Method for characterizing at least one defect of a component by means of an ultrasonic testing method using an ultrasonic device, computer program product, computer-readable storage medium, and ultrasonic device

The method employs a phased-array ultrasonic probe with TFM and parametric modeling to overcome limitations in conventional ultrasonic testing, achieving precise defect characterization in components with limited accessibility, enhancing defect detection and evaluation accuracy.

WO2026233619A1PCT designated stage Publication Date: 2026-11-12SIEMENS ENERGY GLOBAL GMBH & CO KG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/EP2026/060015
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-05
Filing Date
2026-04-07
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Conventional ultrasonic testing methods face challenges in detecting and evaluating small defects in components with limited accessibility due to limitations in probe positioning and signal refraction, leading to inaccurate defect characterization, especially in thin and radiolucent components, and requiring costly alternatives like X-ray inspection with radiation protection measures.

Method used

A method utilizing a phased-array ultrasonic probe with total focus method (TFM) and parametric modeling to reconstruct defect indications from multiple test positions, enabling precise defect characterization by fitting model parameters without echo maximization, suitable for components with limited accessibility.

Benefits of technology

Enables accurate detection and evaluation of small defects with improved resolution and localization, reducing the need for costly alternatives and radiation protection, while providing detailed defect information and classification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2026060015_12112026_PF_FP_ABST
    Figure EP2026060015_12112026_PF_FP_ABST
Patent Text Reader

Abstract

The invention relates to a method for determining a defect (14) of a component (12) using an ultrasonic device (10), having the steps of: applying ultrasound (18) to the component (12) over at least two test positions by means of a phased array ultrasonic test head (16); detecting ultrasound (18) reflected by the component (12) for the at least two test positions; carrying out an information matrix detection process for the at least two test positions; reconstructing the at least two test positions by means of a total focus method; determining at least one component defect (14) relative to the phased array ultrasonic test head (16) on the basis of the at least two reconstructed test positions; comparing the at least one determined component defect (14) with a specified parametric model (22) for component defects; selecting a component defect from the parametric model (22) on the basis of the comparison; adapting at least one model parameter (26) of the selected component defect to the determined component defect; and characterizing the at least one component defect (14) on the basis of the at least one determined model parameter (26).
Need to check novelty before this filing date? Find Prior Art

Description

[0001] 2024PF00055

[0002] Description

[0003] Method for characterizing at least one component defect of a component using an ultrasonic testing method by means of a

[0004] Ultrasound device, computer program product, computer-readable storage medium and ultrasound device

[0005] The following invention relates to a method for characterizing at least one component defect of a component by means of an ultrasonic testing method using an ultrasonic device according to claim 1. The invention further relates to a corresponding computer program product, a corresponding computer-readable storage medium and a corresponding ultrasonic device.

[0006] The following invention relates in particular to the problem of ultrasonic testing of components with limited accessibility. In conventional ultrasonic testing, small defects are evaluated based on the echo amplitude. For this, the defect echo must be generated, meaning it must be maximized by moving the probe on the component surface—in the case of contact technology—until the defect is illuminated perpendicularly. However, the probe's position is limited by the component's shape. A similar situation exists with immersion technology, as the ultrasound is refracted upon entering the component surface. If the defect echo cannot be maximized due to limited accessibility, moving the defects becomes very inaccurate or even impossible. This applies not only to the defect size but also to the defect characteristics, such as the geometry or morphology, which are important for evaluation.

[0007] For very thin components that are easily radiolucent, defects can alternatively be detected using X-ray inspection, 2024PF00055

[0008] 2

[0009] However, this involves greater effort and expense compared to ultrasound testing and also requires radiation protection measures, such as structural modifications like shielding or an X-ray booth, access restrictions, and occupational safety. Even X-rays are insufficient to obtain information about the defect morphology. For this, a suitable X-ray CT scanner would be required, which involves greater effort and higher costs.

