particular modification and / or adjustment, of manufacturing parameters for the production of the light metal component, in particular the method for manufacturing and / or forming the light metal component, can then be carried out based on the predicted material properties. In particular, this allows the light metal component to be manufactured with improved properties and / or the light metal component to be manufactured with greater efficiency. The determination of the conductivity data and / or prediction of the material properties can be carried out during the production of the light metal component, in particular in situ.The electrical conductivity data can comprise, in particular, complex electrical conductivities of the light metal material and / or the real parts and imaginary parts of the electrical conductivities, in particular, usually at different times of the thermal treatment and / or at different temperatures of the light metal material.

[0014] It is expedient if, particularly in step a), a temperature correction of the electrical conductivity data is carried out, wherein the conductivity data are temperature-corrected using a temperature coefficient of the electrical conductivity data, in particular the electrical conductivity. The conductivity data are often temperature-corrected in such a way that they are represented with reference to a, in particular common, reference temperature, for example 20°C. The temperature coefficient of the electrical conductivity usually corresponds to the temperature coefficient of the electrical resistance, often specified as a. The temperature coefficient usually indicates the relative change in the electrical conductivity data, in particular the electrical conductivity, per change in temperature, usually to a, in particular aforementioned, reference temperature.One difficulty that has been shown to be difficult is that the temperature coefficient can change during thermal treatment. This can impair the prediction of the material properties. This is particularly true if the light metal material is formed from an alloy, in particular with at least two or at least three alloying elements, and / or if phase transformations and / or precipitation processes and / or dissolution processes of phases of the alloy of the light metal material occur during the thermal treatment or during the determination of the electrical conductivity data, in particular in the solid state of the alloy. In particular, the temperature coefficient can change depending on the proportion of foreign atoms that are in solution as part of the alloy, whereby the proportion of foreign atoms usually varies during thermal treatment, for example due to precipitation and / or dissolution of phases.The temperature correction may take into account temperature measurement uncertainties. It has been shown that this problem can be mitigated, in particular prevented, if the conductivity data includes real and imaginary data of a complex electrical conductivity. Preferably, the real and imaginary data form input data components of the input data. It has been shown that a temperature correction of the electrical conductivity data, in particular as mentioned above, can then be dispensed with, especially before performing step b).

[0015] Particularly due to the aforementioned difficulty, it is advantageous for a high accuracy of the prediction of the material data if the conductivity data have, in particular are, real part data and imaginary part data of a complex electrical conductivity, in particular of complex electrical conductivities, of the light metal material. The real part data generally represent a real part of the complex electrical conductivity. The imaginary part data generally represent an imaginary part of the complex electrical conductivity. The conductivity data, in particular real part data and imaginary part data, can be determined, in particular measured, using a conductivity measuring device, in particular inductively or using an eddy current sensor. Typically, the conductivity measuring device, in particular the eddy current sensor, is designed to determine, in particular measure, the conductivity data inductively.The conductivity data can be determined contactlessly or by contacting the light metal material with the conductivity measuring device. "Contactless" refers to a measurement without direct contact between the conductivity measuring device and the light metal material. The conductivity data is generally determined non-destructively with respect to the light metal material. The input data for the machine learning procedure can be formed using the real and imaginary data. Calculating electrical conductivity, particularly complex conductivity, using the real and imaginary data typically requires a temperature correction, particularly prior to the calculation, which typically takes temperature deviations into account. It has been shown that, for high accuracy, it is more advantageous if the real and imaginary data form separate input data sections.The input data are then typically generated without calculating a particularly complex electrical conductivity using the real and imaginary data. The real and imaginary data can be raw data determined directly using the eddy current sensor. In particular, a temperature correction of the electrical conductivity data, particularly the real and / or imaginary data, which temperature correction particularly takes into account a temperature coefficient of the electrical conductivity data and / or measurement uncertainties, can generally be omitted before performing step b). The conductivity data, particularly the real and imaginary data, typically each represent a data curve, particularly a measurement curve, as a function of time, particularly the time of the thermal treatment, or the temperature of the light metal material.The electrical conductivity data typically relate to the electrical conductivity within the interior of the light metal material, in particular within a processed body formed by the light metal material and / or within a bulk of the light metal material. The above is particularly advantageous if the light metal material is formed from an alloy, in particular with at least two or at least three alloying elements, and / or if phase transformations and / or precipitation processes and / or dissolution processes of phases of the alloy of the light metal material, in particular in the solid state of the alloy, occur during the thermal treatment or during the determination of the electrical conductivity data.

[0016] A person skilled in the art will understand that the, in particular complex, electrical conductivity of the light metal material can be expressed in an analogous manner by a, in particular complex, electrical resistance, in particular an electrical impedance, of the light metal material or can be implemented as a, in particular complex, electrical resistance, in particular electrical impedance. In an analogous manner, the conductivity data can have, in particular be, real part data and imaginary part data of the complex electrical resistance, in particular the electrical impedance, of the light metal material. A corresponding implementation can be encompassed by an implementation with electrical conductivity described in this document. A corresponding implementation with electrical resistance data is therefore encompassed by the implementation with electrical conductivity data described in this document and / or can be regarded as its specific embodiment variant.A conductivity or conductivity data described in this document usually refers to an electrical conductivity or electrical conductivity data.

