Liquid forging defect detection method and system
By employing ultrasonic phased array probe excitation and signal processing techniques, the problem of suppressing the near-field effect of ultrasonic waves under material inhomogeneity in liquid forgings was solved, enabling efficient and accurate defect detection and improving the accuracy of quality assessment for liquid forgings.
Patent Information
- Application Number
- CN202511866848.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-11
- Publication Date
- 2026-01-13
AI Technical Summary
During liquid forging, factors such as the instability of molten metal flow, uneven temperature field distribution, and improper mold design can lead to defects such as porosity, shrinkage, inclusions, and cracks inside the forging. Existing ultrasonic testing methods are difficult to effectively suppress near-field effects when the material is not uniform, resulting in a decrease in signal-to-noise ratio and masking the true defects.
Ultrasonic phased array probes are used to emit ultrasonic waves in a side-mounted vertically symmetrical excitation mode. The propagation mechanism change points are extracted by the energy attenuation curve, and direct waves and scattered waves are separated. The spatial coherence of the scattered waves is used to suppress the near-field effect, and adaptive weighted fusion is performed to generate a hybrid signal matrix. Finally, defects are identified by full-focus imaging.
It achieves high-resolution, high-contrast defect imaging in liquid forgings with non-uniform material properties, effectively suppresses near-field effects, preserves defect information, and avoids signal loss caused by traditional filtering methods.
Smart Images

Figure CN121324504A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of liquid die forging, and more particularly to a liquid die forging defect detection method and system. BACKGROUND
[0002] Liquid die forging is an advanced metal forming process, which can manufacture near-net-shaped parts with complex shape, dense structure and excellent mechanical properties by injecting molten metal into a mold and solidifying under high pressure. It is widely used in high-end manufacturing fields such as aviation, aerospace and automobile. However, during the liquid die forging process, due to the instability of metal flow, uneven temperature field distribution, solidification shrinkage and improper design of mold or process parameters, micro or macro defects such as porosity, shrinkage, inclusion and crack are easily generated inside the forged parts. These defects can significantly reduce the mechanical properties, fatigue life and reliability of the parts. Therefore, efficient and accurate non-destructive testing of liquid die forgings, especially reliable identification and positioning of internal defects, has become a key technical link to ensure product quality and improve process stability. In the prior art, ultrasonic testing technology is widely used in internal quality evaluation of metal components due to its strong penetration ability, sensitivity to internal defects and quantitative evaluation. However, the traditional ultrasonic testing method usually sets a threshold for filtering and testing based on the material of the part to be tested to avoid near-field interference. However, the liquid die forging part usually has non-uniform grain structure, composition segregation and dispersed micro-pores. This material-level non-uniformity causes strong structural scattering noise during ultrasonic propagation. At this time, the traditional method may not effectively suppress near-field interference, but may filter out the structural scattering signals unique to the material being tested, which may contain defect information, resulting in a decrease in signal-to-noise ratio and even masking the real defects. Therefore, how to suppress the near-field effect of ultrasonic waves in non-destructive defect detection under the condition of non-uniform material of liquid die forgings has become a difficult problem in the industry. SUMMARY
[0003] The present application provides a liquid die forging defect detection method and system, which can suppress the near-field effect of ultrasonic waves in non-destructive defect detection under the condition of non-uniform material of liquid die forgings.
[0004] In a first aspect, the present application provides a liquid die forging defect detection method, comprising: emitting ultrasonic waves into the liquid die forging part by the ultrasonic phased array probe in a side vertical symmetry excitation mode, and collecting a plurality of ultrasonic signals propagated in the liquid die forging part; extract a time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering each time the ultrasonic wave is emitted based on the energy attenuation curve of each ultrasonic signal, and divide all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points; suppress near-field effects on each scattered wave through spatial coherence between all the scattered waves, thereby obtaining a reconstructed signal matrix that retains the scattering characteristics of defects, and performing adaptive weighted fusion of the reconstructed signal matrix and all the direct waves, to obtain a mixed signal matrix that fuses early reflection information and defect scattering characteristics; perform full-focus imaging of defects in the liquid die forging based on the echo characteristics of each signal in the mixed signal matrix, and identify and locate the defects in the liquid die forging according to the imaging result.
[0005] In some embodiments, emitting ultrasonic waves into the liquid die forging by the ultrasonic phased array probe in a side vertical symmetry excitation mode specifically includes: obtaining a nominal thickness of the liquid die forging; determining the type of ultrasonic wave based on the nominal thickness; emitting the type of ultrasonic wave into the liquid die forging by the ultrasonic phased array probe in a side vertical symmetry excitation mode.
[0006] In some embodiments, extracting a time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering each time the ultrasonic wave is emitted based on the energy attenuation curve of each ultrasonic signal specifically includes: determining the energy attenuation curve of each ultrasonic signal; determining a mutation point of the energy variation law of each ultrasonic signal according to the attenuation rate of each energy attenuation curve; determining the time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering each time the ultrasonic wave is emitted based on all the mutation points.
[0007] In some embodiments, dividing all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points specifically includes: selecting one element in the ultrasonic phased array probe as a selected element, and obtaining a time point at which the ultrasonic wave field in the liquid die forging is stable when the selected element emits ultrasonic waves; dividing all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion when the selected element emits ultrasonic waves through the time point at which the ultrasonic wave field in the liquid die forging is stable when the selected element emits ultrasonic waves; The time point at which the ultrasonic field in the liquid die forging is stable when the ultrasonic waves are emitted by the remaining array elements in the ultrasonic phased array probe is used to divide all the ultrasonic signals when the ultrasonic waves are emitted by the remaining array elements in the ultrasonic phased array probe into direct waves in the initial stage of ultrasonic propagation and scattered waves after complete diffusion of the ultrasonic waves.
