Dynamic sparse ultrasonic full focusing method for large-wall-thickness structural member

Abstract

The invention relates to a dynamic sparse ultrasonic full-focusing method for a large-wall-thickness structural member, belongs to the field of nondestructive testing, and solves the problems of low full-focusing imaging efficiency, insufficient deep defect detection sensitivity and unbalanced imaging quality at different depths caused by a fixed sparse rate during ultrasonic testing of the large-wall-thickness structural member (the thickness is greater than or equal to 300mm) in the prior art. The method comprises the following steps: S1, acquiring medium sound velocity and total thickness of a detected component and determining array parameters of the detected component; s2, collecting a full-matrix signal covering a detection depth range; s3, establishing an acoustic beam diffusion model, and calculating acoustic beam diffusion widths at different depths; s4, dynamically distributing the array element sparse rate of each depth; and S5, selecting an effective array element according to the sparse rate, calculating the delay time from the effective array element to the focus point, and reconstructing and imaging to obtain a full-focus detection image covering the whole detection depth. Efficient and accurate nondestructive testing of large-wall-thickness structural members such as pressure containers and heavy pipelines is achieved, and imaging quality and detection efficiency are both considered.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing technology, and in particular to a dynamic sparse ultrasonic full-focusing method for structural components with large wall thickness. Background Technology

[0002] With the rapid development of the petrochemical, nuclear power, and heavy equipment manufacturing industries, the application of thick-walled structural components is becoming increasingly widespread. These components typically exceed 300mm in thickness, and the presence of internal defects directly impacts production safety and equipment lifespan. Therefore, reliable non-destructive testing (NDT) technologies are essential to ensure their quality. Ultrasonic testing technology, due to its strong penetrating power and lack of radiation exposure, has become a primary method for defect detection in thick-walled structural components. Among these, ultrasonic total focusing imaging technology, with its high resolution and high defect detection capability, is gradually being adopted in industrial testing.

[0003] Existing ultrasonic total focusing imaging technology requires all elements of a phased array probe to acquire the full matrix signal. Each element emits ultrasonic waves sequentially, and the remaining elements receive the echoes. The image is then generated by calculating the delay point by point and performing virtual focusing. While this method ensures imaging accuracy, the data volume is enormous, and the delay calculation and signal superposition process is extremely complex, resulting in very low imaging efficiency and making it difficult to meet the needs of real-time detection in industrial settings. Later, sparse array technology reduced the number of elements involved in imaging to decrease data volume and improve efficiency. However, the sparsity and element arrangement of these technologies are fixed, failing to consider the characteristic of "the deeper the sound wave propagates, the more severe the energy diffusion" in thick-walled structures.

[0004] For structural components with a wall thickness exceeding 300mm, the problems of existing technologies are more prominent: on the one hand, if a fixed low sparsity ratio is used, the sound wave diffusion range in the shallow region is small, and the excess array elements will lead to data redundancy and slower calculation; on the other hand, if a fixed high sparsity ratio is used, the sound wave diffusion range in the deep region is large, and too few array elements will lead to insufficient sound beam energy, making it impossible to detect tiny defects, and it is difficult to balance detection efficiency and defect detection capability. Summary of the Invention

[0005] Based on the above analysis, the embodiments of the present invention aim to provide a dynamic sparse ultrasonic full-focusing method for thick-walled structural components, in order to at least solve one of the problems of low imaging efficiency, insufficient sensitivity in detecting deep defects, and uneven imaging quality at different depths when inspecting thick-walled structural components.

[0006] On one hand, embodiments of the present invention provide a dynamic sparse ultrasound full-focusing imaging method for thick-walled structural components, comprising the following steps:

[0007] S1. Obtain the medium sound velocity and total thickness of the thick-walled structural component under inspection, select a phased array probe and determine its array parameters, wherein the thickness of the thick-walled structural component is ≥300mm;

[0008] S2. Control the phased array probe to acquire full matrix signals covering the detection depth range;

[0009] S3. Based on the array parameters and the sound velocity of the medium, establish a sound beam diffusion model and calculate the sound beam diffusion width at different depths;

[0010] S4. The element sparsity of each depth is dynamically allocated according to the beam diffusion width, and the sparsity decreases monotonically as the depth increases.

[0011] S5. Select effective array elements according to the sparsity ratio, calculate the delay time from the effective array elements to the focal point, and reconstruct the image to obtain a fully focused detection image covering the entire detection depth.

[0012] Furthermore, in step S1, the array parameters include the array aperture L0, the total number of array elements n, and the center frequency f, and the medium sound velocity is the longitudinal wave sound velocity c.

[0013] Furthermore, in step S3, the expression for the sound beam diffusion model is:

[0014]

[0015] Where L(h) is the beam spread width at depth h; L0 is the array aperture of the phased array probe; h is the depth of the area to be detected; c is the medium velocity of the thick-walled structure; and f is the center frequency of the phased array probe.

[0016] Furthermore, in step S4, the expression for the dynamic allocation of array element sparsity rate is:

[0017]

[0018] Where S(h) is the sparsity at depth h; S(H) is the reference sparsity at total thickness H; h is the depth of the region to be detected; H is the total thickness of the thick-walled structure; c is the medium sound velocity of the thick-walled structure; f is the center frequency of the phased array probe; and L0 is the array aperture of the phased array probe.

[0019] Furthermore, in step S5, the expression for calculating the delay time is:

[0020]

[0021] Among them, t ij x represents the delay time between the transmitting and receiving array elements; i z is the x-coordinate of the transmitting element; i x is the ordinate of the transmitting element; j z is the x-coordinate of the receiving array element; jdenoted as ordinate of the receiving array element; x is the abscissa of the focal point; z is the ordinate of the focal point; and c is the sound velocity of the medium in the thick-walled structural component.