[0017] The method can comprise a treatment procedure in which the thermal treatment is carried out. The treatment procedure can comprise the thermal treatment and a mechanical deformation of the light metal material. The mechanical deformation can be carried out before or after, or at least partially in parallel with, the thermal treatment. It is often provided that, in the treatment procedure, the light metal material is mechanically deformed before and / or at least partially in parallel with the thermal treatment. Mechanical deformation of the light metal material carried out in parallel with the thermal treatment of the light metal material is usually referred to as thermomechanical deformation. In particular, the mechanical deformation can take place essentially in parallel with the thermal treatment. The mechanical deformation can be a forming of the light metal material.The forming process can comprise, in particular, rolling, free-forming, die-forming, extrusion, forging, and / or bending of the light metal material. Preference is given to extrusion and / or rolling. The thermal treatment can comprise, in particular, tempering of the light metal material. For high application practicality, it is advantageous if the thermal treatment comprises, in particular, artificial aging and / or homogenization of the light metal material. The determination of the electrical conductivity data is usually implemented in the form of an in-situ measurement. The treatment procedure, in particular the thermal treatment and usually the mechanical deformation, can be implemented using a treatment device. The treatment device can have a deformation device for carrying out the mechanical deformation of the light metal material.The treatment facility may comprise a tempering device for tempering the light metal material. The treatment procedure, in particular the thermal treatment and / or the deformation of the light metal material, is typically described, in particular characterized, with respective process parameters. The process parameters may be part of the manufacturing parameters.

[0018] For example, the treatment procedure can comprise homogenization and / or preheating of the light metal material, subsequent mechanical deformation of the light metal material, and typically subsequent artificial ageing or solution annealing of the light metal material. The thermal treatment can be formed by, in particular, homogenization, preheating, or optionally artificial ageing or solution annealing of the light metal material. The mechanical deformation can be implemented by extrusion, forging, or rolling of the light metal material. The preheating can be heating of the light metal material for extrusion, forging, or rolling of the light metal material. Preferably, the mechanical deformation is implemented by extrusion. Preferably, the thermal treatment is implemented by homogenization. The light metal material can be implemented as a billet or ingot.The treatment procedure may include cooling, in particular active cooling, of the light metal material. Cooling of the light metal material is typically performed during or after mechanical deformation of the light metal material. Active cooling may be implemented using a cooling device. Active cooling may include cooling the light metal material with a cooling medium. The cooling device may be part of the treatment device defined below in this document.

[0019] The material properties, in particular predicted, can include a microstructure, grain structure, texture, strength, corrosion, fatigue, optical properties, and / or decorative properties of the light metal component. Preferably, the material properties, in particular predicted, include a microstructure and / or grain structure of the light metal component. The light metal component can be a finished product or a semi-finished product. The light metal material and / or the light metal component is preferably formed with, in particular from, an alloy, in particular an aluminum-based alloy or magnesium-based alloy. The light metal material can be a light metal workpiece. The alloy of the light metal material can be formed with several, in particular with at least two or at least three, alloying elements. The alloy can be an Al-Mg alloy, in particular an Al-Mg-Si alloy.The alloy can be an aluminum-based alloy, with magnesium (Mg) or silicon (Si) forming a second-largest element proportion of the aluminum-based alloy. The aluminum-based alloy typically has aluminum (Al) as a largest element proportion of the aluminum-based alloy. In particular, the aluminum-based alloy can contain Mg and Si, with Mg forming a second-largest element proportion and Si forming a third-largest element proportion, or with Si forming a second-largest element proportion and Mg forming a third-largest element proportion of the aluminum-based alloy. A particularly high level of practical application can be achieved if the alloy of the light metal material is an Al-Mg-Si alloy, in particular from group 6xxx. A high level of practical application can be achieved if the alloy is an Al-Zn-Mg alloy, in particular from group 7xxx, or an Al-Cu alloy or Al-Cu-Mg alloy, in particular from group 2xxx.The alloy is typically an aluminum-based alloy, particularly a wrought aluminum alloy. The aforementioned groups commonly refer to a designation system for wrought aluminum alloys according to EN 573-3 / 4. As part of the process, phase transformations and / or precipitation processes and / or dissolution processes of phases of the alloy of the light metal material, particularly in the solid state of the alloy, can occur during the thermal treatment or during the determination of the electrical conductivity data. If the alloy of the light metal material is an Al-Mg-Si alloy, dissolution and / or precipitation of Mg-Si phases, particularly Mg2Si phases and / or metastable precursor phases of Mg-Si phases, and / or precipitation of dispersoids can occur in the alloy. The dispersoids are typically formed with, in particular, Fe, Mn, Cr, or Al2(Fe,Mn,Cr)3Sii-2, which contains Zr, or AhZr.

[0020] It is advantageous if the conductivity data is determined using an eddy current sensor. The conductivity measuring device can then comprise the eddy current sensor or be the eddy current sensor. In particular, the real part data and imaginary part data can be determined, in particular measured, using the eddy current sensor. It is advantageous if the conductivity measuring device, in particular the eddy current sensor, is designed to determine, in particular measure, the conductivity data in a temperature range of the light metal material greater than 100°C, in particular greater than 200°C. The temperature range is preferably between 100°C and 600°C, in particular between 200°C and 600°C, preferably between 230°C and 580°C, particularly preferably between 300°C and 550°C. The temperature range usually refers to a, in particular time-varying, temperature of the light metal material, usually parallel to the determination of the conductivity data.

[0021] It is advantageous if the eddy current sensor for inductively determining conductivity data has a wire coil formed with, in particular, a non-ferromagnetic metal. It is expedient if the wire coil or the metal is dimensionally stable in an aforementioned temperature range, in particular between 100°C and 600°C. In particular, the wire coil is suitable for inductively determining conductivity data with the wire coil in the aforementioned temperature range. It has proven useful if the eddy current sensor for inductively determining conductivity data has a wire coil formed with, in particular, tungsten or a tungsten alloy, platinum or a platinum alloy, gold or a gold alloy.The wire coil is preferably formed from a metal or an alloy, in particular a metal alloy, in which the respective aforementioned element forms a largest element proportion of an element composition of the metal or metal alloy. This enables a determination, in particular measurement, of the electrical conductivity data at typical temperatures, in particular temperatures as described in this document, of the thermal treatment of the light metal material. Typically, the eddy current sensor has an excitation coil in order to induce eddy currents in the light metal material with the excitation coil, usually by applying an electrical alternating voltage to the excitation coil. As a rule, a magnetic field generated by the eddy currents is detected, in particular measured, with the eddy current sensor, in particular with a receiving coil of the eddy current sensor, in order to determine the electrical conductivity data.The excitation coil and / or the receiving coil can each form the wire coil, in particular each being implemented with a wire coil formed with, in particular, tungsten or a tungsten alloy. This allows for a high level of heat resistance of the sensor, while, in particular, avoiding magnetic transitions. The excitation coil and the receiving coil can be separate coils or formed by the same coil.