[0008] In some embodiments, the near-field effect of each scattered wave is suppressed by spatial coherence between all the scattered waves, and a reconstructed signal matrix retaining defect scattering characteristics is obtained, and specifically includes: combining all the scattered waves into a scattered wave signal matrix; performing cross-correlation processing on the signals in the scattered wave signal matrix to obtain spatial coherence of any two columns of scattered wave signals at different time delays; suppressing the near-field effect along the common array element dimension by all the spatial coherence to obtain a reconstruction function estimate; reconstructing the reconstruction function estimate according to the array element index to obtain a reconstructed signal matrix retaining defect scattering characteristics.
[0009] In some embodiments, the reconstructed signal matrix and all the direct waves are adaptively weighted and fused to obtain a mixed signal matrix of early reflection information and defect scattering characteristics, and specifically includes: determining a fusion weight function based on the time at which the first echo from the bottom of the liquid die forging arrives; determining a compensation factor of ultrasonic propagation attenuation in the liquid die forging according to all the direct waves and the reconstructed signal matrix; fusing all the direct waves and the reconstructed signal matrix into a mixed signal matrix of early reflection information and defect scattering characteristics by the fusion weight function and the compensation factor.
[0010] In some embodiments, full-focus imaging of defects inside the liquid die forging based on echo characteristics of each signal in the mixed signal matrix specifically includes: performing nonlinear full-focus processing on the mixed signal matrix to obtain a preprocessed signal matrix; discretizing an imaging region into a pixel grid; extracting a corresponding signal value from the preprocessed signal matrix based on an acoustic propagation delay corresponding to each pixel in the pixel grid and performing coherent superposition, thereby generating a defect image inside the liquid die forging.
[0011] In some embodiments, a plurality of ultrasonic signals of ultrasonic propagation in the liquid die forging are acquired in a full-matrix capture mode.
[0012] In some embodiments, the ultrasonic phased array probe is a linear ultrasonic phased array probe.
[0013] In a second aspect, the application provides a liquid die forging defect detection system, comprising: The acquisition module is configured to emit ultrasonic waves into the liquid die forging by using the ultrasonic phased array probe in a side-symmetric excitation mode, and collect a plurality of ultrasonic signals propagating in the liquid die forging. The processing module is configured to extract a time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from a deterministic path dominated mode to a random scattering dominated mode based on the energy attenuation curve of each ultrasonic signal, and divide all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points. The processing module is further configured to perform near-field effect suppression on each scattered wave by spatial coherence between all the scattered waves, thereby obtaining a reconstructed signal matrix retaining defect scattering characteristics, and perform adaptive weighted fusion on the reconstructed signal matrix and all the direct waves, thereby obtaining a mixed signal matrix fusing early reflection information and defect scattering characteristics. The execution module is configured to perform full-focus imaging on the defects in the liquid die forging based on the echo characteristics of each signal in the mixed signal matrix, and identify and locate the defects in the liquid die forging according to the imaging result.
[0014] The technical scheme provided by the embodiments disclosed in the application has the following beneficial effects: In the liquid die forging defect detection method and system provided by the application, ultrasonic waves are first emitted into the liquid die forging by using the ultrasonic phased array probe in a side-symmetric excitation mode, and a plurality of ultrasonic signals propagating in the liquid die forging are collected. A time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from a deterministic path dominated mode to a random scattering dominated mode is extracted based on the energy attenuation curve of each ultrasonic signal, and all the ultrasonic signals are divided into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points. Near-field effect suppression is performed on each scattered wave by spatial coherence between all the scattered waves, thereby obtaining a reconstructed signal matrix retaining defect scattering characteristics, and adaptive weighted fusion is performed on the reconstructed signal matrix and all the direct waves, thereby obtaining a mixed signal matrix fusing early reflection information and defect scattering characteristics. Full-focus imaging is performed on the defects in the liquid die forging based on the echo characteristics of each signal in the mixed signal matrix, and the defects in the liquid die forging are identified and located according to the imaging result.
[0015] It can be seen that, by extracting the time point of the change of the ultrasonic wave propagation mechanism based on the energy attenuation curve, the present application accurately divides the ultrasonic signal into direct waves carrying early reflection information and scattered waves containing defect characteristics of the diffusion field, realizes effective separation of signal components, and then uses the spatial coherence between all scattered waves to suppress the near-field effect. This method does not depend on the assumption of a homogeneous medium, can specifically distinguish and suppress the near-field interference generated by the probe, while retaining the structural scattering signals generated by the material micro-heterogeneity and possibly containing defect information, thereby avoiding signal loss caused by traditional filtering methods. Thereafter, the processed scattered wave reconstruction signal and the direct wave are complementary to each other through adaptive weighted fusion to generate a high-quality mixed signal matrix, and finally high-resolution and high-contrast imaging of the defect is realized through full-focus imaging technology, and defect recognition is performed based on the imaging result. In summary, the present application can suppress the near-field effect of ultrasonic waves in non-destructive defect detection under the condition of uneven material of liquid die forgings. BRIEF DESCRIPTION OF DRAWINGS
[0016] Figure 1 is an exemplary flowchart of a liquid die forging defect detection method according to some embodiments of the present application; Figure 2 is an exemplary flowchart of transmitting ultrasonic waves according to some embodiments of the present application; Figure 3 is an exemplary image of full-focus imaging according to some embodiments of the present application; Figure 4 is a structural schematic diagram of a liquid die forging defect detection system according to some embodiments of the present application; Figure 5 is a structural schematic diagram of a computer device for implementing a liquid die forging defect detection method according to some embodiments of the present application. DETAILED DESCRIPTION
[0017] In order to better understand the technical solutions of the present application, the technical solutions of the present application will be described in detail below in combination with the drawings in the specification and specific embodiments.
[0018] Reference Figure 1 is an exemplary flowchart of a liquid die forging defect detection method according to some embodiments of the present application, which mainly includes the following steps: In step 101, an ultrasonic phased array probe is used to excite ultrasonic waves to the inside of a liquid die forging in a side vertical symmetry excitation mode, and a plurality of ultrasonic signals of the ultrasonic waves propagating in the liquid die forging are collected.