[0022] Furthermore, in step S5, the expression for the reconstructed image is:

[0023]

[0024] Where I(x,z) is the amplitude at the focal point (x,z); M is the total number of depth sublayers; m is the index of the depth sublayer; h m S(h) represents the depth corresponding to the m-th depth sublayer. m ) represents the sparsity of the m-th depth sublayer; nS(h m ) represents the number of effective array elements in the m-th depth sublayer, i represents the index of the transmitting array element in the m-th depth sublayer, j represents the index of the receiving array element in the m-th depth sublayer; S ij (t ij (x,z)) represents the normalized amplitude of the filtered and gain-compensated signal; t ij (x,z) represents the time delay between the focal points (x,z) of transmitting element i and receiving element j.

[0025] Furthermore, step S5 includes dividing the detection depth range into M consecutive sub-layers, where M≥2, and the depth span of each sub-layer is 50-100mm;

[0026] The effective array elements are selected at equal intervals, and the center-to-center distance of the effective array elements is ≤2λ0, where λ0 is the ultrasonic wavelength and λ0 = c / f.

[0027] Furthermore, in step S5, the fully focused detection image is obtained in the following way:

[0028] The reconstructed images of each depth sub-layer are aligned according to their spatial location, and the maximum value of the amplitude at the same spatial location is taken as the final amplitude.

[0029] The medium for the thick-walled structural components includes carbon steel, stainless steel, titanium alloy, or composite materials.

[0030] Furthermore, in step S2, during the acquisition of the full matrix signal, the signal sampling time of the receiving array element is not less than the time required for the ultrasonic wave to travel from the probe to the maximum depth of the thick-walled structural component and back.

[0031] Accordingly, the present invention also discloses an ultrasonic nondestructive testing system, comprising:

[0032] A phased array probe used to transmit and receive ultrasonic waves, with an array aperture L0 = 10-50 mm and an element spacing ≤ 0.5λ0;

[0033] The data acquisition module is used to control the probe to acquire full matrix signals, with a sampling rate ≥20MHz;

[0034] The signal processing module is used to execute the dynamic sparse ultrasonic full-focusing imaging method for thick-walled structural components described above.

[0035] The display module is used to display the total focus detection image and imaging parameters.

[0036] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0037] 1) This invention establishes a beam diffusion model based on the phased array probe aperture and the sound velocity of the medium. It can accurately calculate the beam diffusion range at different depths and dynamically allocate the sparsity ratio accordingly. In shallow regions, the beam diffusion is narrow, so the number of effective array elements is appropriately reduced to avoid resource waste caused by element redundancy. In deep regions, the beam diffusion is wide, so the number of effective array elements is appropriately increased to ensure sufficient beam energy to cover the diffusion range. This design ensures consistent defect detection capability at all depths and solves the problems of shallow redundancy and deep inefficiency caused by a fixed sparsity ratio.

[0038] 2) The dynamic sparsity strategy of this invention reduces the number of effective array elements reasonably with depth, and reduces the amount of full matrix signal data by nearly 50%; combined with layered imaging and sub-image fusion, the imaging time is shortened from 25.8s for the traditional full array to 13.5s, the efficiency is improved by 47.7%, and the imaging quality is not significantly reduced, which meets the requirements of real-time detection in industrial fields.

[0039] 3) This invention clarifies the probe parameter range (e.g., center frequency 5MHz, aperture 20mm) and effective array element selection rules (equally spaced arrangement), making it applicable to the inspection of thick-walled structural components in various media and suitable for different application scenarios such as pressure vessels and heavy pipelines. Simultaneously, through a sound beam diffusion model, it accurately matches the sound beam coverage at various depths; combined with the coordinate relationship between the effective array elements and the focal point, it accurately calculates the delay time, enabling reverse location of the actual defect corresponding to the peak signal amplitude. Even when inspecting 500mm ultra-thick components, it can still guarantee defect location accuracy, providing reliable technical support for the quality control of thick-walled structural components.

[0040] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description

[0041] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0042] Figure 1 This is a schematic diagram of the ultrasonic testing system used.

[0043] Figure 2 A schematic diagram of a thick-walled stainless steel structural component with transverse through holes of different depths.

[0044] Figure 3 The image shows the results of transverse aperture imaging using the original full matrix signal;

[0045] Figure 4 The image shows the results of transverse aperture imaging using the dynamic sparse full matrix signal of this invention.

[0046] Figure label:

[0047] 1. Phased array probe; 2. Data acquisition module; 3. Signal processing module; 4. Display module; 5. The structural component with large wall thickness under test. Detailed Implementation

[0048] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0049] In existing technologies, ultrasonic full-focusing imaging of thick-walled structural components faces the dual challenges of low efficiency in full-element mode and uneven imaging quality in fixed sparse mode. This is especially true for structural components with a thickness ≥300mm, where severe deep beam diffusion leads to insufficient defect detection rates. This invention solves the efficiency problem and ensures uniform imaging across the entire depth range by employing a beam diffusion model combined with dynamic sparsity allocation.

[0050] A specific embodiment of the present invention discloses a dynamic sparse ultrasound full-focusing imaging method for thick-walled structural components, comprising the following steps:

[0051] S1. Obtain the medium sound velocity and total thickness of the thick-walled structural component under inspection, select a phased array probe and determine its array parameters, wherein the thickness of the thick-walled structural component is ≥300mm;

[0052] S2. Control the phased array probe to acquire full matrix signals covering the detection depth range;

[0053] S3. Based on the array parameters and the sound velocity of the medium, establish a sound beam diffusion model and calculate the sound beam diffusion width at different depths;

[0054] S4. The element sparsity of each depth is dynamically allocated according to the beam diffusion width, and the sparsity decreases monotonically as the depth increases.