[0022] It is advantageous if, in step a), temperature data of the light metal material is determined during the thermal treatment of the light metal material, wherein in step b), the input data, in particular in addition to the conductivity data, are formed with the temperature data. The temperature data can be temperatures of the light metal material, in particular at different times, during the thermal treatment of the light metal material. The conductivity data and / or the temperatures can be determined as a function of time, in particular a time of the thermal treatment. The temperature data can be determined, in particular measured, using a temperature measuring device during the thermal treatment of the light metal material. The temperature measuring device can be part of the eddy current sensor.The temperature data can be determined contactlessly, in particular pyrometrically, or by contacting the light metal material, in particular with a temperature sensor that contacts the light metal material to determine the temperature data. The temperature sensor can be a thermocouple or resistance temperature sensor. To determine the temperature data, the temperature sensor can be arranged on a surface of the light metal material or at least partially, preferably entirely, recessed into the light metal material. It is advantageous if the temperature sensor is arranged on or recessed into the light metal material in such a way that temperature data of a center of the light metal material can be determined with the temperature sensor. The center can refer to a cross-section, in particular a longitudinal cross-section, of the light metal material, in particular a light metal material geometry of the light metal material.In step b), the input data usually comprise different types of data, usually the conductivity data, the temperature data and frequently one or more of the process parameters, in particular process parameters of the treatment procedure, in particular the thermal treatment and / or process parameters of the mechanical deformation, or the input data are formed with these.

[0023] To ensure high accuracy in temperature data determination, it is advantageous if the temperature data determined with the temperature measuring device are adjusted with temperature data of a reference material which essentially has the same geometry as the light metal material. This can be done as part of the method, in particular in step a). It is advantageous if the reference material is essentially a pure metal material. The temperature data of the reference material can be determined as part of the method. For example, this can be done by subjecting the reference material to the same thermal treatment, in particular tempering, as the light metal material, with temperature data of the reference material being determined, in particular measured, using a temperature measuring device, usually during the thermal treatment. This can be done in a similar way to that described for the light metal material.The reference material may, for example, be formed from substantially pure aluminum, for example aluminum with a purity of at least 99.5%, in particular at least 99.99%.

[0024] Typically, the input data of the machine learning procedure includes a chemical composition of the light metal material, typically in step b), in particular the input data is formed using the chemical composition. Typically, the conductivity data is determined, in particular measured, in a temperature range of the light metal material greater than 100°C, in particular greater than 200°C. The temperature range is preferably between 100°C and 600°C, in particular between 200°C and 600°C, preferably between 230°C and 580°C, particularly preferably between 300°C and 550°C. The conductivity data is typically determined using the conductivity measuring device.

[0025] It has proven useful if the input data of the machine learning procedure includes process parameters of the treatment procedure, in particular the thermal treatment and / or the mechanical deformation. Accordingly, the input data can include process parameters of the thermal, in particular thermomechanical, treatment and / or process parameters of a mechanical deformation of the light metal material, in particular the aforementioned one, carried out before and / or at least partially in parallel with the thermal treatment. This makes it possible to increase the accuracy of the prediction of the material properties. The process parameters preferably include a temperature program for the tempering of the light metal component and / or deformation parameters for the deformation, in particular the forming, of the light metal component.For example, the process parameters can include a forming tool temperature of a forming tool of the forming device with which the light metal material was mechanically deformed and / or a deformation speed of the mechanical deformation. In the case of forming with extrusion, the process parameters can include an extrusion speed and / or a billet temperature of a billet with which a force is exerted on the light metal material for extrusion. In the case of forming with rollers, the process parameters can include a pass schedule.

[0026] In step c), the manufacturing parameters for the light metal component can be adjusted based on the predicted material properties. This adjustment is typically dependent on the result of a comparison of the predicted material properties with the reference material properties. It is expedient for the adjustment to be carried out conditionally if there is a significant deviation between the predicted material properties and the reference material properties. Typically, a significant deviation refers to a deviation between the predicted material properties and the reference material properties that is greater than a customarily tolerated maximum deviation between material properties considered to be essentially identical.Adjusting the manufacturing parameters may include adjusting one or more of the process parameters, in particular the thermal treatment and / or the mechanical deformation. In this way, a change in the treatment of the light metal material can be carried out efficiently. This is especially true if the treatment procedure represents a continuous treatment of the light metal material and / or if different regions, in particular segments, of the light metal material are subjected to the treatment procedure one after the other.

[0027] The reference material properties can be implemented analogously to the material properties described, in particular predicted, in this document. Typically, in step c), corresponding features of the predicted material properties and the reference material properties are compared. The reference material properties can be provided, in particular as part of the process.