[0019] In some embodiments, referring to Figure 2Fig. 1 is a schematic flow chart of transmitting ultrasonic waves according to some embodiments of the present application, and the present application can be implemented by the following steps to transmit ultrasonic waves into a liquid die casting by a side vertical symmetry excitation mode through an ultrasonic phased array probe: In step 1011, a nominal thickness of the liquid die casting is acquired. In step 1012, a type of ultrasonic wave is determined based on the nominal thickness. In step 1013, the type of ultrasonic wave is transmitted into the liquid die casting by a side vertical symmetry excitation mode through the ultrasonic phased array probe.
[0020] It should be noted that the ultrasonic phased array probe in the present application is an ultrasonic phased array probe.
[0021] In specific implementation, the nominal thickness of the liquid die casting can be acquired by the following methods, i.e., the nominal thickness of the liquid die casting can be directly acquired from design drawings of the liquid die casting, and in other embodiments, the nominal thickness of the liquid die casting can also be directly acquired from process files or product specification of the liquid die casting, which is not limited herein, and the nominal thickness refers to a thickness of a finished product as specified in design of the liquid die casting.
[0022] In specific implementation, the type of ultrasonic wave can be determined based on the nominal thickness by the following methods, i.e., according to common knowledge in the field of ultrasonic detection, when a nominal thickness of a workpiece and a wavelength of an ultrasonic wave are in the same order of magnitude or smaller, a Lamb wave is selected for detection to excite a complete plate wave structure, and when the nominal thickness of the workpiece is much larger than the wavelength of the ultrasonic wave, a bulk wave is selected to effectively scan an internal volume, specifically, an empirical thickness threshold can be preset, for example, the thickness threshold is set to 2 mm in the present application, if the nominal thickness is less than or equal to the threshold, the type of ultrasonic wave is set to a Lamb wave, and if the nominal thickness is greater than the threshold, the type of ultrasonic wave is set to a bulk wave.
[0023] In specific implementation, the type of ultrasonic wave can be transmitted into the liquid die casting by a side vertical symmetry excitation mode through the ultrasonic phased array probe by the following methods, i.e., first, a material of the liquid die casting is acquired, and a sound velocity of the ultrasonic wave propagating in the material is queried, then, according to the sound velocity and a preset beam vertical incidence requirement, a transmission delay of all array elements of the linear phased array is set to zero or a same constant value, this delay rule synchronously excites all array elements, thereby synthesizing a plane wave parallel to a side surface of the probe array and perpendicular to a surface of the workpiece, and finally, according to the delay rule, each array element of the linear ultrasonic phased array is synchronously driven, and the specified type of ultrasonic wave is incident into the liquid die casting at a vertically symmetric beam angle through a sound coupling agent.
[0024] It should be noted that each probe in the phased array of the present application is an array element.
[0025] In a specific implementation, the multiple ultrasonic signals of the ultrasonic wave propagating in the liquid die forging can be acquired in the following manner: the multiple ultrasonic signals of the ultrasonic wave propagating in the liquid die forging are acquired in a full-matrix capture mode, that is, each array element in the array is sequentially excited as a transmitting source, and when the ultrasonic wave is transmitted by each array element, all array elements (including the transmitting array element itself) in the array are simultaneously used as receivers to capture and digitally record the corresponding complete ultrasonic signals, and the process is repeated until each array element in the array has completed the process of being the only transmitting source, thereby obtaining the multiple ultrasonic signals.
[0026] In step 102, the time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering is extracted based on the energy attenuation curve of each ultrasonic signal, and all ultrasonic signals are divided into direct waves in the initial stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all time points.
[0027] In some embodiments, the time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering can be extracted based on the energy attenuation curve of each ultrasonic signal in the following steps: determining the energy attenuation curve of each ultrasonic signal; determining the mutation point of the energy change rule of each ultrasonic signal according to the attenuation rate of each energy attenuation curve; determining the time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering based on all mutation points.
[0028] In a specific implementation, the energy attenuation curve of each ultrasonic signal can be determined in the following manner: first, Hilbert transform is performed on each ultrasonic signal to calculate the analytic signal of each ultrasonic signal, then the modulus, i.e., the absolute value, of each analytic signal is calculated, and finally, the modulus of each analytic signal is squared, and each curve obtained after squaring is used as the energy attenuation curve of each ultrasonic signal.
[0029] It should be noted that the energy attenuation curve in the present application is a curve used to describe the energy attenuation of the ultrasonic wave propagating in the liquid die forging. The energy of the ultrasonic wave is concentrated and attenuates quickly in the initial stage of propagation, and then enters a diffusion stable stage formed by multiple scattering, in which the energy slowly attenuates. The curve can effectively depict this transition process.
[0030] In a specific implementation, the determination of the mutation point of the energy variation law of each ultrasonic signal according to the attenuation rate of each energy attenuation curve can be implemented in the following manner: the derivative of each energy attenuation curve is calculated, and the obtained derivative is taken as the attenuation rate curve of each energy attenuation curve, respectively. Subsequently, the Hurst index of each attenuation rate curve in each sub-interval is determined, and the trend of the attenuation rate curve in each sub-interval is identified through the Hurst index. When the sub-interval of random walk is first identified, the starting time of the sub-interval is taken as the mutation point of the energy variation law of the corresponding ultrasonic signal. That is, for each attenuation rate curve, the Hurst index of the attenuation rate curve in each sub-interval with an interval length of 0.5 seconds can be calculated. Subsequently, the relative error between each Hurst index and 0.5 is calculated (as known in the industry, the closer the Hurst index is to 0.5, the closer the curve is to random walk). If the relative error is less than 10%, it is determined that it is random walk.