[0055] S5. Select effective array elements according to the sparsity ratio, calculate the delay time from the effective array elements to the focal point, and reconstruct the image to obtain a fully focused detection image covering the entire detection depth.

[0056] This solution employs a sparse strategy that dynamically adapts to the acoustic beam diffusion characteristics, unlike existing technologies that use fixed-parameter imaging. Thick-walled structural components refer to metal or composite material components with a thickness ≥300mm. These components have longer acoustic propagation paths, resulting in significantly greater acoustic beam diffusion and energy attenuation compared to conventional thin-walled components. The dynamic sparsity rate means that the sparsity rate is not a fixed value but dynamically adjusts with the detection depth, showing a negative correlation with the acoustic beam diffusion width: the shallower the depth, the narrower the acoustic beam diffusion and the higher the sparsity rate (fewer effective array elements); the deeper the depth, the wider the acoustic beam diffusion and the lower the sparsity rate (more effective array elements).

[0057] During implementation, the core parameters of the component under test (total thickness H, longitudinal wave velocity c) and the probe parameters (array aperture L0, total number of array elements n, center frequency f) are first defined. This is the basis for the subsequent construction of the sound beam diffusion model. Although existing technologies also select probe parameters, they are not associated with the depth-adaptive sparse strategy.

[0058] Then the full matrix signal is acquired, which consists of n 2 The time-domain signals of each transmit-receive array element pair constitute the complete detection depth range [h]. a ,h β ], providing raw data support for layered imaging, where h a The starting depth for the detection depth range corresponds to the depth near the probe side surface of a thick-walled structural component, and is typically set to 0 mm, h. β To determine the termination depth of the detection depth range, the depth corresponding to the far-probe side surface of the thick-walled structural component is equal to the total thickness H of the component. The acquired full-matrix signal needs to undergo preprocessing: high-frequency noise and low-frequency interference are removed by bandpass filtering (filter range 0.5f-2f), and then depth gain compensation (DGC) is used to compensate for energy attenuation during sound beam propagation, avoiding the influence of noise and attenuation in the original signal on subsequent sparsity processing and imaging quality.

[0059] This step is consistent with the signal acquisition logic of traditional all-focus imaging. However, traditional methods do not optimize preprocessing parameters for the acoustic attenuation characteristics of thick-walled structures. In contrast, the filtering and gain compensation parameters of this scheme are adapted in coordination with the subsequent acoustic beam diffusion model and sparsity allocation. For example, when f = 5MHz, the filtering range is set to 2.5MHz-10MHz, and the depth gain compensation coefficient is adjusted synchronously with the acoustic beam diffusion width to further improve the signal-to-noise ratio.

[0060] Next, a sound beam diffusion model is established, and the sound beam diffusion width L(h) at different depths h is derived using geometric acoustics principles, accurately quantifying the diffusion law of the sound beam with depth. Existing technologies do not have such a model, resulting in a lack of theoretical support for sparsity allocation. Based on this sound beam diffusion model, the sparsity S(h) corresponding to each depth is dynamically calculated to ensure that the effective number of array elements at each depth matches the sound beam diffusion requirements, avoiding redundancy in shallow layers and insufficient elements in deep layers.

[0061] Finally, by selecting effective array elements at equal intervals, calculating the delay time, reconstructing and fusing the images layer by layer, full-focus imaging is completed. The entire process retains the high-resolution advantage of full-focus imaging while significantly reducing the computational load through dynamic sparsity.

[0062] The method provided in this embodiment has two main advantages. First, it solves the problem that a fixed sparsity ratio cannot adapt to the diffusion characteristics of thick-walled sound beams. By accurately calculating the number of effective array elements required at each depth through a model, the method eliminates the need for redundant array elements in shallow layers, ensures sound beam energy in deeper layers, and achieves uniform imaging quality at all depths. Second, it significantly reduces the amount of data and computation time without sacrificing imaging accuracy. This is because the number of effective array elements is reasonably reduced with depth, resulting in a significant reduction in the amount of full matrix signal processing, thus meeting the real-time detection needs of industrial sites.

[0063] The construction of the acoustic beam diffusion model provides a clear basis for sparsity allocation, rather than blindly reducing array elements; the dynamic adjustment strategy precisely matches the acoustic propagation characteristics of thick-walled structural components. The two work together to overcome the low efficiency of the traditional full-element mode and make up for the uneven imaging quality of the fixed sparse mode, ultimately achieving the dual goals of high efficiency and accuracy.

[0064] Furthermore, in step S1, the array parameters include the array aperture L0, the total number of array elements n, and the center frequency f, and the medium sound velocity is the longitudinal wave sound velocity c.

[0065] Specifically, the array aperture L0 determines the width of the initial sound beam, the total number of array elements n affects the imaging resolution and data volume, and the center frequency f is related to the sound wavelength λ0 (λ0 = c / f), directly affecting the sensitivity of defect detection. The longitudinal wave velocity c is chosen because longitudinal waves have the strongest penetrating power in ultrasonic testing and are the main wave type for detecting thick-walled structural components. Other wave types, such as transverse waves, are not suitable due to their rapid attenuation.

[0066] In step S2, during the acquisition of the full matrix signal, the signal sampling time of the receiving array element is not less than the time required for the ultrasonic wave to travel from the probe to the maximum depth of the thick-walled structural component and back.