[0010] In the field of ultrasonic testing, there is TOFD (time-of-flight diffraction) testing, which does not require the vertical acquisition of defects. However, TOFD testing requires the use of two probes in a V-configuration, which generally entails even higher access requirements. Furthermore, it can happen that the actual defect indication is masked by small, adjacent reflectors, for example, a weld defect with many small pores in the surrounding area.

[0011] Ultrasonic testing is an established method for the non-destructive inspection of components for internal defects such as cracks, pores, or inclusions. In this process, ultrasonic waves are introduced into the component and reflected at the interfaces between different materials. Sensors capture and analyze these reflections to obtain information about the location, shape, and size of the defects. However, as already mentioned, a problem with ultrasonic testing of components is their limited accessibility, meaning that smaller defects often cannot be clearly detected and assessed. This is because the echo amplitude, i.e., the strength of the reflected signal, depends on various factors, such as the orientation of the defect relative to the probe position, as well as the shape and depth of the defect.

[0012] To generate a maximum defect echo, the probe must be moved so that the defect is illuminated perpendicularly. 2024PF00055

[0013] 3

[0014] In contact technology, where the probe is placed directly on the component surface, the component shape limits the available options. In immersion technology, where the ultrasonic device is immersed in a liquid-filled test medium and the ultrasonic signal is transmitted through the liquid into the component, the signal is refracted upon entering the component surface, which can also impair the detectability of defects. If the defect echo cannot be maximized due to limited accessibility, defect evaluation becomes highly inaccurate or even impossible. This applies not only to defect size but also to defect characteristics, i.e., the geometry and morphology of the defect, which are equally important for a correct assessment. Therefore, it is crucial to develop suitable testing techniques and evaluation strategies to reliably detect and evaluate even small defects in components with limited accessibility.

[0015] The object of the present invention is to provide a method, a computer program product, a computer-readable storage medium and an ultrasound device by means of which the disadvantages of the prior art can be overcome.

[0016] In particular, the object of the present invention is to provide a method, a computer program product, a computer-readable storage medium and an ultrasound device by means of which a component defect can be reliably characterized even in the case of limited accessibility to a component.

[0017] This problem is solved by a method, a corresponding computer program product, a corresponding computer-readable storage medium, and a corresponding ultrasound device according to the independent claims. Advantageous embodiments are specified in the dependent claims. 2024PF00055

[0018] 4

[0019] One aspect of the invention relates to a method for determining at least one component defect using an ultrasonic testing method with an ultrasonic device. The component is subjected to ultrasound at at least two test positions using a phased-array ultrasonic probe of the ultrasonic device. The phased-array probe detects the ultrasound reflected by the component at these two test positions. An information matrix is ​​acquired for these two test positions using an electronic computing unit of the ultrasonic device. The data from these two test positions are then reconstructed using a total focus method (TFM) with the electronic computing unit.The process involves determining at least one component defect relative to the phased-array ultrasonic probe, depending on the at least two reconstructed test positions, using the electronic computing unit. The at least one identified component defect is then compared with a predefined parametric model for component defects using the electronic computing unit. Based on this comparison, one component defect is selected from the parametric model. Finally, at least one model parameter of the selected component defect is fitted to the identified defect using the electronic computing unit, and the at least one component defect is characterized as a function of this at least one specific model parameter using the electronic computing unit.

[0020] In particular, the system is designed to scan the component and perform a corresponding measurement, for example, FMC (Full Matrix Capture), for each test position. The same test area is reconstructed using TFM for each test position. The position of any detected defects relative to the probe is then determined. (2024PF00055)

[0021] 5

[0022] A parametric model is fitted to these defect indications. The model parameters that provide the best fit are then determined. A defect assessment can then be performed based on these model parameters.

[0023] In particular, it is therefore intended that even with correspondingly limited space for testing the component, a reliable defect detection can still be achieved using the proposed method.