[0031] It should be noted that the mutation point in the present application refers to the critical time when the physical mechanism of ultrasonic wave propagation fundamentally changes. Before the mutation point, the ultrasonic energy is dominated by direct waves that follow a certain geometric path, and its attenuation process is controlled by the macroscopic geometric boundary of the workpiece. Therefore, its attenuation rate curve shows strong long-range correlation or trend (i.e., the Hurst index significantly deviates from 0.5). After the mutation point, the ultrasonic wave is scattered multiple times in the non-uniform microstructure of the liquid die forging, forming a diffusion field dominated by random scattered waves. Therefore, the attenuation rate curve shows the random walk feature of no memory. Therefore, the extraction of the mutation point is essentially to accurately capture the critical point of the physical transition of the ultrasonic wave from a deterministic propagation state to a random diffusion state by quantifying the trend change of the attenuation rate curve.
[0032] In a specific implementation, the determination of the time point at which the ultrasonic wave propagation mechanism in the liquid die forging changes from a deterministic path to a random scattering dominant for each ultrasonic wave emission based on all the mutation points can be implemented in the following manner: first, select one element in the ultrasonic phased array probe as a selected element, and obtain all the ultrasonic signals when the selected element emits ultrasonic waves. The average of the mutation points of all the ultrasonic signals when the selected element emits ultrasonic waves is taken as the time point at which the ultrasonic field in the liquid die forging is stable when the selected element emits ultrasonic waves. The time point at which the ultrasonic field in the liquid die forging is stable when the remaining elements in the ultrasonic phased array probe emit ultrasonic waves is determined.
[0033] It should be noted that the deterministic path dominance in the present application refers to the early stage after the ultrasonic wave is emitted, the energy propagation is mainly composed of wave components following a clear geometric path, at this time the propagation direction, arrival time and energy attenuation of the sound wave are mainly determined by the macroscopic shape, size and probe position of the workpiece and other deterministic factors, the propagation trajectory and energy attenuation mode are theoretically predictable, further, the random scattering dominance refers to the later stage of ultrasonic wave propagation, due to the non-uniform microstructure of the liquid die forging, the original wave component following the deterministic path has been significantly attenuated after experiencing sufficient scattering, the sound field gradually evolves into a diffusion field composed of scattered waves with random propagation direction and chaotic phase, at this stage, the propagation behavior of the sound wave is no longer dominated by the macroscopic geometric boundary, but by the statistical properties of the internal microstructure of the material, the sound field state and the signal show strong randomness and unpredictability.
[0034] In some embodiments, dividing all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete diffusion of the ultrasonic waves according to all time points can be achieved by the following steps: selecting an array element in the ultrasonic phased array probe as a selected array element, obtaining a time point at which the ultrasonic wave field in the liquid die forging is stable when the selected array element emits ultrasonic waves; dividing all the ultrasonic signals when the selected array element emits ultrasonic waves into direct waves in the early stage of ultrasonic propagation and scattered waves after complete diffusion of the ultrasonic waves through the time point at which the ultrasonic wave field in the liquid die forging is stable when the selected array element emits ultrasonic waves; continuing to divide all the ultrasonic signals when the remaining array elements in the ultrasonic phased array probe emit ultrasonic waves into direct waves in the early stage of ultrasonic propagation and scattered waves after complete diffusion of the ultrasonic waves through the time point at which the ultrasonic wave field in the liquid die forging is stable when the remaining array elements in the ultrasonic phased array probe emit ultrasonic waves.
[0035] In specific implementation, dividing all the ultrasonic signals when the selected array element emits ultrasonic waves into direct waves in the early stage of ultrasonic propagation and scattered waves after complete diffusion of the ultrasonic waves through the time point at which the ultrasonic wave field in the liquid die forging is stable when the selected array element emits ultrasonic waves can be achieved by the following manner, that is, dividing all the ultrasonic signals when the selected array element emits ultrasonic waves into two segments through the time point at which the ultrasonic wave field in the liquid die forging is stable when the selected array element emits ultrasonic waves, and taking the signal segment before the time point as the direct wave in the early stage of ultrasonic propagation and taking the signal segment after the time point as the scattered wave after complete diffusion of the ultrasonic waves.
[0036] It should be noted that the early stage of ultrasonic propagation in the present application refers to the stage in which the sound field is not fully diffused after the ultrasonic wave is emitted, and the wave following the deterministic geometric path is dominant, further, the direct wave refers to the wave component directly reaching the receiving array element from the transmitting array element, including the early reflection wave and the near-field interference component.
[0037] In addition, it needs to be explained that the complete diffusion of ultrasound in the present application refers to the state that the sound field reaches statistical uniformity after the ultrasound is scattered in the component for a sufficient number of times. Further, the scattered wave refers to a set of wave components with random propagation directions generated by the interaction between the material microstructure or small defects and the diffusion field, which contains the defect information inside the liquid die.
[0038] In step 103, the near-field effect of each scattered wave is suppressed by the spatial coherence between all the scattered waves, and a reconstructed signal matrix retaining the defect scattering characteristics is obtained. Then, the reconstructed signal matrix and all the direct waves are adaptively weighted and fused to obtain a mixed signal matrix of the fused early reflection information and the defect scattering characteristics.
[0039] In some embodiments, the step of suppressing the near-field effect of each scattered wave by the spatial coherence between all the scattered waves to obtain a reconstructed signal matrix retaining the defect scattering characteristics can be implemented by the following steps: combining all the scattered waves into a scattered wave signal matrix; performing cross-correlation processing on the signals in the scattered wave signal matrix to obtain the spatial coherence of any two columns of scattered wave signals at different time delays; suppressing the near-field effect along the common array element dimension by all the spatial coherence to obtain a reconstruction function estimate; reconstructing the reconstruction function estimate according to the array element index to obtain a reconstructed signal matrix retaining the defect scattering characteristics.