[0067] Specifically, the sampling duration T satisfies the formula T≥2H max / c, where H maxWhere c is the maximum depth of the thick-walled structural component, and c is the longitudinal wave velocity determined in step S1. This limitation ensures that the round-trip signal of the ultrasonic wave is completely acquired across the entire depth range of the structural component, avoiding the loss of deep defect reflection signals due to insufficient sampling time, and providing a comprehensive data foundation for subsequent dynamic sparse focusing processing.

[0068] Furthermore, in step S3, the expression for the sound beam diffusion model is formula (1):

[0069]

[0070] Where L(h) is the beam spread width at depth h, 3.83 is the first-order minimum coefficient of circular aperture diffraction; L0 is the array aperture of the phased array probe; h is the depth of the region to be detected; c is the medium velocity of the thick-walled structure; and f is the center frequency of the phased array probe. Formula (1) is derived based on geometric acoustics and diffraction theory, and can accurately calculate the beam spread width at different depths, providing a quantitative basis for sparsity allocation.

[0071] For example, when L0 = 20mm, c = 5890m / s, and f = 5MHz, L(h) = 34.36mm at a depth h = 50mm and L(h) = 91.80mm at a depth h = 450mm, clearly demonstrating the diffusion law of the sound beam with depth.

[0072] Furthermore, in step S4, the expression for the sparsity rate of the dynamically allocated array elements is formula (2):

[0073]

[0074] Where S(h) is the sparsity at depth h, and S(H) is the reference sparsity at total thickness H; h is the depth of the area to be detected; H is the total thickness of the thick-walled structure; c is the medium velocity of the thick-walled structure; f is the center frequency of the phased array probe; and L0 is the array aperture of the phased array probe. Formula (2) uses the reference sparsity S(H) at total thickness H as a reference to dynamically calculate the sparsity at each depth h, ensuring that S(h) decreases monotonically as h increases.

[0075] For example, when H = 500 mm and S(H) = 1.0, S(h) = 0.37 at h = 50 mm and S(h) = 0.92 at h = 450 mm, which ensures sufficient effective array elements in the deep layer and reduces redundancy in the shallow layer.

[0076] It is worth noting that the baseline sparsity S(H) ranges from 0.5 to 1.0, and can be flexibly adjusted according to the dielectric attenuation characteristics of the structural components: the larger the dielectric attenuation coefficient (e.g., carbon fiber reinforced composite materials, longitudinal wave attenuation coefficient α = 0.5 dB / mm), the larger the value of S(H), preferably 0.8 to 1.0, to compensate for attenuation by increasing the number of effective deep array elements; the smaller the dielectric attenuation coefficient (e.g., stainless steel, α = 0.01 dB / mm), the smaller the value of S(H), preferably 0.5 to 0.8, to further reduce the amount of data while ensuring the beam energy. This optimized selection makes the dynamic sparsity allocation more suitable for thick-walled structural components with different media, avoiding the problems of blurred deep imaging in media with large attenuation or data redundancy in media with small attenuation caused by blindly selecting S(H).

[0077] Based on this, effective array elements need to be selected according to the sparsity corresponding to each depth, and the delay time from the effective array element to the focal point needs to be calculated. Specifically, the selection method is to screen from all array elements of the probe according to the principle of equal intervals. The center-to-center spacing of the effective array elements ≤ 2λ0 (λ0 is the ultrasonic wavelength, λ0 = c / f); the number of effective array elements is determined by the product of the sparsity S(h) at the corresponding depth and the total number of array elements; the effective array elements can be identified by their two-dimensional coordinate position (x, y) in the probe array. i ,z i ) clearly express, where x i For the lateral position parameters of the array element, z i The longitudinal position parameter can also be expressed by the array element number. That is, all array elements of the probe are numbered 1, 2, ..., n in order from left to right and from top to bottom. The effective array elements are selected at equal intervals according to the sparsity S(h) of the corresponding depth, corresponding to the specific number set {1, 3, 5, ...}.

[0078] Furthermore, in step S5, the calculation expression for the delay time is formula (3):

[0079]

[0080] Among them, t ij x represents the delay time between the transmitting and receiving array elements; i z is the x-coordinate of the transmitting element; i x is the ordinate of the transmitting element; j z is the x-coordinate of the receiving array element; j denoted as ordinate of the receiving array element; x is the abscissa of the focal point; z is the ordinate of the focal point; and c is the sound velocity of the medium in the thick-walled structural component.

[0081] (x i ,z i ) and (x j ,z j(x, z) represent the coordinates of the transmitting and receiving elements, respectively (both can be expressed using two-dimensional coordinates in the probe array or the corresponding element numbers), and (x, z) represent the coordinates of the focal point. Specifically, the detection area of ​​the structural component needs to be divided into a uniform grid according to the required detection resolution. The detection depth range is [h]. a ,h β Each grid point within the range corresponds to a focal point (x, z).

[0082] Formula (3) is based on the principle of geometric sound path to calculate the total propagation time t of the sound wave from the transmitting element i to the focal point (x,z) and then reflected to the receiving element j. ij In the subsequent reconstruction imaging process, this delay time will be directly used, that is, t involved in formula (4). ij (x,z). By capturing the signals of all relevant transmit-receive array element pairs, and performing [analysis] with t ij The corresponding delay superposition can provide the basis for the amplitude calculation in formula (4), and finally realize the coherent synthesis of the focal point, so that its reflected energy is concentrated in the image.

[0083] Based on the above calculation results of the delay time, the reconstruction of the fully focused detection image covering the entire detection depth can be completed. The reconstruction process is based on the dynamic sparsity of each depth sub-layer, and combines the delay superposition signals of different focal points to achieve coherent synthesis of the reflected energy of each focal point, and finally stitches together the detection image of the full depth range.

