An ultrasonic non-destructive testing method based on full matrix data migration and phase coherence
By employing a frequency domain phase shift transfer algorithm and a phase coherence factor weighted full-focus imaging method, the problems of low imaging efficiency and insufficient accuracy in multi-layer structures in traditional ultrasonic testing are solved, enabling rapid and high-resolution damage detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- EAST CHINA UNIV OF SCI & TECH
- Filing Date
- 2026-05-12
- Publication Date
- 2026-06-30
Smart Images

Figure CN122306949A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of nondestructive testing technology, and more specifically, to an ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence. Background Technology
[0002] Multi-layered industrial structures (such as multi-walled pressure vessels, weld overlay plates, composite plates, and cladding structures) are widely used in petrochemical, energy, and metallurgical industries. Their structural safety directly affects the long-term stable operation of the equipment and the safety of personnel and the environment. However, during long-term service, these structures often simultaneously endure high-temperature and high-pressure conditions, corrosive media erosion, and fatigue caused by alternating loads. Poor bonding, interlayer debonding, and localized stress concentration may also occur at multi-layer interfaces, inducing various damage types such as cracks, corrosion thinning, inclusions, and delamination / debonding. Because damage usually originates in concealed locations and expands gradually, failure to detect and implement timely repair or replacement measures can easily lead to major accidents such as media leaks, explosions, and production shutdowns, causing severe loss of life and property and environmental risks.
[0003] Currently, non-destructive testing (NDT) technologies for complex industrial structures mainly include X-ray inspection, magnetic particle inspection, penetrant testing, eddy current testing, and ultrasonic testing. X-ray inspection has certain advantages for volumetric defects, but it suffers from drawbacks such as high radiation safety management requirements, limited on-site conditions, and significantly affected testing efficiency and cost by operating conditions. Magnetic particle inspection and penetrant testing are mainly suitable for surface or near-surface defects, with limited applicability to internal and interface defects in multilayer structures. Eddy current testing is suitable for near-surface areas of conductive materials but is easily affected by the material's electromagnetic properties and surface condition. In contrast, ultrasonic testing has advantages such as strong penetration capability, sensitivity to internal delamination and cracks, quantitative localization, portable equipment, and moderate cost, making it one of the important methods for condition assessment and in-service inspection of industrial equipment.
[0004] The introduction of ultrasonic phased array technology provides a new technical means for damage detection of multi-layered planar structures in industrial applications. Ultrasonic phased arrays, through electronic delay control, can achieve beam deflection, dynamic focusing, and multi-angle scanning, theoretically improving detection coverage and resolution without significantly increasing mechanical scanning complexity. Furthermore, to enhance phased array imaging quality, researchers have introduced Full Matrix Capture (FMC) and Total Focusing Method (TFM) into phased array detection. By acquiring full matrix data and performing time-delayed superposition in the imaging domain, high focusing quality and imaging resolution can be achieved in homogeneous media or simple structures, thereby improving the accuracy and reliability of defect detection. However, for planar multilayer structures, traditional ultrasonic phased array detection methods are easily limited by the following factors: First, the multilayer interface causes multiple reflections, transmissions, and mode conversions of the sound beam during propagation, resulting in complex echo components. The superposition of defect echoes and interface echoes can easily obscure defect features. Second, differences in acoustic parameters of multilayer materials, interface roughness, and the non-uniformity of the composite layer structure introduce significant scattering and attenuation, leading to a reduced signal-to-noise ratio and increased imaging artifacts. Third, if geometric acoustics-based propagation path tracing (refraction angle calculation, layer travel time calculation, etc.) is used for imaging, the computational load is large and sensitive to structural parameters, making it difficult to meet the dual requirements of "rapid detection and high-precision positioning" in engineering applications. Fourth, although some imaging methods based on frequency domain signal extrapolation can achieve imaging of multilayer structures, there is a lack of an integrated solution that organically combines "rapid wavefield extrapolation" and "high-resolution noise-suppressed focusing imaging," often resulting in a contradiction between "speed and quality" in engineering applications. Therefore, traditional methods often struggle to achieve rapid, high-resolution, and high-reliability imaging of defects within the target layer of multilayer structures.