[0040] In a specific implementation, the step of combining all the scattered waves into a scattered wave signal matrix can be implemented in the following manner. First, each array element in the ultrasonic phased array is uniquely numbered. If there are n array elements in total, the array elements are numbered from 1 to n in the order of first row and then column, starting from the top left corner of the ultrasonic phased array. Then, the scattered wave signal corresponding to each specific transmitting array element i and receiving array element j is filled into a two-dimensional matrix with the same dimension as the ultrasonic phased array according to its original transmitting index i and receiving index j, thereby directly constructing a scattered wave signal matrix. The matrix element in the ith row and jth column is the scattered wave corresponding to the combination of transmitting array element i and receiving array element j.
[0041] It needs to be explained that the receiving array element in the present application refers to the array element in the ultrasonic phased array, which is only distinguished according to the working state of the array element. For example, the same array element is a transmitting array element when it transmits a signal, and is a receiving array element when it receives a signal.
[0042] It needs to be explained that the scattered wave signal matrix in the present application refers to a data set containing all the late echo signals formed by the multiple chaotic reflections of ultrasonic waves inside the workpiece.
[0043] In a specific implementation, the spatial coherence between any two columns of the scattered wave signal matrix can be obtained by cross-correlating the signals in the scattered wave signal matrix, which can be achieved by the following method: traversing the scattered wave signal matrix, and for any two receiving elements j and k receiving the scattered wave under the excitation of the same transmitting element i, the cross-correlation function between them is calculated, which represents the spatial coherence of the sound field between the receiving points j and k under the condition that the sound source i is fixed. Each spatial coherence can be uniquely determined by the index (i, j, k).
[0044] It should be noted that the spatial coherence in the present application is a function for describing the degree of similarity between the scattered wave signals received at different positions.
[0045] In a specific implementation, the near-field effect is suppressed by averaging all spatial coherences along the common element dimension to obtain the reconstruction function estimate, which can be achieved by the following method: first, select an element in the ultrasonic phased array as the selected element k, then average the spatial coherences with the same transmitting element index and the selected element index k to obtain the reconstruction function from each transmitting element to the selected element, and continue to obtain the reconstruction function from each transmitting element to the remaining elements in the ultrasonic phased array. Finally, the derivative of all reconstruction functions is obtained, and the set of results obtained by the derivative is used as the reconstruction function estimate. For example, if there are 5 elements, numbered 1 to 5, to determine the reconstruction function from element 3 to element 2, 5 spatial coherences are obtained by the index (3, 2, x), where x is a variable and the value range is the number of elements in the ultrasonic phased array. The function obtained by averaging the 5 spatial coherences and then taking the derivative is used as the reconstruction function from element 3 to element 2.
[0046] It should be noted that the reconstruction function estimate in the present application is a response function reflecting the true propagation path of sound waves between two points.
[0047] In a specific implementation, the reconstruction function estimate can be reconstructed into a reconstruction signal matrix that retains the scattering characteristics of the defect by the following method: all reconstruction functions in the reconstruction function estimate are arranged in a matrix, i.e., the reconstruction function from element e to element w is used as the element in the e-th row and w-th column of the matrix, and then the obtained matrix is used as the reconstruction signal matrix.
[0048] It should be noted that the reconstruction signal matrix in the present application is a data set composed of all reconstruction functions.
[0049] In some embodiments, the reconstruction signal matrix and all direct waves are adaptively weighted and fused to obtain a mixed signal matrix that combines early reflection information and defect scattering characteristics, which can be achieved by the following steps: determine a fusion weight function based on the time of arrival of the first bottom echo of the liquid forging; determine a compensation factor of the attenuation of ultrasonic wave propagation in the liquid forging according to all the direct arrivals and the reconstructed signal matrix; fuse all the direct arrivals and the reconstructed signal matrix into a mixed signal matrix of fused early reflection information and defect scattering characteristics through the fusion weight function and the compensation factor.
[0050] In a specific implementation, the fusion weight function based on the time of arrival of the first bottom echo of the liquid forging can be determined in the following manner: in the early stage of ultrasonic signal propagation, the direct arrivals contain clear defect reflection information and are less disturbed by noise, and are highly reliable; in the late stage of ultrasonic signal propagation, the signal-to-noise ratio of the direct arrivals decreases due to attenuation and mode conversion, and the signals reconstructed from the diffuse field can better extract defect scattering information submerged by noise. Therefore, the Sigmoid function can be selected to calculate the weight coefficient. Specifically, first, the time of arrival of the first bottom echo of the liquid forging is determined. The time of arrival of the first bottom echo of the liquid forging can be calculated according to the nominal thickness of the liquid forging and the propagation speed of the ultrasonic wave therein, and the theoretical time required for the ultrasonic wave to propagate one round-trip thickness distance, and then the time of arrival of the first bottom echo of the liquid forging is obtained. Then, the current time point and the time of arrival of the first bottom echo of the liquid forging are substituted into the Sigmoid function to output a fusion weight function continuously changing between 0 and 1, that is, the fusion weight function expressed as a(t) = 1 / (1+exp(-r(t-T))) is obtained, where a(t) is the fusion weight function, t is the time variable, T is the time of arrival of the first bottom echo of the liquid forging, r is a smoothing degree parameter greater than zero set by experience, used to control the transition time width from direct arrival dominance to reconstructed signal dominance, and exp is the natural exponential function.
[0051] It should be noted that the fusion weight function in the present application is a function composed of weight coefficients continuously changing with time, and its value range is [0, 1], which is used to dynamically adjust the proportion of direct arrival signals and reconstructed signals in the mixed signal in the fusion process.