[0028] It is advantageous if, in step c), a deviation error is formed based on the comparison of the predicted material properties with the reference material properties. The deviation error can represent a deviation between corresponding characteristics of the predicted material properties and the reference material properties. It is advantageous if, after the comparison, the manufacturing parameters are adjusted so that the deviation error is reduced, in particular minimized. It is advantageous if, in step c), the manufacturing parameters are adjusted if the deviation error is greater than a predetermined deviation limit. Typically, the manufacturing parameters are not adjusted in step c) if the deviation error is smaller than the deviation limit.The deviation limit can represent a customarily tolerated deviation between material properties considered to be essentially identical, in particular their corresponding characteristics, the predicted material properties, and the reference material properties. In this way, an efficient adjustment of the manufacturing parameters toward the reference material properties and, in particular, an efficient production of a light metal component according to the reference material properties can be achieved. The reference material properties can be target material properties. The deviation limit can be specified, in particular as part of the process. High efficiency can be achieved if the adjustment of the manufacturing parameters takes place with iterative adjustment of process parameters that form the input data, so that the deviation error is reduced.Process parameters that form the input data are typically process parameters used to generate input data for the machine learning procedure in step b). The iterative adaptation can be implemented by including the treatment procedure for the light metal material and steps a) to c) as part of an iteration loop that is run through in multiple iterations. Typically, in step c) of the respective iteration, an adaptation of process parameters that form the input data is performed, and the treatment procedure, in particular the thermal treatment and / or mechanical deformation of the light metal material, is carried out in the respective subsequent iteration according to the adapted process parameters that form the input data.

[0029] It is advantageous if the conductivity data is reduced before performing step b) using a compression method, in particular by performing a Fourier transformation and / or an encoder-decoder architecture. This can be done in a step b0), which is usually performed between step a) and step b). By performing the compression method, the amount of data representing the conductivity data and / or the data complexity can be reduced. Performing the Fourier transformation usually involves a Fourier transformation of the conductivity data. The encoder-decoder architecture usually includes an encoder, with which the conductivity data is transferred into a latent space, in which latent space the conductivity data is represented as a compressed representation.The encoder-decoder architecture can have a decoder to convert the conductivity data from the latent space back into a reconstructed form, which typically represents the original conductivity data. The original conductivity data is typically the conductivity data used by the encoder for conversion into the latent space. The encoder-decoder architecture can be implemented in a form known to those skilled in the art. For example, the encoder-decoder architecture can be implemented as an autoencoder, in particular a multilayer autoencoder. Preferably, the temperature data, in particular the conductivity data as a function of the temperature data, are also reduced in the aforementioned manner, in particular in step b0). In step b), the input data can then be formed using the reduced conductivity data or reduced temperature data.

[0030] The conductivity data can be determined and / or described as a function of a temperature, in particular the temperature of the light metal material, or a time, in particular a time of thermal, in particular thermomechanical, treatment. The conductivity data can represent a conductivity curve. The input data of the machine learning procedure are typically fed to the machine learning procedure, in particular in step b), in order to predict the material properties based on these data, in particular by the machine learning procedure.

[0031] It has proven effective if the machine learning procedure is formed using a decision tree and / or a neural network, in particular a feed-forward neural network or a recurrent neural network. The neural network can be designed and in particular trained to predict the material properties, in particular a microstructure, of the light metal component based on the input data. The machine learning procedure can be designed to predict the material properties for different coordinates of the light metal material. The coordinates can be different coordinates of a cross-section of the light metal material geometry. It is advantageous if the coordinates are distributed across the cross-section of the light metal material geometry, in particular in a grid-like manner, in order to determine and / or display a distribution of material properties in the cross-section.The respective coordinate can represent a spatial region of the light metal material, in particular the light metal material geometry, and in particular can be such a region. It has proven useful if the machine learning procedure, in particular the neural network, is constructed in such a way that regions of the light metal material represented by different coordinates are each processed separately by the machine learning procedure.

[0032] Predict material properties of the respective area. This is especially true if the machine learning procedure is a feed-forward neural network. The machine learning procedure can be configured to suggest an adjustment of manufacturing parameters based on the predicted material properties. The adjustment of manufacturing parameters in step c) can be carried out taking into account, in particular according to, the proposed adjustment of manufacturing parameters. The manufacturing parameters can be the process parameters. Steps b) and c), and in particular b0), are typically implemented by computer.

[0033] It has been shown that electrical conductivity data of the light metal material determined during thermal treatment are suitable for predicting production deviations during the manufacture of the light metal component based on the electrical conductivity data. Specifically, it has been found that a prediction of production deviations during the manufacture of the light metal component can be implemented using a computer-implemented machine learning procedure, with the input data of the machine learning procedure being formed using the conductivity data.

[0034] Accordingly, the object is achieved according to the invention by a further method of the type mentioned at the outset for producing a light metal component, if the method comprises the following steps: a) determining electrical conductivity data of the light metal material (9) during the thermal treatment of the light metal material (9); b) predicting production deviations in the manufacture of the light metal component using a computer-implemented machine learning procedure (5), wherein input data of the machine learning procedure (5) are formed using the conductivity data; c) determining a process anomaly in the manufacture, in particular in series production, of the light metal component based on the predicted production deviations and preferably adapting manufacturing parameters for the manufacture of the light metal component based on the predicted production deviations.

[0035] The further method for producing a light metal component relating to a prediction of production deviations can be implemented in a similar manner, in particular with similar features and effects, to the method for producing a light metal component relating to a prediction of material properties described in this document. This also applies to the method for producing a light metal component relating to a prediction of material properties with regard to the further method relating to a prediction of production deviations.

[0036] The production deviations in the manufacture of the light metal component can be deviations of the light metal component, in particular its design, from a comparison design of the light metal component and / or deviations in a treatment process of the light metal material, which treatment process the light metal material is treated and / or processed with to form the light metal component, in particular in comparison to a comparison treatment process. The treatment process is usually part of the process for manufacturing the light metal component. The comparison design and / or the comparison treatment process, in particular the implementation features characterizing them, can be specified, in particular for the machine learning procedure.