[0005] Therefore, there is an urgent need for an ultrasonic nondestructive testing method for planar multilayer structures: under the premise of ensuring engineering feasibility, it should be able to realize rapid extrapolation or migration of wavefield based on full matrix data, reduce the dependence on point-by-point solution of time-domain refraction path, and further combine coherence measures such as phase coherence factor to weight and enhance the full-focus imaging results, so as to achieve rapid, accurate localization and high-resolution imaging of internal damage defects in complex multilayer structures, thereby meeting the actual needs of multilayer wall inspection and maintenance of equipment such as pressure vessels. Summary of the Invention
[0006] To overcome the aforementioned shortcomings of existing technologies, this invention provides an ultrasonic nondestructive testing method based on full-matrix data transfer and phase coherence. By using a frequency-domain phase-shift transfer algorithm to extrapolate the wavefield layer by layer from the full-matrix capture dataset, the point-by-point solution of the refraction path at multi-layer interfaces in the time domain is avoided, effectively improving processing efficiency. Combined with a phase coherence factor-weighted full-focusing imaging algorithm, the method enhances the echo of real defects and suppresses noise and artifacts, solving the problems of low computational efficiency and poor imaging signal-to-noise ratio in traditional detection methods for planar multi-layer structures.
[0007] To achieve the above objectives, the present invention provides the following technical solution:
[0008] An ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence includes the following steps:
[0009] Step 1: Select the ultrasonic frequency according to the object to be detected, determine the parameter configuration of the phased array elements, arrange the ultrasonic phased array volume wave probe on the surface of the multi-layer structure of the plane to be detected, excite the ultrasonic volume wave signal, collect the echo signal between each array element, and construct the time-domain full matrix capture dataset.
[0010] Step 2: Perform a Fast Fourier Transform on the time-domain full matrix capture dataset to obtain a frequency-wavenumber domain full matrix capture dataset;
[0011] Step 3: Based on the frequency domain phase shift migration algorithm, according to the thickness and sound velocity of each layer of medium in the planar multilayer structure, the frequency-wavenumber domain full matrix capture dataset is extrapolated in the imaging depth direction to obtain the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer.
[0012] Step 4: Based on the explosion reflection model, perform a fast inverse Fourier transform on the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer to extract the time domain signal data corresponding to the target detection layer.
[0013] Step 5: Perform delay compensation processing on each imaging pixel based on the time-domain signal data, and calculate the phase coherence factor at each imaging pixel.
[0014] Step six: Using a phase coherence factor-weighted full-focusing imaging algorithm, the compensated data is reconstructed by time-delay weighting to obtain the imaging results of internal damage and defects of multi-layer structural target detection layers below any imaging depth.
[0015] As a further aspect of the present invention, the parameter configuration of the phased array elements in step one includes the transducer type, center frequency, and number of elements.
[0016] As a further aspect of the present invention, the time-domain full matrix capture dataset is obtained by sequentially exciting each element in the ultrasonic phased array probe and simultaneously recording the echo signals received by all elements, forming a complete data matrix containing the combined propagation information of all transmitting and receiving elements.
[0017] As a further aspect of the present invention, the fast Fourier transform process in step two specifically involves: capturing the acquired time-domain full matrix dataset. Perform a three-dimensional Fourier transform to obtain the frequency-wavenumber domain full matrix capture dataset at the imaging depth of 0. The transformed variables correspond Dimension of location correspond Dimension of location correspond Dimension in which it is located.
[0018] As a further aspect of the present invention, the frequency-domain phase-shift migration algorithm in step three applies an extrapolation factor to the full matrix capture dataset in the frequency-wavenumber domain to achieve layer-by-layer extrapolation of the wavefield along the depth direction, thereby obtaining the frequency-wavenumber domain wavefield data at the interface of the target detection layer; for example, if the thickness of the first layer of medium is known to be... The speed of sound is Capture the entire frequency-wavenumber domain dataset at the imaging depth of 0. Multiply by extrapolation factor to obtain depth Frequency-wavenumber domain full matrix capture dataset For a multi-layer structure, the frequency-wavenumber domain full matrix capture dataset at the interface of the nth layer of the target detection layer can be represented as: The result can be obtained by calculating using the method described above.
[0019] As a further aspect of the present invention, in step four, the explosion reflection model treats the scattering points in the target detection layer as virtual emission sources, performs time inversion processing on the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer, extracts the virtual source signal corresponding to the target detection layer, and constructs the corresponding time-domain dataset. That is, a three-dimensional inverse Fourier transform is performed on the frequency-wavenumber domain full matrix capture dataset at the interface of the nth layer of the target layer to obtain the time-domain full matrix dataset of the virtual array elements at the interface of the target layer.
[0020] .