[0024] In the prior art, the value of the TFM amplitude sum is not used for evaluation. In contrast to a conventional TFM, the invention involves multiple TFM reconstructions of the same test area, utilizing data from differently positioned apertures, for example, sub-apertures of a phased-array probe or probe movement. This allows for the evaluation of amplitude changes when the probe is illuminated from different directions, enabling defect assessment by fitting the parametric model to the test results without echo maximization.

[0025] This allows for a precise evaluation of small TFM indications, especially when smaller relative to the wavelength. The indication height, which depends on the relative location, can also be taken into account for FMC / TFM. In particular, the method can be used even in areas with limited accessibility, or is especially advantageous in such situations. The method also provides additional information for defect assessment. It can also be used for wall thickness reduction or wear testing as part of reversal inspection. The direction of the load, for example, mechanical stresses, can also be considered when deciding on admissibility. The strength of the directivity can also be taken into account when deciding on admissibility; for example, the magnitude of the directivity typically correlates with greater material separation or cross-sectional weakening.

[0026] This results in a reduction in load-bearing capacity and the appearance of crack tips, which can lead to increased crack growth. The invention also enables the differentiation of defect types. Furthermore, it allows for the highlighting of specific reflector types in the results display, particularly for better interpretation of critical indications. This enables a very high level of defect resolution, similar to that achieved with prior art TFM methods.

[0027] In other words, the advantage of this method lies in the fact that it provides a method for characterizing component defects through ultrasonic testing, based on a total focus method and a parametric model. Using the total focus method, defects located at different depths within the component can be detected, enabling more precise localization and evaluation of defects. By analyzing this measurement data, the electronic computing unit can detect and locate potential defects by comparing the reflected signal with the parametric model. The method also allows the parametric model to be adapted to the potential defect, leading to a more accurate determination and characterization of the defect.Finally, the defect can be determined by identifying at least one parameter of the parametric model, enabling a more precise analysis and evaluation of the defect. Overall, this method thus offers a more accurate and reliable approach for detecting and evaluating component defects using ultrasonic testing, particularly in applications where accessibility is limited or when testing complex components.

[0028] Information matrix capture includes, for example, full matrix capture. However, in addition to full matrix capture, half matrix capture or sparse matrix capture is also possible. (2024PF00055)

[0029] 7

[0030] Capture methods can be used. In particular, methods for generating A-scans or time-signal images can be employed. Full matrix capture (FMC) is preferred. It is crucial to understand that FMC does not represent the complete inspection procedure, but merely the data acquisition level. The core of FMC is the information matrix. This matrix contains the A-scans (signal waveforms) of all possible combinations of transmitting and receiving phased array elements. For a phased array with N elements, the information matrix then comprises N x N A-scans. Each A-scan represents the echo that travels from a specific transmitting element to a specific receiving element. Acquiring the full matrix data is thus one step in the overall procedure. The interpretation of this data is then performed using the total focus method. Only the combination of full matrix capture and the total focus method constitutes a complete inspection procedure.

[0031] Full-matrix acquisition offers several advantages over conventional ultrasonic testing: The simultaneous measurement of multiple sound paths enables the generation of high-resolution images, allowing for more precise analysis and inspection. The use of a full array of transmitters and receivers ensures comprehensive coverage of the inspection area, reducing the likelihood of hidden defects. Parallel data acquisition, instead of the sequential measurement used in single-transducer systems, significantly reduces inspection time. Rotating and shifting the full-matrix sensor allows for the creation of a three-dimensional image of the test object, which is advantageous for more complex geometries. The simultaneous measurement of multiple sound paths enables a more accurate determination of the ultrasound propagation direction, leading to improved defect characterization.Full matrix acquisition is used in various application areas, including the automotive industry, aerospace, and energy generation.2024PF00055

[0032] as well as in shipbuilding, to detect material defects such as cracks, delaminations, porosities or corrosion.