[0037] In step c), a comparison of the predicted production deviations with one or more, in particular predefined, normal ranges can be carried out, in particular to determine the process anomaly. In particular, production deviations outside the normal range can be assessed as a process anomaly. Expediently, several such process anomalies can be identified in step c). It is advantageous if the process anomaly relates to a production of the light metal component, which production is implemented as series production, in particular of light metal components. Series production can be the production of several light metal components in series, whereby in particular the light metal components can be manufactured as described in this document. The production of the light metal component described in this document can take place as part of a series production of light metal components.

[0038] The aim is achieved according to the invention with a data processing device of the type mentioned at the outset, if the device comprises means, in particular at least one processor, which are adapted to carry out step b) of the method, in particular as described in this document. Typically, the means are designed to predict material properties, in particular a microstructure, of the light metal component using a computer-implemented machine learning procedure, wherein input data of the machine learning procedure is formed using the conductivity data. The data processing device, in particular the means, can be used for the method, in particular as part of the method, for producing the light metal component, in particular as described in this document, and can be implemented with corresponding features and effects.The processor can, for example, be formed with, in particular from, one or more graphics processors (also referred to as GPU or graphics processing unit).

[0039] The data processing device, in particular the means, can be adapted to carry out step b0) of the method for producing a light metal component, in particular as described in this document. In particular, the means can be adapted to reduce the conductivity data before carrying out step b) by performing a Fourier transformation and / or an encoder-decoder architecture, in particular as described in this document. The data processing device, in particular the means, can be adapted to carry out step c) of the method for producing a light metal component, in particular as described in this document.

[0040] The further object is achieved according to the invention with an eddy current sensor of the type mentioned above for determining electrical conductivity data of a metal material, in particular a light metal material, if the eddy current sensor for inductively determining the electrical conductivity data has a wire coil formed with, in particular from, a non-ferromagnetic metal. It is advantageous if the wire coil is formed with, in particular from, tungsten or a tungsten alloy or platinum or a platinum alloy or gold or a gold alloy. Preferably, the wire coil is formed with, in particular from, a metal or an alloy, in particular a metal alloy, in which the respective aforementioned element forms a largest element proportion of an element composition of the metal or metal alloy. It is expedient if the wire coil or the metal is dimensionally stable in an aforementioned temperature range.The eddy current sensor can be designed as described in this document, in particular to determine, in particular to measure, electrical conductivity data as described in this document. The electrical conductivity data can include, in particular, real part data and imaginary part data of a complex electrical conductivity. The eddy current sensor can be designed to determine, in particular to measure, the conductivity data in a temperature range of the metal material greater than 100°C, in particular greater than 200°C. The temperature range can be between 100°C and 600°C, in particular between 200°C and 600°C, preferably between 230°C and 580°C, particularly preferably between 300°C and 550°C. This can be achieved in particular by the aforementioned implementation of the wire coil.

[0041] The eddy current sensor can be used for the method, in particular as part thereof, for producing a light metal component, in particular as described in this document. The eddy current sensor can be designed according to the features and effects which are described in the context of the method for producing a light metal component, in particular above, in this document. The same applies analogously to the method for producing a light metal component with regard to the eddy current sensor. The person skilled in the art will understand that the, in particular complex, electrical conductivity of the Meta II material, in particular light metal material, can be expressed in an analogous manner by a, in particular complex, electrical resistance, in particular an electrical impedance, of the Meta II material, in particular light metal material, and that a corresponding implementation of a implementation described in this document with electrical conductivity orelectrical conductivity data is included.

[0042] Typically, the eddy current sensor comprises an excitation coil for inducing eddy currents in the metal material, typically by applying an alternating electrical voltage to the excitation coil. Typically, a magnetic field generated by the eddy currents is detected, in particular measured, by the eddy current sensor, in particular by a receiving coil of the eddy current sensor, to determine the electrical conductivity data. The excitation coil and / or the receiving coil can each form an aforementioned wire coil, in particular each being implemented with a wire coil formed with, in particular, tungsten or a tungsten alloy. Typically, the excitation coil and the receiving coil are arranged electromagnetically coupled to one another.The eddy current sensor can be configured to determine the conductivity data, in particular an electrical conductivity, based on a relative change in the real part data and imaginary part data of a particularly complex electrical conductivity, in particular analogous to an electrical impedance, of the metal material, in particular a light metal material, measured with the eddy current sensor. For this purpose, the eddy current sensor can have a control device to which the excitation coil and the receiving coil are connected for controlling the determination of the conductivity data, and which is in particular configured accordingly.

[0043] The eddy current sensor can define a receptacle into which the metal material can be at least partially inserted for measuring the conductivity data. In particular, the metal material can be passed through the receptacle at least partially, in particular entirely. The eddy current sensor can comprise an annular receptacle body having the receptacle. The receptacle can be a through-channel, in particular a cylindrical one, of the receptacle body, which through-channel leads in particular through the receptacle body. The receptacle is usually formed with one or more receiving wall surfaces of the receptacle body, which generally surround the metal material inserted into the receptacle at least partially around its circumference when the eddy current sensor is in use.Use of the eddy current sensor typically refers to a state in which the metal II material, in particular light metal material, is at least partially inserted into the receptacle for measuring the electrical conductivity data. It is advantageous if the receptacle body, in particular the receptacle wall surfaces, are made of an electrically insulating material, usually essentially with, in particular from, a ceramic material. As a rule, the excitation coil and / or the receiving coil are arranged on a side of the receptacle wall surfaces facing away from the receptacle, in particular connected to the receptacle body. The excitation coil and / or the receiving coil can be arranged at least partially, preferably essentially entirely, within the receptacle body.Alternatively and / or cumulatively, the eddy current sensor can have electrical insulation connected to the receiving body, wherein the excitation coil and / or the receiving coil are arranged between the electrical insulation and the receiving body. The electrical insulation can be formed with, in particular from, a ceramic material, for example a ceramic fiber layer, in particular a ceramic fiber braid. The receiving body can be circumferentially surrounded at least in sections by the excitation coil and / or the receiving coil. For example, the excitation coil and / or the receiving coil can each be wound spirally around the receiving body.Typically, in an insert of the eddy current sensor, the metal material, in particular light metal material, is cylindrical, wherein the metal material can be at least partially inserted into the receptacle for measuring electrical conductivity data of the metal material along a height direction of the cylindrical shape of the metal material.