[0084] Specifically, obtaining the focal point amplitude I(x,z) requires, for each focal point (x,z), calling the acquisition signal of the corresponding transmit-receive array element pair, and calculating the delay time t according to formula (3). ij The signal is time-aligned and superimposed, and then the sparsity S(h) of the m-th depth sub-layer is obtained by signal amplitude extraction. m Then, based on the sound beam diffusion model established in step S4, combined with the corresponding depth h, m The acoustic beam diffusion characteristics and the attenuation coefficient of the structural component medium were selected from the reference sparsity range (0.1 to 0.5); while the effective number of array elements nS(h) of this depth sublayer was obtained. m ), simply adjust the sparsity S(h) of the corresponding depth sublayer. m The parameters can be obtained by multiplying the total number of array elements n of the probe. These parameters together provide core data support for subsequent reconstruction imaging.

[0085] Furthermore, in step S5, the expression for the reconstructed image is formula (4):

[0086]

[0087] Where I(x,z) is the amplitude at the focal point (x,z); M is the total number of depth sublayers; m is the index of the depth sublayer; h m S(h) represents the depth corresponding to the m-th depth sublayer. m ) represents the sparsity of the m-th depth sublayer; nS(h m ) represents the number of effective array elements in the m-th depth sublayer, i represents the index of the transmitting array element in the m-th depth sublayer, j represents the index of the receiving array element in the m-th depth sublayer; S ij (t ij (x,z)) represents the normalized amplitude of the filtered and gain-compensated signal; t ij (x,z) represents the time delay between the focal points (x,z) of transmitting element i and receiving element j.

[0088] Formula (4) achieves hierarchical reconstruction and fusion through triple summation. The outer summation symbol ∑ corresponds to M deep sub-layers, and the inner double ∑ summation symbol targets the effective array element pairs in each sub-layer. First, the signals of all effective transmit-receive array element pairs in a single sub-layer are delayed and superimposed, i.e., S ij (t ij The summation of (x,z) yields the full-focus sub-image of that sub-layer; then, the results of all sub-layer sub-images are summed and aggregated through the outer layer to finally form a full-focus image covering the entire detection depth, ensuring that no imaging information in each depth region is missed, while also guaranteeing the continuity and integrity of the image.

[0089] The present invention transforms the technical solution from a qualitative description to a quantitative implementation through the above formulas (1)-(4), which not only clarifies the specific operation steps of the technical solution, but also combines the sound beam diffusion model with dynamic sparsity, delay time and reconstruction algorithm through formula form.

[0090] Furthermore, in step S5, the detection depth range is first divided into M consecutive sub-layers, where M≥2, and the depth span of each sub-layer is 50-100mm; then, for each depth sub-layer, the sparsity S(h) of that sub-layer is first determined. m Effective array elements are selected at equal intervals, with the center-to-center spacing of the effective array elements ≤ 2λ0, where λ0 is the ultrasonic wavelength, λ0 = c / f. The number of effective array elements is determined by the sparsity S(h) of the corresponding sublayer. m The product of the total number of array elements is determined.

[0091] Specifically, the sublayer span is set to 50-100mm because in thick-walled structural components, the change in sound beam diffusion width within the 50-100mm range is relatively gradual, and the sparsity adaptability can be guaranteed without further fine division. If the span is too large (e.g., exceeding 100mm), it will lead to a large difference in sound beam diffusion at different depths within the same sublayer, affecting imaging uniformity. If the span is too small (e.g., less than 50mm), it will increase the number of sublayers and calculation steps, reducing efficiency.

[0092] The selection of effective array elements at equal intervals is to ensure that the effective array elements are uniformly distributed on the array aperture, avoiding beam distortion caused by the concentration of array elements. The limitation of center spacing ≤ 2λ0 is based on the sampling theorem, ensuring that the distribution density of effective array elements can completely capture the beam information, and avoiding the decrease in imaging resolution caused by excessive array element spacing.

[0093] For example, when λ0 = 1.178 mm, the effective array element center spacing is ≤ 2.356 mm. When selecting probes with n = 32 and L0 = 20 mm, the minimum spacing when selecting at equal intervals is 0.625 mm, which fully meets the requirements.

[0094] Furthermore, to address complex scenarios such as non-uniform internal media and abnormal local attenuation within structural components, this solution can also supplement the dynamic adjustment strategy for effective array elements:

[0095] According to the sparsity S(h) m After selecting effective array elements and completing the imaging of the sub-layer, the imaging signal-to-noise ratio (SNR) is calculated in real time. If the SNR < 10dB (the conventional threshold for industrial testing), the sparsity of the sub-layer is automatically reduced and the number of effective array elements is increased until the SNR meets the standard.

[0096] This strategy can avoid missing defects due to local medium anomalies. For example, if there are local inclusions in a certain depth area of ​​a structural component, the sound attenuation will suddenly increase. By dynamically adjusting the number of effective array elements, the energy loss caused by attenuation can be compensated, ensuring that the defect signal is clearly identifiable.

[0097] This invention achieves layered adaptation through sub-layer division, making the sparsity of each sub-layer more accurate; the selection of array elements at equal intervals ensures the uniformity and resolution of the sound beam. The combination of the two further improves the stability of imaging quality, while balancing computational efficiency and avoiding problems caused by overly fine sub-layer division or uneven array element selection, thus ensuring a high signal-to-noise ratio and defect detection capability in the final image.

[0098] Furthermore, in step S5, the fully focused detection image is obtained by aligning the reconstructed images of each depth sub-layer according to their spatial positions. The reconstructed image of each layer is obtained by spatially arranging the amplitude I(x,z) of all focal points within that sub-layer, and then taking the maximum value of the amplitude at the same spatial position as the final amplitude.