[0033] Total focus is a technique in ultrasonic testing where every point within the test area is focused. All sound waves in the beam are bundled and concentrated on a specific area of ​​the test object, resulting in increased sensitivity and resolution. Unlike conventional focusing techniques, where the ultrasonic beam is always focused in the same way, total focus allows each point within the test volume to be individually (optimally) focused differently.

[0034] According to an advantageous embodiment, at least one defect size and / or defect orientation and / or the strength of a defect directional effect and / or a defect morphology and / or a roughness of the component defect are determined and, in particular, characterized. Specifically, a defect evaluation can thus be performed based on the determined model parameters. The corresponding evaluation parameters regarding the severity, for example, a rice-grain-like defect in forged parts if, for instance, cavities were not flattened during forging, or the like, can be determined. Furthermore, only subsets of the model parameters can be created in the parametric model. This makes it possible to use different evaluation parameters for the defect in order to perform a reliable and detailed evaluation of the component defect.

[0035] Another advantageous embodiment provides that the specified parametric model is simulated and created based on a progression of maximum amplitudes from a total focus method and / or that a sum amplitude dynamic is determined for the specific component defect 14. In particular, this can be used as a preliminary measure for the 2024PF00055

[0036] 9

[0037] Determining the component defect can be considered. For this purpose, it may be particularly useful to simulate the location-dependent progression of the maximum amplitude of TFM defect indications. Advantageously, this can be achieved by simulating the TFM scan where the width of the phased array elements (pitch) is equal to the scan grid, enabling rapid computation. The creation of the parametric model of the TFM maximum amplitude progression based on the simulation results can be carried out, for example, using parameter variations such as directional / non-directional reflector, different defect sizes, and different defect orientations.

[0038] Furthermore, it has proven advantageous to calculate the given parametric model on the basis of a modified total focus method, for which a scattering characteristic is used.

[0039] In particular, during the preparatory simulation, a modified TFM is calculated for each test position. This modified TFM is based on the scattering characteristic, where the distance and probe directional characteristics of the transmit and receive sound fields have been subtracted. This modified TFM is then used to determine the model parameters that provide the best fit. The scattering characteristic of a defect is calculated by subtracting the distance and probe directional characteristics from the echo signal using the following formula, where A(...) is the distance characteristic and D(...) denotes the probe directional characteristic:

[0040]

[0041] With r t as a sound path from the transmitter element to the defect, r r as sound path from the defect to the receiver element, θ t as angle between transmitter element and defect and θ r as angle between receiver element and defect. 2024PF00055

[0042] 10

[0043] Furthermore, it has proven advantageous to generate the parametric model using linear regression or machine learning. This allows the parametric model to be determined based on various mathematical methods. Depending on the application and the available data, the best possible model can thus be provided. This ensures a highly reliable parametric model, which in turn leads to a reliable determination of component defects.

[0044] In a further advantageous embodiment, it is provided that a line scan and / or an area scan of the component is performed. Thus, the component can be scanned / proven in different ways.

[0045] In particular, the method is therefore applicable to different ultrasonic devices that perform different scanning processes. Furthermore, depending on the shape of the component, it can be determined whether line scanning or area scanning should be implemented. With line scanning, the scan is performed only in one direction, and the defect characteristics can only be displayed in this direction. With area scanning, the probe is additionally moved perpendicular to the line scan, which allows for defect characterization in this additional direction.

[0046] Another advantageous embodiment provides that the at least two test positions are generated by means of a probe movement of the phased-array ultrasonic probe and / or electronically by means of partial aperture formation. Thus, the scan can be performed in different ways, making the method suitable for various ultrasonic devices. Furthermore, the method can therefore also be used under different external conditions, for example in 2024PF00055

[0047] 11

[0048] The process depends on the component or the available space.