[0044] Alternatively, the eddy current sensor can be designed such that, for determining the conductivity data, the excitation coil and / or the receiving coil are or will be arranged on only one side of the metal material, in particular the light metal material. The excitation coil and / or the receiving coil can be implemented, in particular separately or together, as part of a measuring head of the eddy current sensor. For determining the conductivity data, the respective measuring head, in particular the excitation coil and / or the receiving coil, can be arranged in a region of a surface of the metal II material, for example an end face of a rod-shaped form of the metal material, for example when the metal II material is an extrusion billet. It can be provided that, when determining the conductivity data, the measuring head is arranged adjacent to the surface of the metal II material or at a distance from the surface of the metal material.The measuring head can have a flat, in particular, plate-like shape, in particular such that the measuring head forms a flat sensor.

[0045] The eddy current sensor can have a temperature measuring device for determining, in particular measuring, temperature data of a metal material at least partially inserted into the receptacle for measuring electrical conductivity data. The temperature measuring device can be designed for contactless, in particular pyrometric, temperature data measurement or for contacting the metal material. The temperature measuring device can have a temperature sensor designed to be contacted with the metal material for measuring the temperature measurement data of the metal material. The temperature sensor can be a thermocouple or resistance temperature sensor. In use, the temperature sensor for determining the temperature data can be arranged on a surface of the metal material or at least partially, preferably entirely, recessed into the metal material. The eddy current sensor can expediently have several such temperature sensors.

[0046] The further object is achieved according to the invention by using an eddy current sensor for determining, in particular measuring, electrical conductivity data of a metal material, in particular a light metal material. The eddy current sensor can be designed as described in this document and / or used for determining, in particular measuring, electrical conductivity data. This applies in particular when a temperature of the metal II material is greater than 100°C, in particular greater than 200°C. In particular, the electrical conductivity data of the metal material can be measured with the eddy current sensor while the metal material has a temperature in a temperature range between 100°C and 600°C, in particular between 200°C and 600°C, preferably between 230°C and 580°C, particularly preferably between 300°C and 550°C.The conductivity measuring device, in particular the eddy current sensor, can have a data processing device, as described in particular in this document, comprising means, in particular at least one processor, adapted to carry out step b) of the method. The data processing device can be integrated into the eddy current sensor, in particular into the measuring head.

[0047] The further aim is achieved with a device of the type mentioned at the outset for producing a light metal component, if the device comprises a treatment device for the thermal, in particular thermomechanical, treatment of a light metal material, a conductivity measuring device, in particular an eddy current sensor, for determining electrical conductivity data of the light metal material during the thermal treatment of the light metal material, in particular according to step a) of the method, and devices, in particular a device for data processing, which are adapted to carry out the steps b) described in this document and preferably c) and / or optionally b0).In particular, the device for producing a light metal component, in particular as described in this document, can comprise a data processing device, which typically comprises at least one processor, and / or a temperature measuring device for determining temperature data of the light metal material during thermal treatment, in particular as part of the devices. The devices can be configured to perform the steps described in this document. The eddy current sensor can be implemented as described in this document, in particular with corresponding features and / or effects.

[0048] The treatment device can be designed to carry out a treatment procedure for the light metal material, as described in particular in this document. The treatment device can have a tempering device for carrying out the thermal treatment, in particular a tempering, of the light metal material. The treatment device can have a deformation device for carrying out the mechanical deformation of the light metal material, in particular before and / or at least partially in parallel with the thermal treatment.

[0049] The device for producing a light metal component is particularly designed to carry out the method for producing a light metal component described in this document. The device for producing a light metal component can be designed according to the features and effects described in this document within the framework of the method for producing a light metal component. The same applies analogously to the method for producing a light metal component with regard to the device for producing a light metal component.

[0050] Further features, advantages, and effects will become apparent from the following exemplary embodiments. The drawings, to which reference is made, show:

[0051] Fig. 1 is a schematic representation of a process sequence for producing a light metal component;

[0052] Fig. 2 is a circuit diagram of an eddy current sensor;

[0053] Fig. 3 to Fig. 5 are schematic representations of eddy current sensors in an operating state;

[0054] Fig. 6 and Fig. 7 diagrams of mean absolute errors of grain sizes for a cross-section of a light metal material, wherein Fig. 6 shows an example diagram for predicted grain sizes and Fig. 7 shows a common mean respective error variability as a comparison reference.

[0055] Fig. 1 schematically illustrates a process sequence for producing a light metal component, wherein the process comprises a treatment procedure in which a light metal material 9 is subjected to a thermal treatment to form the light metal component. It is often provided that, in the treatment procedure, the light metal material 9 is mechanically deformed before and / or at least partially in parallel with the thermal treatment. The mechanical deformation can be an extrusion of the light metal material 9. The thermal treatment can be an artificial aging or a homogenization of the light metal material 9, wherein the light metal material 9 is typically subjected to a tempering, in particular a temperature program.In step a) of the method, electrical conductivity data and typically temperature data of the light metal material 9 are determined during the thermal treatment, represented by reference numeral 2. It is advantageous if the conductivity data comprise real and imaginary part data of a complex electrical conductivity. The conductivity data are typically measured using a conductivity measuring device designed as an eddy current sensor 8, as shown, for example, in Fig. 2 and Fig. 3. The mechanical deformation and / or the thermal treatment are generally characterized by process parameters, represented by reference numeral 3.The process parameters can, for example, with respect to extrusion, include a bolt temperature of a bolt of a deformation device designed as an extrusion device, with which bolt a force is exerted on the light metal material 9 for extrusion, and / or an extrusion speed of the extrusion. Typically, the process parameters include a chemical composition of the light metal material 9.