[0021] As a further aspect of the present invention, the method for calculating the phase coherence factor in step five is as follows: analytical signal processing is performed on the echo signals corresponding to each transmit-receive array element combination in the time-domain signal data to extract the instantaneous phase information at each moment. At each imaging pixel, the instantaneous phase information of all array element signals corresponding to the propagation path through that point is statistically analyzed. A phase consistency index is constructed to characterize the coherence between signals, and this index is used as a weighting coefficient in the imaging calculation; [The remaining text appears to be incomplete and requires further context.] Each element emits and the first The echo signal received by each array element Corresponding single-signal phase coherence factor The calculation formula is:
[0022] ,
[0023] In the formula, For the first Each element emits and the first The time-domain echo signal received by each array element For Hilbert transform operations, To perform the operation of taking the real part, To perform operations on the imaginary part, For arctangent operation, For standard deviation calculation, To adjust the parameters, A preset phase threshold is used; the imaging grid points are obtained by delaying and summing in a manner similar to that of a total focusing algorithm. Phase coherence results at the point :
[0024] ,
[0025] In the formula, This is the delay amount for virtual focusing. The number of array elements of the array transducer; the phase coherence factor is used to enhance the coherent echo signal from the actual defect location, while suppressing random scattering noise and artifact signals.
[0026] As a further aspect of the present invention, in step six, the all-focusing imaging algorithm calculates the propagation time from each transmitting and receiving array element to the imaging pixel, performs delay compensation on the corresponding echo signals, and performs superposition processing; during the delay summation and reconstruction, a phase coherence factor is introduced to weight the echo signals of each array element, and the grid points in the imaging region are... The pixel amplitude value after introducing the phase coherence factor is:
[0027] ,
[0028] In the formula, After introducing the phase coherence factor, the grid points The amplitude value of the pixel at that location. Grid points obtained by the full-focus imaging algorithm The pixel amplitude value at the location; after normalizing the image reconstruction result, logarithmic compression is performed, and the image is thresholded to obtain the imaging result used to identify defects.
[0029] As a further embodiment of the present invention, the planar multi-layer structure includes a composite layer structure, a weld overlay structure, or a multi-layer metal structure in the pressure vessel wall. The detection method is applicable to internal damage detection and condition assessment of pressure vessels, storage tanks, reactors, and other equipment with multi-layer wall structures.
[0030] Compared with existing technologies, the beneficial effects of the ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence of this invention are as follows:
[0031] This invention employs a frequency-domain phase-shift migration algorithm to extrapolate the wavefield of the full-matrix capture dataset, transferring the reconstruction of ultrasonic wave propagation in a multi-layered planar structure to the frequency-wavenumber domain. This avoids the tedious calculations required by traditional methods to solve the time-domain refraction path point by point, effectively improving data processing efficiency while ensuring detection accuracy.
[0032] This invention introduces a phase coherence factor for weighting during the time-delay superposition process of the full-focus imaging algorithm. By taking advantage of the high phase consistency of the real defect echo signal and the strong phase dispersion of the noise and artifact signals, the effective signal at the defect location is enhanced and random scattering noise is suppressed. Compared with the traditional full-focus imaging method, it can achieve a higher signal-to-noise ratio and spatial resolution.
[0033] This invention organically combines rapid extrapolation of frequency domain wavefield with phase coherence factor weighted full-focus imaging to form a complete planar multilayer structure ultrasonic non-destructive testing scheme. It can simultaneously take into account detection speed and imaging quality, and is suitable for internal damage detection and condition assessment of industrial equipment with multilayer wall structures such as coke towers, reactors and high-pressure heat exchangers. Attached Figure Description
[0034] Figure 1 This is a schematic diagram of the process of an ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to the present invention.
[0035] Figure 2 This is a schematic diagram of the overall framework of an ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to the present invention.
[0036] Figure 3 This is a schematic diagram of the bulk wave phased array setup and planar multilayer structure experimental specimen for an ultrasonic nondestructive testing method based on full matrix data migration and phase coherence according to the present invention.
[0037] Figure 4 This is a schematic diagram illustrating the wavefield extrapolation principle of an ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to the present invention.
[0038] Figure 5 This is an experimental imaging result of a planar multilayer structure target layer based on an ultrasonic nondestructive testing method using full matrix data transfer and phase coherence, according to the present invention. Detailed Implementation
[0039] The technical solutions of this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.
[0040] Example 1
[0041] This embodiment provides an ultrasonic nondestructive testing method based on full matrix data migration and phase coherence to achieve localization imaging of internal defects in planar multilayer structures.