[0049] According to a further advantageous embodiment, a defect database for future use is created using the parametric model and / or at least one specific component defect and / or a characterized component defect. In particular, a defect database can thus be created, and the similarity of detected indications to entries in the defect database can be determined. The database can also be expanded with metallographic representations of previously detected / analyzed defects. This allows the method to be further refined in the future by storing the corresponding defects and then, for example, considering or querying them when adapting / fitting the potential component defect.

[0050] In a further advantageous embodiment, a defect classification is determined based on the specific component defect. For example, a classification algorithm can be used to determine the similarity to entries in the defect database, such as through logistic regression, a classification algorithm, or AI-based classification.

[0051] Another advantageous design provides that the usability of the component is determined depending on the specific component defect.

[0052] For example, a threshold value can be defined for this purpose. If the component defect is sufficiently severe and exceeds the threshold value, the component can be classified as defective and, for example, not used. However, if the component defect is below a corresponding threshold value, it is possible that the component can continue to be used despite the detected defect. 2024PF00055

[0053] 12

[0054] can be used. It is of course also possible to use the other parameters to perform a "OK / not OK" classification.

[0055] Furthermore, it has proven advantageous to classify the component defect as either a directional or non-directional reflector, depending on the specific defect. In particular, the parametric model can be extended to differentiate between further reflector types beyond directional / non-directional, or by including a parameter to describe roughness. Additionally, shadow indication can be considered, for example, an additional indication outside the reflector surface, such as the propagation of surface waves at the defect.

[0056] In a further advantageous embodiment, the selection of a component defect from the parametric model involves determining the similarity of the reconstructed component defect using logistic regression and / or a classification algorithm and / or an artificial intelligence-based classification. This allows for the determination of similarities and thus enables appropriate adjustment / fitting. In particular, a best-fit algorithm can be used to identify the best possible similarity and thus reliably determine the component defect.

[0057] The presented method is in particular a computer-implemented method. Therefore, a further aspect of the invention relates to a computer program product with program code means which cause an electronic computing device to execute the program code means when the electronic computing device is encoded.

[0058] 13

[0059] to be processed, a procedure according to the preceding aspect.

[0060] Furthermore, the invention also relates to a computer-readable storage medium with at least one computer program product according to the preceding aspect.

[0061] A further aspect of the invention relates to an ultrasonic device for determining at least one component defect of a component using an ultrasonic testing method, comprising at least one ultrasonic probe and an electronic computing unit, wherein the ultrasonic device is configured to carry out a method according to the preceding aspect. In particular, the method is carried out using the ultrasonic device.

[0062] Advantageous embodiments of the process are to be regarded as advantageous embodiments of the computer program, the computer-readable storage medium, and the ultrasound device. In this respect, the ultrasound device possesses, in particular, tangible features to enable the corresponding process steps to be carried out.

[0063] A computing unit / electronic computing device can be understood, in particular, as a data processing device containing a processing circuit. The computing unit can therefore process data to perform arithmetic operations. This may also include operations to perform indexed access to a data structure, such as a lookup table (LUT).

[0064] The computing unit can, in particular, contain one or more computers, one or more microcontrollers and / or one or more integrated circuits, for example, one or more application-specific 2024PF00055

[0065] 14

[0066] The computing unit may consist of integrated circuits (ASICs, application-specific integrated circuits), one or more field-programmable gate arrays (FPGAs), and / or one or more systems on a chip (SoCs). It may also include one or more processors, such as one or more microprocessors, one or more central processing units (CPUs), one or more graphics processing units (GPUs), and / or one or more signal processors, in particular one or more digital signal processors (DSPs). The computing unit may also include a physical or virtual cluster of computers or other units of the aforementioned type.

[0067] In various embodiments, the computing unit includes one or more hardware and / or software interfaces and / or one or more storage units.