[0056] In a step b) of the method, input data of a computer-implemented machine learning procedure 5 is generated, wherein the input data comprises the conductivity data, the temperature data, and typically the process parameters. The machine learning procedure 5 is typically implemented as a neural network. The machine learning procedure 5 is used to predict material properties, in particular a microstructure, for example a grain structure, of the light metal component, represented by reference numeral 6. It is advantageous if, in a step b0) of the method carried out before step b), the conductivity data are reduced by performing a Fourier transformation and / or an encoder-decoder architecture. In this way, a data volume representing the conductivity data and / or data complexity can be reduced. For example, the encoder-decoder architecture can be implemented as an autoencoder.

[0057] In a step c) of the method, a comparison of the predicted material properties with comparison material properties is carried out and preferably manufacturing parameters for the production of the light metal component are adapted based on the predicted material properties, represented by reference numeral 7. In particular, the manufacturing parameters comprise, in particular are, one or more of the process parameters.

[0058] A further method for producing a light metal component can be implemented in an analogous manner if, in step b), instead of predicting material properties, a prediction of production deviations in the production of the light metal component is implemented using a computer-implemented machine learning procedure (5). And if, in step c), instead of comparing the predicted material properties with comparative properties and preferably adapting the production parameters based on the predicted material properties, a determination of a process anomaly in the production, in particular in series production, of the light metal component is implemented based on the predicted production deviations and a preferably adaptation of production parameters in the production of the light metal component is implemented based on the predicted production deviations. This can be carried out in a manner analogous to the implementation described above.

[0059] Fig. 2 shows a circuit diagram of an eddy current sensor 8 and Figs. 3 to 5 show schematic representations of correspondingly implemented eddy current sensors 8. The eddy current sensor 8 can be used to determine, in particular measure, the conductivity data in the method described for Fig. 1. The eddy current sensor 8 is designed for the inductive determination of electrical conductivity data of a metal material, in particular a light metal material 9, wherein the eddy current sensor 8 has an excitation coil 11 and a receiving coil 12 in order to induce eddy currents in the metal II material with the excitation coil 11, usually by applying an electrical alternating voltage to the excitation coil 11, and to detect, in particular measure, a magnetic field generated by the eddy currents with the receiving coil 12 in order to determine the electrical conductivity data.The excitation coil 11 and / or the receiving coil 12 are each implemented as a wire coil, which is formed, in particular, from tungsten or a tungsten alloy, or platinum or a platinum alloy, or gold or a gold alloy. This makes it possible to determine the conductivity data in a temperature range of the metal material greater than 100°C, in particular greater than 200°C, preferably in a temperature range of the metal II material between 200°C and 600°C. With the eddy current sensor 8, real part data and imaginary part data of a complex electrical conductivity or the conductivity data can be determined, in particular measured. R. gen , Rterm and R CU rrent denotes electrical resistances in the circuit diagram.

[0060] In a first variant of the eddy current sensor 8, the eddy current sensor 8 can have an annular receiving body with a receptacle, wherein the metal material for measuring electrical conductivity data of the Meta II material can be at least partially inserted into the receptacle, in particular can be passed through it, as shown in Fig. 3. The receptacle can be a through-channel, in particular a cylindrical one, of the receiving body. The receptacle is formed with one or more receiving wall surfaces of the receiving body, which, during use, usually surround the metal material inserted into the receptacle at least in sections. The receiving body, in particular the receiving wall surfaces, are made of an electrically insulating material, usually essentially with, in particular from, a ceramic material. The excitation coil 11 and / or the receiving coil 12 are arranged on a side of the receiving wall surfaces facing away from the receptacle.The excitation coil 11 and / or the receiving coil 12 can be arranged at least partially, preferably substantially entirely, within the receiving body. Alternatively and / or additionally, the eddy current sensor 8 can have electrical insulation connected to the receiving body, with the excitation coil 11 and / or the receiving coil 12 being arranged between the electrical insulation and the receiving body. The electrical insulation can be formed with, in particular, a ceramic material, for example, a ceramic fiber braid.

[0061] In a second variant, the eddy current sensor 8 can comprise a measuring head having the excitation coil 11 and the receiving coil 12, wherein the measuring head is designed for the one-sided arrangement of the excitation coil 11 and / or the receiving coil 12 on the Meta II material. Such eddy current sensors 8 are shown schematically in Fig. 4 and Fig. 5. To determine the electrical conductivity data, the measuring head can be arranged in a region of a surface of the Meta II material, for example on an end face of a rod-shaped form of the Metal I material. The measuring head can have a flat, in particular plate-shaped, shape or form a flat sensor. The excitation coil 11 and the receiving coil 12, in particular essentially a totality of their respective windings, can be arranged one behind the other, in particular spaced apart from one another, and in particular wound around a coaxial winding axis W. This is shown schematically in Fig. 5.The winding axis W can be oriented transversely, in particular substantially orthogonally, to a measuring surface of the measuring head facing a material surface of the metal material.