[0042] The wavelength of the ultrasonic bulk wave is calculated using known frequencies and sound speeds. A suitable frequency is selected based on the defect size to avoid missed defects due to diffraction. Figure 3 As shown, after applying an appropriate amount of coupling agent, the ultrasonic phased array volume wave probe is placed on the surface of the structure to be tested in order to detect defects in the effective detection area below the probe.
[0043] This method first requires capturing the acquired time-domain full matrix dataset. Perform a three-dimensional Fourier transform to obtain the frequency-wavenumber domain full matrix capture dataset at the imaging depth of 0. The specific calculation method can be considered as performing a Fast Fourier Transform on each of the three dimensions of a matrix sequentially, i.e., a three-dimensional Fourier Transform:
[0044] ,
[0045] Transformed variables correspond Dimension of location correspond Dimension of location correspond Dimension in which it is located.
[0046] After obtaining the full matrix capture dataset in the frequency-wavenumber domain, extrapolation of the frequency-domain wavefield is performed. Taking the extrapolation of the first-layer structure as an example: the thickness of the first-layer medium is known to be... The speed of sound is Capture the entire frequency-wavenumber domain dataset at the imaging depth of 0. Multiply by extrapolation factor to obtain depth Frequency-wavenumber domain full matrix capture dataset .like Figure 4 As shown, this scheme requires obtaining the frequency-wavenumber domain wavefield of the virtual array element at the interface of the nth layer of the target detection layer. The calculation method is as follows: ,
[0047] In the formula, For the first The thickness of the medium layer, For the first The speed of sound in the medium layer.
[0048] Then, a three-dimensional inverse Fourier transform is performed on the frequency-wavenumber domain wavefield to obtain the time-domain full matrix capture dataset of the virtual array elements at the interface of the nth layer of the target detection layer:
[0049] .
[0050] After extracting the temporal full-matrix capture dataset corresponding to the target detection layer, the phase coherence factor is calculated. The phase coherence factor only applies phase coherence weighting to a single focal point. However, after introducing virtual focusing in the full-focus imaging algorithm, phase processing needs to be performed on each echo signal in the temporal full-matrix capture dataset. (Temporal full-matrix capture dataset) In the middle, by the first The array element emitted and the first The time-domain echo signal received by each array element is denoted as For this signal, the single-signal phase coherence factor It can be obtained from the following formula:
[0051] ,
[0052] In the formula, For the first Each element emits and the first The time-domain echo signal received by each array element For Hilbert transform operations, To perform the operation of taking the real part, To perform operations on the imaginary part, For arctangent operation, For standard deviation calculation, To adjust the parameters, A preset phase threshold is used. The characteristic value of the phase distribution during the phase coherence factor calculation process is represented as a time-domain echo signal. The standard deviation of the instantaneous phase at each moment. The imaging grid points are obtained using a delayed summation method similar to that used in all-focusing imaging algorithms. Phase coherence factor at the location :
[0053] ,
[0054] In the formula, This is the delay amount for virtual focusing. The number of array elements of the array transducer.
[0055] Total Focusing Imaging (TFM) based on the Full Matrix Acquisition (FMC) dataset can virtually focus on every point in the imaging region by phase-shifting and superimposing the array signals from the FMC dataset. The FMC dataset, acquired using the full matrix acquisition method, is considered to contain complete geometric features and defect information of the target object. For a total number of array elements... The acquisition process of a one-dimensional linear array transducer with a full matrix acquisition dataset can be described as follows: the first array element is independently excited, and all array elements, including the excited element, receive and store the echo signal within the pulse repetition period, thus obtaining... Group echo data Then, the next element is instructed to fire, and the operation is repeated, and so on, until all elements are excited. The echo signals were collected into a dataset matrix, thus obtaining a complete dataset containing the transducer transmit and receive combinations.
[0056] Assume the total number of array elements is With a grid in the imaging region For example, the method for calculating the flight time of a given transmit-receive array element combination for point P is as follows:
[0057] ,
[0058] In the formula, c is the speed of sound in the medium. and These represent the lateral positions of the transmitting and receiving array elements, respectively. The sound wave propagation from each transmitting element to the virtual focal point is calculated. Then, by considering the acoustic path difference between each receiving array element and performing phase-shifting superposition processing on all signals in the array dataset, the signal can be obtained at the point. This creates an equivalent focusing enhancement effect. By extracting the corresponding amplitude of the virtual focusing beam through Hilbert transform, the pixel value representing the scattering information at that point can be obtained. The pixel amplitude value can be expressed as:
[0059] ,
[0060] In the formula, For the launch array element Transmitting and receiving array elements The received time-domain echo signal, and These represent the serial numbers of the transmitting and receiving array elements, respectively. The above operation is repeated for all grid pixels within the imaging region. After obtaining the amplitude information of all pixels in the entire imaging detection area, the image of the entire target region can be reconstructed.