[0068] A storage unit can be volatile data storage, for example as dynamic random access memory (DRAM) or static random access memory (SRAM), or as non-volatile data storage, for example as read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory or flash EEPROM, ferroelectric random access memory (FRAM), or magnetoresistive random access memory.MRAM (English: "magnetoresistive random access memory") or as phase-change memory with random access, PCRAM2024PF00055,

[0069] 15

[0070] (English: "phase-change random access memory") ), be designed.

[0071] For use cases or application situations that may arise in a method according to the invention and that are not explicitly described herein, it may be provided that, according to the method, an error message and / or a request for user feedback is issued and / or a default setting and / or a predetermined initial state is set.

[0072] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identities are included.

[0073] Further features and combinations of features of the invention will become apparent from the figures and their descriptions, as well as from the claims. In particular, further

[0074] Embodiments of the invention do not necessarily include all features of any one of the claims. Further embodiments of the invention may have features or combinations of features not mentioned in the claims.

[0075] This shows:

[0076] Fig. 1 shows a schematic perspective view of an embodiment of an ultrasound device;

[0077] Fig. 2 shows a schematic flowchart according to one embodiment of the method;

[0078] Fig. 3 is a schematic diagram of amplitude sum curves according to the prior art; and

[0079] Fig. 4 shows another schematic diagram of amplitude sums according to one implementation of the method. 2024PF00055

[0080] 16

[0081] In the figures, identical or functionally equivalent elements are provided with the same reference symbols.

[0082] Fig. 1 shows a schematic perspective view of an embodiment of an ultrasonic device 10 for performing an ultrasonic testing procedure on a component 12. In the present embodiment, the component 12 has a component defect 14. Using the method according to the invention, the component defect 14 can be determined and characterized accordingly via the ultrasonic device 10.

[0083] For this purpose, the ultrasound device 10 comprises at least one phased-array ultrasound transducer 16 for emitting and receiving ultrasound 18. Furthermore, the ultrasound device 10 comprises at least one electronic computing unit 20, in particular for performing information matrix acquisition, preferably full-matrix capture (FMC) and / or total-focus method (TFM) testing. The electronic computing unit 20 has, in particular, a parametric model 22. Furthermore, the electronic computing unit 20 can communicate with a defect database 24. The defect database 24 can be part of the electronic computing unit 20 or the ultrasound device 10, or it can be provided, for example, as a backend server or cloud server.

[0084] Fig. 2 shows a schematic flowchart according to one embodiment of the method. In step S1, the component 12 is subjected to ultrasound 18 at at least two test positions using the phased-array ultrasonic transducer 16. In a second step S2, ultrasound 18 reflected by the component 12 is detected for the at least two test positions using the phased-array ultrasonic transducer 16. In a third step S3, an information matrix acquisition, in particular the FMC-2024PF00055, is performed.

[0085] 17

[0086] Testing is performed for which at least two test positions are carried out using an electronic computing device 20. In a fourth step S4, the at least two test positions are reconstructed using a total focus method (TFM) with the electronic computing device 20. In a fifth step S5, the at least one component defect is determined relative to the phased-array ultrasonic probe as a function of the at least two reconstructed test positions using the electronic computing device 20. In a sixth step S6, the at least one specific component defect is determined using the predefined parametric model 22 for component defects with the electronic computing device 20. In an optional further step, the sum amplitude dynamics are determined for the previously determined component defect 14, i.e., the amplitude profile in the several TFM reconstructions at the location of the potential component defect 14.In a seventh step S7, a component defect is selected from the parametric model 22 (Fig. 2) based on a comparison using the electronic computing device 20. In an eighth step S8, at least one model parameter 26 of the selected component defect is adapted to the specific component defect using the electronic computing device 20. In a ninth step S9, the at least one component defect 14 is characterized based on the at least one specific model parameter 26 using the electronic computing device 20.

[0087] In particular, it may be provided that at least one defect size and / or one defect orientation and / or one defect morphology and / or one defect directional effect and / or one defect morphology and / or one roughness of the component defect 14 is determined and characterized.