[0099] Specifically, the maximum amplitude is chosen because the reconstructed image of each sub-layer has the highest amplitude and the best signal-to-noise ratio within its corresponding depth range. Taking the maximum value preserves the best imaging effect of each depth sub-layer and avoids signal attenuation or noise enhancement caused by simple superposition. For example, the reconstructed image of a shallow sub-layer has a high amplitude in the shallow region, and the reconstructed image of a deep sub-layer has a high amplitude in the deep region. After taking the maximum value, the defect signal can be clearly presented throughout the entire detection depth range.

[0100] The medium for the thick-walled structural components includes carbon steel, stainless steel, titanium alloy, or composite materials. These are commonly used materials for thick-walled structural components in industrial fields, covering major application scenarios such as petrochemicals, nuclear power, and heavy equipment. Specific types include pressure vessel shells, heavy pipelines, or thick-walled steel structural components.

[0101] In summary, the method of this invention achieves dynamic sparsity adaptive allocation based on detection depth by establishing a sound beam geometric diffusion model; then, it layers the imaging region and selects effective array elements within each layer for independent full-focus reconstruction according to the sparsity; finally, it fuses the sub-images of each layer to form a complete detection result. This method significantly reduces data volume and computational complexity while ensuring imaging quality and defect detection capability across the entire depth range, thus improving imaging efficiency. It is particularly suitable for rapid, high-reliability non-destructive testing of thick-walled structures such as pressure vessels and heavy pipelines.

[0102] To achieve compatibility between the method and the equipment, this invention also proposes an ultrasonic non-destructive testing system. For example... Figure 1 As shown, the system includes:

[0103] Phased array probe 1 is used to transmit and receive ultrasonic waves, with an array aperture L0 = 10-50 mm and an element spacing ≤ 0.5λ0; data acquisition module 2 is used to control the probe to acquire full matrix signals, with a sampling rate ≥ 20 MHz; signal processing module 3 is used to execute the methods described above; display module 4 is used to display the fully focused detection image and imaging parameters.

[0104] Specifically, the probe array aperture L0 = 10-50mm, adaptable to structural components of different thicknesses (the smaller the thickness, the smaller the aperture can be selected; the larger the thickness, the larger the aperture is required to ensure the initial beam width); the element spacing ≤ 0.5λ0 is to avoid grating lobes and ensure the directivity of the beam; the sampling rate ≥ 20MHz is more than 4 times the center frequency f, satisfying the sampling theorem and ensuring the integrity of the entire matrix signal.

[0105] in addition, Figure 1The system also illustrates a typical working object: a thick-walled structural component 5 under test. In practical use, the system couples the phased array probe 1 with the surface of the thick-walled structural component 5 to achieve non-destructive testing and imaging of its internal structure (such as possible defects like transverse through-holes and cracks). This system is particularly suitable for rapid and highly reliable testing of thick-walled structural components such as pressure vessels and heavy pipelines with a thickness of not less than 300 mm.

[0106] This system provides the hardware requirements for the method, avoiding the problem that general-purpose equipment cannot meet the method requirements, and further improves the reliability and stability of the detection. It is especially suitable for complex detection environments in industrial sites, ensuring the practical application effect of the technical solution.

[0107] The present invention will be described in more detail below through specific embodiments. These embodiments are merely descriptions of the best implementation of the invention and do not limit the scope of the invention in any way.

[0108] Example 1

[0109] A dynamic sparse ultrasonic full-focusing method for thick-walled structural components, employing an ultrasonic testing system such as... Figure 1 As shown, the system includes:

[0110] Phased array probe 1 is used to transmit and receive ultrasonic waves, with an array aperture L0 = 10-50 mm and an element spacing ≤ 0.5λ0; data acquisition module 2 is used to control the probe to acquire full matrix signals, with a sampling rate ≥ 20 MHz; signal processing module 3 is used to execute the methods described below; display module 4 is used to display the fully focused detection image and imaging parameters.

[0111] The steps are as follows:

[0112] S1. In this embodiment, the workpiece under inspection is a thick-walled stainless steel structural component with a total thickness H = 500 mm and a longitudinal wave velocity c = 5890 m / s. Figure 2 As shown, to verify the method, two transverse through holes with a diameter of 2 mm and different center-to-center spacing were machined every 100 mm within a depth range of 50 mm to 450 mm inside the workpiece to simulate defects. Based on the detection requirements, a linear phased array probe with a center frequency f = 5 MHz, an array element number n = 32, and an array aperture L0 = 20 mm was used.

[0113] S2. The probe is coupled to the surface of the stainless steel structural component under inspection using a commonly used industrial ultrasonic coupling agent to ensure efficient ultrasonic wave transmission. The probe is controlled to operate in full-matrix acquisition mode, sequentially exciting each array element to emit ultrasonic waves and simultaneously recording the echo signals received by all array elements, thereby acquiring a full-matrix signal covering the detection depth range of [0, 500] mm (i.e., 0-500 mm). This dataset contains n2 (i.e., 32×32=1024) time-domain A-scan signals.

[0114] S3. Based on the probe aperture L0 = 20 mm, sound velocity c = 5890 m / s and frequency f = 5 MHz, establish the ultrasonic beam diffusion model:

[0115]

[0116] After dividing the 0-500mm depth range into 5 equal layers (each layer 100mm), the calculated diffusion width L(h) at the center depth of each layer is as follows: 34.36mm, 48.72mm, 63.08mm, 77.44mm, and 91.80mm.