[0061] For a grid in the imaging region After introducing the phase coherence factor (PCF) to perform phase coherence weighting on a single focal point, the focal point... The pixel amplitude value can be expressed as:
[0062] ,
[0063] Finally, the imaging program normalizes the image reconstruction results, performs logarithmic compression, and then thresholds the image to obtain the imaging results used for defect identification. Figure 3 The detection results of the double-layer sample obtained after imaging the experimental sample are shown in the figure. Figure 5 As shown in the image, areas with an amplitude greater than -6dB in the imaging results are generally considered defects.
[0064] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0065] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An ultrasonic non-destructive testing method based on full matrix data migration and phase coherence, characterized in that, Includes the following steps: Step 1: Use an ultrasonic phased array volume wave probe to excite and receive the planar multilayer structure to be tested, collect the echo signals between each array element, and construct a time-domain full matrix capture dataset. Step 2: The time-domain full matrix capture dataset is subjected to a fast Fourier transform to obtain a frequency-wavenumber domain full matrix capture dataset; Step 3: Based on the frequency domain phase shift migration algorithm, extrapolate the frequency-wavenumber domain full matrix capture dataset according to the thickness and sound velocity of each layer of the medium in the planar multilayer structure to obtain the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer. Step 4: Based on the explosion reflection model, perform a fast inverse Fourier transform on the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer to extract the time domain signal data corresponding to the target detection layer. Step 5: Perform delay compensation processing on the time-domain signal data and calculate the phase coherence factor at each imaging pixel. Step six: Based on the phase coherence factor weighted full-focusing imaging algorithm, the compensated data is reconstructed by time-delay weighting to obtain the imaging and localization of the damage and defects inside the target detection layer in the planar multilayer structure.
2. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The time-domain full matrix capture dataset is obtained by sequentially exciting each element in the ultrasonic phased array probe and simultaneously recording the echo signals received by all elements, forming a complete data matrix containing the propagation information of all transmitting and receiving elements.
3. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The frequency-domain phase-shift migration algorithm applies an extrapolation factor to the full matrix capture dataset in the frequency-wavenumber domain in the frequency-wavenumber domain, thereby achieving layer-by-layer extrapolation of the wave field along the depth direction and obtaining the frequency-wavenumber domain wave field data at the interface of the target detection layer.
4. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The explosion reflection model treats the scattering points in the target detection layer as virtual emission sources, performs time inversion processing on the frequency-wavenumber domain full matrix capture dataset at the interface of the target detection layer, extracts the virtual source signal corresponding to the target detection layer, and constructs the corresponding time domain dataset.
5. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The total focusing imaging algorithm calculates the propagation time from each transmitting and receiving array element to the imaging pixel, performs delay compensation on the corresponding echo signals, and then performs superposition processing.
6. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The method for calculating the phase coherence factor is as follows: the echo signals corresponding to each transmit-receive array element combination in the time domain signal data are analyzed and processed to extract the instantaneous phase information at each moment. At each imaging pixel, the instantaneous phase information of all array element signals corresponding to the propagation path through that point is statistically analyzed. The phase consistency index is constructed to characterize the coherence between signals, and this index is used as a weighting coefficient in the imaging calculation.
7. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 6, characterized in that, By the Each element emits and the first The echo signal received by each array element Corresponding single-signal phase coherence factor The calculation formula is: , In the formula, For the first Each element emits and the first The time-domain echo signal received by each array element For Hilbert transform operations, To perform the operation of taking the real part, To perform operations involving the imaginary part, For arctangent operation, For standard deviation calculation, To adjust the parameters, This is the preset phase threshold.
8. The ultrasonic nondestructive testing method based on full matrix data transfer and phase coherence according to claim 1, characterized in that, The planar multi-layer structure includes a composite layer structure, a weld overlay structure, or a multi-layer metal structure in the pressure vessel wall. The detection method is applicable to internal damage detection and condition assessment of pressure vessels, storage tanks, reactors, and other equipment with multi-layer wall structures.