[0088] Furthermore, it may be provided that the specified parametric model 22 is simulated and created based on a profile of maximum amplitudes of a total focus method. It may also be provided that the 2024PF00055

[0089] 18

[0090] A predefined parametric model 22 is modified based on a scattering characteristic with respect to the distance and probe directional characteristics of the transmit and receive sound fields. The parametric model 22 can be generated using linear regression or machine learning.

[0091] Furthermore, it may be provided that a line scan of the component 12 and / or an area scan of the component 12 is performed. According to one embodiment, the at least two test positions can be generated by means of a probe movement of the phased-array ultrasonic probe 16 and / or electronically by means of partial aperture formation.

[0092] It can also be provided that a defect database 24 is created for future use by means of the parametric model 22 and / or by means of at least one specific component defect 14. Depending on the specific component defect 14, a defect classification can be determined. Likewise, depending on the specific component defect 14, the usability of the component 12 can be determined. Furthermore, depending on the specific component defect 14, the component defect 14 can be classified as a directional reflector or as a non-directional reflector.

[0093] It may also be provided that, in order to adapt the component defect 14 to the parametric model 22, a similarity is carried out by means of a logistic regression and / or a classification algorithm and / or a classification based on artificial intelligence.

[0094] In other words, the proposed embodiment involves scanning component 12 and performing an FMC measurement for each test position. The same test area is used for each test step with TFM2024PF00055.

[0095] 19

[0096] The system is reconstructed. The position of a detected defect indication relative to the ultrasonic probe 16 is then determined. The parametric model 22 is fitted to the detected defect indication. At least one model parameter 26 is determined that provides the best fit. The defect evaluation can then be performed based on the determined model parameters.

[0097] Advantageously, in the preparatory simulation, a modified TFM is calculated for each test position, based on the scattering characteristics where the distance and probe directional characteristics of the transmit and receive sound fields have been factored out. This modified TFM is then used to determine the model parameters 26 that provide the best fit. It is also possible to determine evaluation parameters such as the strength of the directivity.

[0098] Furthermore, the parametric model 22 can also create only a subset of the possible model parameters 26, in particular depending on the component 12 or the component defect 14. The parametric model 22 can also be extended, for example by differentiating further reflector types or by adding a parameter to describe the roughness.

[0099] The parametric model 22 can be determined via linear regression or machine learning. The creation of a defect database 24 and the determination of the similarity of found displays to entries in the defect database 24 can also be implemented. Furthermore, a classification algorithm can be used to determine the similarity with entries in the defect database 24, whereby, for example, logistic regression, classification algorithms, and AI-based classifications can be used. The evaluation parameter and an admissibility threshold can be defined.

[0100] 20

[0101] Defects can also be entered into the defect database 24, for example, from the simulation of crack tip stresses. The defect database 24 can also be expanded with metallographic characteristics of previously found / analyzed defects. Furthermore, a reliability measure or probability for evaluation based on the scan length can be determined. Separate evaluation of partial indications can also be implemented. Shadow indications can also be taken into account.

[0102] Fig. 3 shows two schematic amplitude sum curves for three defects. On the left side of Fig. 3, a reconstruction from the echo signal is shown, which corresponds to the prior art, and on the right side of Fig. 4, a reconstruction from the scattering characteristics of the defects according to the inventive method is shown, which allows conclusions to be drawn about the morphology of the defect. The scan positions in mm are shown on the abscissa and the amplitude sums on the ordinate. Three lines 28, 30, and 32 are also shown. Line 28 shows a spherical defect of 1 mm, line 30 shows a flat-bottomed bore of 1 mm, and line 32 shows a circular crack of 1 mm. 2024PF00055