[0062] The eddy current sensor 8 has a temperature measuring device 10 with one, in particular several, temperature sensors in order to determine, in particular to measure, temperature data of a metal material at least partially inserted into the receptacle for measuring electrical conductivity data. The temperature sensor can be, for example, a resistance temperature sensor, in particular a platinum resistance temperature sensor, for example a Pt 1000. It is preferably provided that, in use, the temperature sensor for determining the temperature data is at least partially, preferably entirely, sunk into the metal material. The temperature measuring device 10, in particular the temperature sensor, can be designed as a separate component with respect to the measuring head, shown in Fig. 4, or can be designed as part of the measuring head, shown in Fig. 5. In particular, the temperature sensor can form part of the measuring surface of the measuring head.

[0063] Fig. 6 and Fig. 7 show diagrams of mean absolute errors or uncertainties, referred to as MAE, of grain sizes, in pm, of a predicted grain structure of a cross-section of a light metal material 9. The diagram in Fig. 6 relates, by way of example, to grain sizes of an aluminum-based alloy, in particular an EN AW-6005A alloy or AISiMg(A) alloy, predicted using a computer-implemented machine learning procedure 5 of the method described in Fig. 1. The diagram in Fig. 7 represents a typical mean error variability of the predicted grain sizes as a comparison reference. The grain structure in the diagram in Fig. 6 has an average mean absolute error of 7.19 pm and the grain structure in the diagram in Fig. 7 has an average mean absolute error of 31.73 pm.It can be seen that the grain structure can be predicted with the computer-implemented machine learning procedure 5 based on the electrical conductivity data with very low mean absolute errors.

Claims

Patent claims 1. A method for producing a light metal component, in particular from an aluminum-based alloy, wherein a light metal material (9) is subjected to a thermal, in particular thermomechanical, treatment for forming the light metal component, the method comprising the following steps: a) determining electrical conductivity data of the light metal material (9) during the thermal treatment of the light metal material (9); b) predicting material properties, in particular a microstructure, of the light metal component using a computer-implemented machine learning procedure (5), wherein input data of the machine learning procedure (5) are formed using the conductivity data; c) comparing the predicted material properties with comparative material properties and preferably adapting manufacturing parameters for the production of the light metal component based on the predicted material properties.

2. Method according to claim 1, characterized in that the conductivity data comprise real part data and imaginary part data of a complex electrical conductivity.

3. Method according to one of claims 1 or 2, characterized in that the conductivity data are determined with an eddy current sensor (8).

4. Method according to claim 3, characterized in that the eddy current sensor (8) for an inductive determination of the conductivity data has a wire coil formed with a non-ferromagnetic metal, preferably with tungsten or a tungsten alloy or platinum or a platinum alloy or gold or a gold alloy, wherein the wire coil is dimensionally stable in a temperature range between 100°C and 600°C.

5. Method according to one of claims 1 to 4, characterized in that in step a) a determination of temperature data of the light metal material (9) takes place during the thermal treatment of the light metal II material (9), wherein in step b) the input data are formed with the temperature data.

6. Method according to one of claims 1 to 5, characterized in that the conductivity data are determined in a temperature range of the light metal material greater than 100°C, in particular between 200°C and 600°C.

7. Method according to one of claims 1 to 6, characterized in that the input data of the machine learning procedure (5) comprise process parameters of the thermal treatment and / or process parameters of a mechanical deformation of the light metal material (9) carried out before and / or at least partially in parallel with the thermal treatment.

8. The method according to claim 7, characterized in that the adjustment of the manufacturing parameters comprises an adjustment of one or more of the process parameters.

9. Method according to one of claims 1 to 8, characterized in that in step c) a deviation error is formed based on the comparison of the predicted material properties with the comparison material properties, after which the adjustment of the manufacturing parameters takes place so that the deviation error is reduced.

10. Method according to one of claims 1 to 9, characterized in that the conductivity data are reduced before carrying out step b) using a compression method, in particular carrying out a Fourier transformation and / or an encoder-decoder architecture.

11. The method according to one of claims 1 to 10, characterized in that the machine learning procedure (5) is formed with a decision tree and / or a neural network, in particular a feed-forward neural network or a recurrent neural network.

12. A method for producing a light metal component, in particular from an aluminum-based alloy, wherein a light metal material (9) is subjected to a thermal, in particular thermomechanical, treatment for the formation of the light metal component, the method comprising the following steps: a) Determination of electrical conductivity data of the light metal material (9) during the thermal treatment of the light metal material (9); b) Prediction of production deviations in the manufacture of the light metal component using a computer-implemented machine learning procedure (5), wherein input data of the machine learning procedure (5) are formed using the conductivity data; c) Determination of a process anomaly in the manufacture, in particular in series production, of the light metal component based on the predicted production deviations and preferably adjustment of manufacturing parameters for the manufacture of the light metal component based on the predicted production deviations.

13. Device for data processing, comprising means, in particular at least one processor, which are adapted to carry out step b) of the method according to one of claims 1 to 12.

14. Eddy current sensor (8) for determining electrical conductivity data of a metal material, characterized in that the eddy current sensor (8) for an inductive determination of the conductivity data has a wire coil formed with a non-ferromagnetic metal, preferably with tungsten or a tungsten alloy or platinum or a platinum alloy or gold or a gold alloy, wherein the wire coil is dimensionally stable in a temperature range between 100°C and 600°C.

15. Use of an eddy current sensor (8) according to claim 14 for determining electrical conductivity data of a Meta II material, in particular at a temperature of the Meta II material greater than 100°C, in particular in a temperature range between 200°C and 600°C.

16. Device for producing a light metal component, comprising a treatment device for the thermal, in particular thermomechanical, treatment of a light metal material (9), a conductivity measuring device, in particular an eddy current sensor (8), for determining electrical conductivity data of the light metal material (9) during the thermal treatment of the light metal material (9) and devices, in particular a device for data processing, which are adapted to carry out step b) of the method according to one of claims 1 to 12.

Images