[0117] S4. Using the diffusion width L(H) = 91.80 mm at the deepest layer (h = 500 mm) as a baseline, set the sparsity S(H) = 1 (i.e., use all array elements). Based on the diffusion width ratio, dynamically determine the sparsity at other depths:

[0118]

[0119] The calculated dynamic sparsity S(h) of each layer from shallow to deep is as follows: 0.37, 0.53, 0.69, 0.84, 1.00.

[0120] S5. For each depth sub-layer (5 layers) partitioned in step S4, according to the sparsity S(h) assigned to that layer... m ), calculate the effective number of array elements N m =n·S(h m N elements were selected from all 32 array elements using an equal-interval method. m Each element serves as an effective array element for imaging this layer.

[0121] Next, according to the detection resolution requirements, the detection area of ​​each sub-layer is divided into a uniform grid of 0.5mm × 0.5mm (horizontal x-direction × vertical z-direction). Each grid point is a focal point (x, z). For example, the focal point coordinates in the first sub-layer (0-100mm) can be represented as (0.5mm, 0.5mm), (1.0mm, 0.5mm), ..., (20.0mm, 99.5mm). For all focal points (x, z) in each layer, the total propagation delay t of the sound wave from each effective transmitting element i to the focal point and then reflected to each effective receiving element j is calculated. ij :

[0122]

[0123] The focal point (x, z) = (10 mm, 250 mm) within the third sublayer (200-300 mm) is selected, where x is the lateral coordinate and z is the depth coordinate of the sublayer center. Since the probe array has 32 elements and there are 31 gaps between adjacent elements, with an array aperture L0 = 20 mm, the element spacing d = 20 mm / 31 ≈ 0.645 mm. Taking the first element as the x-axis reference point (x = 0 mm), the lateral coordinate of the i-th element is d × (i-1), and the reference plane where the probe element is located is z = 0 mm. Therefore, the coordinates of the effective transmitting element i = 10 are (x, z = 10 mm ... 10 ,z 10 )=(0.645mm×9,0mm)=(5.805mm,0mm), the effective receiving array element j=20 has coordinates (x 20 ,z 20 = (0.645mm × 19, 0mm) = (12.255mm, 0mm). Substituting the above parameters into formula (3), we get the sound path from the transmitting element to the focal point ≈ 250.035mm and the sound path from the receiving element to the focal point ≈ 250.010mm. Combining this with the longitudinal wave velocity of the inspected workpiece c = 5890m / s, we calculate the total propagation delay time t. ij ≈(250.035+250.010) / 5890≈0.0849ms.

[0124] Using this delay time, the corresponding A-scan signals in the full matrix signal are aligned and coherently superimposed, and the full-focus sub-image of that layer is synthesized point by point. Finally, the imaging results of all depth sub-layers are fused by taking the maximum amplitude of the corresponding pixels to generate a complete full-focus image covering the entire detection depth range of 0-500mm (as shown in the attached image). Figure 4 ).

[0125] To verify the effectiveness of this invention, imaging was performed using both the original full-matrix signal (unsparsed) and the signal after the aforementioned dynamic sparsity processing. The results are as follows: Figure 3 and Figure 4 As shown.

[0126] As can be seen from the comparison, the image obtained by the dynamic sparse method ( Figure 4 The clarity and positioning accuracy of transverse through-hole defects at various depths in the image are compared with those of the original method image. Figure 3 The results were basically consistent, and no significant decrease in image quality was observed.

[0127] In terms of imaging efficiency, actual measurements show that it takes about 25.8 seconds to complete a full-focus imaging using the original full-matrix signal, while the imaging time is shortened to about 13.5 seconds after adopting the dynamic sparse method of the present invention, with an efficiency improvement of about 47.7%, which significantly speeds up the detection speed of thick-walled structural components.

[0128] Example 2

[0129] The structural component under inspection is the same as in Example 1; the only difference in the method and steps is:

[0130] In S1, the probe parameters are: center frequency f = 2.25MHz, total number of array elements n = 16, array aperture L0 = 10mm, and array element spacing 0.63mm.

[0131] In S3, the depth range of 0-500mm is divided into 4 layers (each layer is 125mm), and the calculated L(h) are 38.75mm, 61.23mm, 83.71mm, and 106.19mm respectively.

[0132] In S4, the baseline sparsity S(H) = 0.8, and S(h) are 0.42, 0.54, 0.66, and 0.80 respectively.

[0133] Example 3

[0134] The difference from Example 1 is that the inspected structural component is made of carbon steel with a thickness of H=300mm, a longitudinal wave velocity of c=5900m / s, and two transverse through holes with a diameter of 1.5mm are machined every 100mm at a depth of 50-250mm inside.

[0135] In S1, the probe parameters are: center frequency f = 4MHz, total number of array elements n = 64, array aperture L0 = 50mm, and array element spacing 0.78mm.

[0136] In S3, the depth range of 0-300mm is divided into 3 layers (each layer is 100mm), and the calculated L(h) are 52.18mm, 79.34mm, and 106.50mm respectively;

[0137] In S4, the baseline sparsity S(H) = 0.6, and S(h) are 0.36, 0.48, and 0.60 respectively.

[0138] Comparative Example 1

[0139] This comparative example uses a conventional full-focus imaging method. The inspected structure, probe parameters, and full-matrix signal acquisition process are all the same as in Example 1. The only difference from Example 1 is that this comparative example does not employ any sparsity strategy or perform layered processing.

[0140] Specifically, it uses the full matrix signals acquired by all 32 array elements and directly performs single-delay superposition and reconstruction of the signals of all array element pairs within the entire 0-500mm detection depth range. That is, it only performs the conventional full-focus imaging algorithm without dynamic sparsity allocation, layered imaging and image fusion steps.