[0103] Reference symbol list

[0104] 10 Ultrasound device

[0105] 12 components

[0106] 14 Component defect

[0107] 16 Ultrasonic probe

[0108] 18 Ultrasound

[0109] 20 electronic computing device 22 parametric model

[0110] 24 Defect database

[0111] 26 model parameters

[0112] Line 28

[0113] Line 30

[0114] Line 32

[0115] S1 - S9 Steps of the procedure

Claims

2024PF00055 22 Patent claims 1. Method for characterizing at least one component defect ( 14 ) of a component ( 12 ) by means of an ultrasonic testing method using an ultrasonic device ( 10), comprising the steps: Applying ultrasound (18) to the component (12) via at least two test positions using a phased-array ultrasonic test head (16) of the ultrasonic device (10); Detection of ultrasound (18) reflected by the component (12) for the at least two test positions using the phased-array ultrasound probe (16); performance of an information matrix acquisition for the at least two test positions using an electronic computing unit (20) of the ultrasound device (10); Reconstructing at least two test positions using a total focus method with the electronic computing device (20); Determining at least one component defect ( 14 ) relative to the phased-array ultrasonic probe (16) depending on the at least two reconstructed test positions using the electronic computing device (20); Comparing at least one specific component defect ( 14 ) with a predefined parametric model (22 ) for component defects using the electronic computing device (20); Selecting a component defect from the parametric model (22 ) depending on the comparison using the electronic computing device (20); Adapting at least one model parameter (26) of the selected component defect to the specific component defect using the electronic computing device (20); and Characterizing the at least one component defect ( 14 ) depending on the at least one specific 2024PF00055 23 Model parameters (26) using the electronic computing device (20).

2. Method according to claim 1, characterized in that at least one defect size and / or one defect orientation and / or one defect strength and / or one defect directional effect and / or one defect morphology and / or one roughness of the component defect ( 14 ) are determined.

3. Method according to claim 1 or 2, characterized in that the specified parametric model (22 ) is simulated and created based on a progression of maximum amplitudes of a total focus method and / or an amplitude sum dynamic is determined for the specific component defect ( 14 ).

4. Method according to one of the preceding claims, characterized in that the specified parametric model (22 ) is calculated on the basis of a modified total focus method, for the calculation of which a scattering characteristic is used.

5. Method according to one of claims 3 or 4, characterized in that the parametric model (22 ) is generated using linear regression or machine learning.

6. Method according to one of the preceding claims, characterized in that a line scan of the component and / or an area scan of the component ( 12 ) is performed.

7. Method according to one of the preceding claims, characterized in that which generate at least two test positions by means of a probe movement of the phased-array ultrasonic probe (16) and / or electronically by means of a partial aperture formation.

8. Method according to one of the preceding claims, characterized in that a defect database (24) is created for future use by means of the parametric model (22) and / or by means of at least one specific component defect (14).

9. Method according to one of the preceding claims, characterized in that Depending on the specific component defect ( 14 ), a defect classification is determined.

10. Method according to one of the preceding claims, characterized in that Depending on the specific component defect ( 14 ), the usability of the component ( 12 ) is determined.

11. Method according to one of the preceding claims, characterized in that depending on the specific component defect ( 14 ) the component defect ( 14 ) is classified as a directional reflector or as an undirectional reflector.

12. Method according to one of the preceding claims, characterized in that To select a component defect from the parametric model (22 ), a similarity of the reconstructed component defect is performed via a logistic regression and / or a classification algorithm and / or an artificial intelligence-based classification.

13. Computer program product with program code means which cause an electronic computing device (20) to, when the program code means are removed from the 2024PF00055 25 electronic computing device (20) to carry out a method according to one of claims 1 to 12.

14. Computer-readable storage medium comprising at least one computer program product according to claim 13.

15. Ultrasonic device ( 10) for determining at least one component defect ( 14 ) of a component ( 12 ) by means of an ultrasonic testing method, with at least one ultrasonic probe ( 16 ) and with an electronic computing device (20 ), wherein the ultrasonic device ( 10 ) is configured to carry out a method according to one of claims 1 to 12.