[0052] In a specific implementation, the compensation factor of the attenuation of the ultrasonic wave in the liquid die forging can be determined according to all the direct waves and the reconstructed signal matrix in the following manner: first, the time at which the first echo from the bottom of the liquid die forging arrives is obtained, then, a time window is selected with the time as the center point, for example, assuming that the time is S, the time window can be selected as [S-5, S+5] with the unit of second, then, the sum U1 of the absolute amplitudes of all the direct waves in the time window is calculated, and the sum U2 of the absolute amplitudes of all the reconstructed signals in the time window is calculated, and finally, the ratio of U1 to U2 is taken as the compensation factor of the attenuation of the ultrasonic wave in the liquid die forging.
[0053] It should be noted that the compensation factor in the present application is a parameter value for compensating for the energy difference caused by the acoustic attenuation characteristics and the diffusion field reconstruction process when the ultrasonic wave propagates in the material of the liquid die forging.
[0054] In a specific implementation, the fusion of all the direct waves and the reconstructed signal matrix into a mixed signal matrix that fuses the early reflection information and the defect scattering characteristics can be realized in the following manner: first, all the direct waves are arranged into a matrix, the element in the qth row and the pth column of the matrix corresponds to the direct wave of the transmitting element q and the receiving element p, then, the compensation factor is multiplied with the reconstructed signal matrix, and the fusion is performed through the fusion weight function and the matrix of the direct waves, that is, according to the formula y=a*V+(1-a)*b*W, wherein y is the mixed signal matrix, a is the fusion weight function, V is the matrix of the direct waves, b is the compensation factor, and W is the reconstructed signal matrix.
[0055] It should be noted that the reconstructed signal matrix in the present application is a matrix composed of ultrasonic signals after suppressing noise interference in the near field range.
[0056] In step 104, the defects inside the liquid die forging are fully focused and imaged based on the echo characteristics of each signal in the mixed signal matrix, and the defects inside the liquid die forging are identified and positioned according to the imaging result.
[0057] In some embodiments, the defects inside the liquid die forging can be fully focused and imaged based on the echo characteristics of each signal in the mixed signal matrix in the following steps: The mixed signal matrix is subjected to nonlinear full focusing processing to obtain a preprocessed signal matrix; The imaging area is discretized into a pixel grid; The signal value corresponding to each pixel in the pixel grid is extracted from the preprocessed signal matrix based on the sound propagation delay corresponding to each pixel, and coherent superposition is performed, thereby generating a defect image inside the liquid die forging.
[0058] In a specific implementation, the nonlinear all-focus processing of the mixed signal matrix to obtain the preprocessed signal matrix can be implemented in the following manner: first, Hilbert transform is performed on each signal in the mixed signal matrix to calculate an analytic signal, then the envelope of each analytic signal is extracted, and the envelope is subjected to nonlinear scaling processing of p-th root to enhance the contrast while keeping the instantaneous phase of the signal unchanged, and finally all signals subjected to the nonlinear scaling processing are organized into a preprocessed signal matrix in the original matrix structure, where p is an arbitrary real number greater than 1. In other embodiments, other existing technologies can also be used to perform nonlinear all-focus processing on the mixed signal matrix, which is not limited here.
[0059] It should be noted that the preprocessed signal matrix in the present application is a mixed signal matrix after enhancing the contrast between the defect signal and the background noise.
[0060] In a specific implementation, the imaging area can be discretized into a pixel grid in the following manner: a two-dimensional rectangular imaging area is defined in a Cartesian coordinate system according to the arrangement position of the ultrasonic phased array probe and the detection area range of the liquid die, and the rectangular area is divided along the depth direction and the width direction with a fixed pixel spacing to form a uniformly distributed pixel grid, wherein the pixel spacing is set by considering the imaging resolution requirement and the calculation efficiency, and is usually set to be between one half of the wavelength of the ultrasonic wave in the medium and one wavelength.
[0061] In a specific implementation, the signal value corresponding to each pixel in the pixel grid is extracted from the preprocessed signal matrix based on the sound wave propagation delay corresponding to each pixel, and then the coherent superposition is performed to generate a defect image inside the liquid die in the following manner: for each pixel in the pixel grid, the total sound wave propagation delay from each transmitting element to the pixel and back to each receiving element is calculated according to the propagation speed of the ultrasonic wave in the liquid die, the signal amplitude at the corresponding time is extracted from the signal of the corresponding transmitting-receiving channel in the preprocessed signal matrix based on the propagation delay, the signal amplitudes of all transmitting-receiving channels are subjected to coherent superposition, the modulus of the superposition result is taken as the imaging intensity value of the pixel, and the above calculation is performed for all pixels in the pixel grid to finally generate a defect image reflecting the defect distribution inside the liquid die.
[0062] It should be noted that the defect image in the present application is an image for visually displaying the position, size and morphological characteristics of the defects inside the liquid die.
[0063] In a specific implementation, the identification and positioning of the defects inside the liquid die forging based on the imaging result can be achieved in the following manner: first, an intensity threshold based on the overall gray scale statistical features of the defect image is set, all pixel points in the imaging result with an intensity value higher than the threshold are identified as potential defect regions, then connected domain analysis is performed on adjacent potential defect pixel points to determine independent defect regions, and finally the centroid coordinates of each defect region are calculated as the defect position, and the maximum size and average intensity thereof are recorded as the defect feature parameters, thereby completing the identification and positioning of the defects inside the liquid die forging. For example, the average value of the gray scale values of the defect image can be used as the intensity threshold.
[0064] In some embodiments, the reference Figure 3 The horizontal axis x (unit: millimeter, mm) corresponds to the width direction of the liquid die forging, covering the transverse scanning range detected by the ultrasonic phased array probe, and the numerical range is matched with the array element arrangement length and the detection coverage area of the probe. The vertical axis z (unit: millimeter, mm) corresponds to the thickness (depth) direction of the liquid die forging, extending from the surface of the forging (z=0 mm) to the inside, and the upper limit of the numerical value is consistent with the nominal thickness of the liquid die forging, which is used to locate the depth position of the defect inside the forging. The image is a defect image obtained by full-focus imaging of a plate-shaped liquid die forging sample in a laboratory using the method of the present application. The plate-shaped liquid die forging sample has a size of 300 mm x 1 mm x 150 mm. Two defect regions are clearly visible in the image, and the near-field effect and clutter interference are weak.
[0065] In addition, another aspect of the present application provides a liquid die forging defect detection system in some embodiments, referring to Figure 4 The figure is a structural schematic diagram of a liquid die forging defect detection system according to some embodiments of the present application. The liquid die forging defect detection system 400 includes a collection module 401, a processing module 402, and an execution module 403, which are described as follows: The collection module 401 is mainly used for transmitting ultrasonic waves to the inside of the liquid die forging by the ultrasonic phased array probe in a side vertical symmetry excitation mode, and collecting a plurality of ultrasonic signals propagating in the liquid die forging. The processing module 402 is mainly used for extracting the time point at which the ultrasonic wave propagation mechanism inside the liquid die forging changes from deterministic path dominance to random scattering dominance based on the energy attenuation curve of each ultrasonic signal, and dividing all the ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points. It should be noted that the processing module 402 is further configured to perform near-field effect suppression on each scattered wave by spatial coherence between all scattered waves, so as to obtain a reconstructed signal matrix retaining defect scattering characteristics, and to perform adaptive weighted fusion on the reconstructed signal matrix and all direct waves, so as to obtain a mixed signal matrix of fused early reflection information and defect scattering characteristics. The execution module 403 is mainly configured to perform full-focus imaging on the defects in the liquid die forging based on echo characteristics of each signal in the mixed signal matrix, and to identify and locate the defects in the liquid die forging according to the imaging result.
[0066] In addition, the present application further provides a computer device, which comprises a memory and a processor, the memory stores code, and the processor is configured to acquire the code and execute the liquid die forging defect detection method described above.
[0067] In some embodiments, reference Figure 5 The figure is a structural schematic diagram of a computer device for implementing the liquid die forging defect detection method according to some embodiments of the present application. The liquid die forging defect detection method in the above embodiments can be implemented by the computer device shown in the figure, which is a computer device for implementing the liquid die forging defect detection method. Figure 5 The computer device 500 comprises at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.
[0068] The processor 501 can be a general central processing unit (CPU) or an application-specific integrated circuit (ASIC).
[0069] The communication bus 502 can be used to transmit information between the above components.
[0070] The memory 503 can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM), or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk or other magnetic storage device, or any other medium capable of storing instructions or data that can be accessed by a computer, without limitation. The memory 503 can exist independently of the processor 501, and can be connected to the processor 501 via the communication bus 502. The memory 503 can also be integrated with the processor 501.
[0071] The memory 503 is configured to store program codes of the embodiments of the present application, and the processor 501 is configured to execute the program codes stored in the memory 503. The program codes can include one or more software modules. The liquid die forging defect detection method in the above embodiments can be implemented by one or more software modules in the program codes of the processor 501 and the memory 503.
[0072] The communication interface 504 is configured to communicate with other devices or communication networks, such as an Ethernet, a radio access network (RAN), a wireless local area network (WLAN), and the like, using any transceiver-like mechanism.
[0073] In specific implementations, as an example, the computer device can include multiple processors, each of which can be a single-CPU processor or a multi-CPU processor. The processor herein can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).
[0074] The computer device described above can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a laptop computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of computer device.
[0075] In addition, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the liquid die forging defect detection method described above.
[0076] In summary, in the liquid die forging defect detection method and system disclosed by the embodiments of the present application, first, an ultrasonic phased array probe transmits ultrasonic waves to the inside of a liquid die forging according to a side vertical symmetry excitation mode, and collects a plurality of ultrasonic signals propagating in the liquid die forging; based on the energy attenuation curve of each ultrasonic signal, a time point is extracted at which the ultrasonic wave propagation mechanism in the liquid die forging changes from being dominated by a deterministic path to being dominated by random scattering, and all ultrasonic signals are divided into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all time points; near-field effect suppression is performed on each scattered wave through the spatial coherence between all scattered waves, and then a reconstructed signal matrix retaining defect scattering characteristics is obtained, and adaptive weighted fusion is performed on the reconstructed signal matrix and all direct waves to obtain a mixed signal matrix combining early reflection information and defect scattering characteristics; based on the echo characteristics of each signal in the mixed signal matrix, full-focus imaging is performed on the defects inside the liquid die forging, and the defects inside the liquid die forging are identified and positioned according to the imaging result.
[0077] As can be seen, the present application extracts the time point of the change of the ultrasonic wave propagation mechanism based on the energy attenuation curve, accurately divides the ultrasonic signals into direct waves carrying early reflection information and scattered waves containing diffusion field defect characteristics, realizes effective separation of signal components, and then uses the spatial coherence between all scattered waves to perform near-field effect suppression. This method does not depend on the assumption of a homogeneous medium, can specifically distinguish and suppress the near-field interference generated by the probe, while retaining the structural scattering signals generated by the material micro-heterogeneity and possibly containing defect information, thereby avoiding signal loss caused by traditional filtering methods. Thereafter, the processed scattered wave reconstructed signal and the direct wave are complementary to each other through adaptive weighted fusion to generate a high-quality mixed signal matrix, and finally high-resolution and high-contrast imaging of the defects is realized through full-focus imaging technology, and defect identification is performed based on the imaging result. In summary, by using the scheme of the present application, the near-field effect of the ultrasonic wave in non-destructive defect detection can be suppressed in the case of uneven material of the liquid die forging.
[0078] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those skilled in the art once they learn of the basic inventive concepts. Therefore, the appended claims are intended to encompass within their scope all such variations and modifications as are included within the scope of the application.
[0079] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.
Claims
1. A method of liquid die casting defect detection, characterized by, The method comprises the following steps: ultrasonic waves are emitted into the liquid-state die forging by an ultrasonic phased array probe in a side face vertical symmetry excitation mode, and a plurality of ultrasonic signals of ultrasonic wave propagation in the liquid-state die forging are collected; a time point at which ultrasonic wave propagation mechanism in the liquid-state die forging changes from being dominated by a deterministic path to being dominated by random scattering is extracted based on an energy attenuation curve of each ultrasonic signal, and all the ultrasonic signals are divided into direct waves in an early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points; near-field effect suppression is performed on each scattered wave through spatial coherence between all the scattered waves, a reconstructed signal matrix retaining defect scattering characteristics is obtained, and a mixed signal matrix of fused early reflection information and defect scattering characteristics is obtained through adaptive weighted fusion of the reconstructed signal matrix and all the direct waves; defects in the liquid-state die forging are fully focused and imaged based on echo characteristics of signals in the mixed signal matrix, and the defects in the liquid-state die forging are identified and positioned according to the imaging result.
2. The method of claim 1, wherein, The method of emitting ultrasonic waves into the liquid-state die forging by the ultrasonic phased array probe in the side face vertical symmetry excitation mode specifically comprises the following steps: a nominal thickness of the liquid-state die forging is obtained; a type of ultrasonic wave is determined based on the nominal thickness; the type of ultrasonic wave is emitted into the liquid-state die forging by the ultrasonic phased array probe in the side face vertical symmetry excitation mode.
3. The method of claim 1, wherein, The method of extracting a time point at which ultrasonic wave propagation mechanism in the liquid-state die forging changes from being dominated by a deterministic path to being dominated by random scattering based on an energy attenuation curve of each ultrasonic signal specifically comprises the following steps: an energy attenuation curve of each ultrasonic signal is determined; a mutation point of energy variation law of each ultrasonic signal is determined according to an attenuation rate of each energy attenuation curve; a time point at which ultrasonic wave propagation mechanism in the liquid-state die forging changes from being dominated by a deterministic path to being dominated by random scattering is determined based on all the mutation points.
4. The method of claim 1, wherein, The method of dividing all the ultrasonic signals into direct waves in an early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all the time points specifically comprises the following steps: one element in the ultrasonic phased array probe is selected as a selected element, and a time point at which an ultrasonic wave field in the liquid-state die forging is stable when the selected element emits ultrasonic waves is obtained; all the ultrasonic signals when the selected element emits ultrasonic waves are divided into direct waves in an early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion through the time point at which the ultrasonic wave field in the liquid-state die forging is stable when the selected element emits ultrasonic waves; the method of dividing all the ultrasonic signals when the remaining elements in the ultrasonic phased array probe emit ultrasonic waves into direct waves in an early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion is continued through the time point at which the ultrasonic wave field in the liquid-state die forging is stable when the remaining elements emit ultrasonic waves.
5. The method of claim 1, wherein, The method of performing near-field effect suppression on each scattered wave through spatial coherence between all the scattered waves, and then obtaining a reconstructed signal matrix retaining defect scattering characteristics specifically comprises the following steps: all the scattered waves are combined into a scattered wave signal matrix; signals in the scattered wave signal matrix are cross-correlated to obtain spatial coherence of any two columns of scattered wave signals under different time delays; The near-field effect is suppressed along a common array element dimension by all spatial coherence to obtain a reconstruction function estimate; The reconstruction function estimate is reconstructed by array element index into a reconstruction signal matrix that retains defect scattering characteristics.
6. The method of claim 1, wherein, The reconstruction signal matrix and all direct waves are adaptively weighted and fused to obtain a mixed signal matrix that fuses early reflection information and defect scattering characteristics, specifically including: A fusion weight function is determined based on the time when the first bottom echo of the liquid die forging arrives; A compensation factor for ultrasonic wave propagation attenuation in the liquid die forging is determined according to all direct waves and the reconstruction signal matrix; All direct waves and the reconstruction signal matrix are fused into a mixed signal matrix that fuses early reflection information and defect scattering characteristics by the fusion weight function and the compensation factor.
7. The method of claim 1, wherein, Based on the echo characteristics of each signal in the mixed signal matrix, the defects inside the liquid die forging are fully focused and imaged, specifically including: The mixed signal matrix is nonlinearly fully focused to obtain a preprocessed signal matrix; The imaging area is discretized into a pixel grid; Based on the corresponding acoustic propagation delay of each pixel in the pixel grid, the corresponding signal value is extracted from the preprocessed signal matrix and coherently superimposed, thereby generating a defect image inside the liquid die forging.
8. The method of claim 1, wherein, A plurality of ultrasonic signals propagating in the liquid die forging are collected in a full matrix capture mode.
9. The method of claim 1, wherein, The ultrasonic phased array probe is a linear ultrasonic phased array probe.
10. A liquid die casting defect detection system, characterized by, It includes: A collection module for transmitting ultrasonic waves into the liquid die forging by the ultrasonic phased array probe in a side vertical symmetry excitation mode and collecting a plurality of ultrasonic signals propagating in the liquid die forging; A processing module for extracting a time point when the ultrasonic wave propagation mechanism in the liquid die forging changes from deterministic path dominance to random scattering dominance based on the energy attenuation curve of each ultrasonic signal for each ultrasonic wave transmission, and dividing all ultrasonic signals into direct waves in the early stage of ultrasonic propagation and scattered waves after complete ultrasonic diffusion according to all time points; The processing module is also used to suppress the near-field effect of each scattered wave by the spatial coherence between all scattered waves, thereby obtaining a reconstruction signal matrix that retains defect scattering characteristics, and adaptively weighting and fusing the reconstruction signal matrix and all direct waves to obtain a mixed signal matrix that fuses early reflection information and defect scattering characteristics; An execution module for fully focusing and imaging the defects inside the liquid die forging based on the echo characteristics of each signal in the mixed signal matrix, and identifying and positioning the defects inside the liquid die forging according to the imaging result.