[0141] Comparative Example 2

[0142] This comparative example employs a layered full-focusing method with a fixed sparsity ratio. The inspected structure, probe parameters, and detection depth range division (5 layers) are all the same as in Example 1. The only difference from Example 1 is that this comparative example uses a fixed sparsity ratio (S = 0.5), meaning that all depth sub-layers use the same number of effective array elements (16) for imaging, without dynamic adjustment based on the acoustic beam diffusion model. Although its reconstructed imaging uses layered summation, the number of effective array elements in each sub-layer is fixed.

[0143] Characterization results and analysis

[0144] The characterization results of the above-described embodiments and comparative examples are shown in Table 1 below.

[0145] Table 1

[0146] serial number Example 1 Example 2 Example 3 Comparative Example 1 Comparative Example 2 Imaging time (s) 13.5 6.6 12.1 27.8 14.1 Shallow defect detection rate (%) 100 100 100 100 100 Deep defect detection rate (%) 100 100 100 100 80 Uniformity of imaging at various depths (%) 98 95 97 98 60 Signal-to-noise ratio (SNR, dB) 18.6 15.2 17.8 18.9 10.0

[0147] As can be seen from Table 1, Examples 1-3 have excellent and stable overall performance: the imaging time of 13.5-18.7s meets the requirements for real-time detection, the detection rate of shallow defects is 100%, the detection rate of deep defects is 95%-100%, the imaging uniformity is 95%-98%, the signal-to-noise ratio is 15.2-18.6dB, and the scene adaptability is strong.

[0148] However, Comparative Examples 1 and 2 have obvious shortcomings: Comparative Example 1 has satisfactory imaging quality but takes 27.8s, which is less than 50% of the efficiency of this invention; Comparative Example 2 has improved efficiency (14.1s), but the deep detection rate is only 80%, the uniformity is 60%, the signal-to-noise ratio is 10.0dB, the imaging quality is degraded, and the scene adaptability is poor.

[0149] In summary, the solution of this invention achieves both high efficiency and accuracy in the inspection of structural components with wall thicknesses of 300-500mm, solves the core pain points of existing technologies, and has significant technical advantages and industrial application value.

[0150] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A dynamic sparse ultrasonic total focusing imaging method for thick-walled structural components, characterized in that, Includes the following steps: S1. Obtain the medium sound velocity and total thickness of the thick-walled structural component under inspection, select a phased array probe and determine its array parameters, wherein the thickness of the thick-walled structural component is ≥300mm; S2. Control the phased array probe to acquire full matrix signals covering the detection depth range; S3. Based on the array parameters and the sound velocity of the medium, establish a sound beam diffusion model and calculate the sound beam diffusion width at different depths; S4. The element sparsity of each depth is dynamically allocated according to the beam diffusion width, and the sparsity decreases monotonically as the depth increases. S5. Select effective array elements according to the sparsity ratio, calculate the delay time from the effective array elements to the focal point, and reconstruct the image to obtain a fully focused detection image covering the entire detection depth.

2. The method according to claim 1, characterized in that, In step S1, the array parameters include array aperture L0, total number of array elements n, and center frequency f, and the medium sound velocity is the longitudinal wave sound velocity c.

3. The method according to claim 1, characterized in that, In step S3, the expression for the sound beam diffusion model is: Where L(h) is the beam spread width at depth h; L0 is the array aperture of the phased array probe; h is the depth of the area to be detected; c is the medium velocity of the thick-walled structure; and f is the center frequency of the phased array probe.

4. The method according to claim 1, characterized in that, In step S4, the expression for the dynamic allocation of array element sparsity is: Where S(h) is the sparsity at depth h; S(H) is the reference sparsity at total thickness H; h is the depth of the region to be detected; H is the total thickness of the thick-walled structure; c is the medium sound velocity of the thick-walled structure; f is the center frequency of the phased array probe; and L0 is the array aperture of the phased array probe.

5. The method according to claim 1, characterized in that, In step S5, the expression for calculating the delay time is: Among them, t ij x represents the delay time between the transmitting and receiving array elements; i z is the x-coordinate of the transmitting element; i x is the ordinate of the transmitting element; j z is the x-coordinate of the receiving array element; j denoted as ordinate of the receiving array element; x is the abscissa of the focal point; z is the ordinate of the focal point; and c is the sound velocity of the medium in the thick-walled structural component.

6. The method according to claim 1, characterized in that, In step S5, the expression for the reconstructed image is: Where I(x,z) is the amplitude at the focal point (x,z); M is the total number of depth sublayers; m is the index of the depth sublayer; h m S(h) represents the depth corresponding to the m-th depth sublayer. m ) represents the sparsity of the m-th depth sublayer; nS(h m ) represents the number of effective array elements in the m-th depth sublayer, i represents the index of the transmitting array element in the m-th depth sublayer, j represents the index of the receiving array element in the m-th depth sublayer; S ij (t ij (x,z)) represents the normalized amplitude of the filtered and gain-compensated signal; t ij (x,z) represents the time delay between the focal points (x,z) of transmitting element i and receiving element j.

7. The method according to claim 1, characterized in that, Step S5 includes dividing the detection depth range into M consecutive sub-layers, where M≥2, and the depth span of each sub-layer is 50-100mm; The effective array elements are selected at equal intervals, and the center-to-center distance of the effective array elements is ≤2λ0, where λ0 is the ultrasonic wavelength and λ0 = c / f.

8. The method according to claim 1, characterized in that, In step S5, the fully focused detection image is obtained in the following way: The reconstructed images of each depth sub-layer are aligned according to their spatial location, and the maximum value of the amplitude at the same spatial location is taken as the final amplitude. The medium for the thick-walled structural components includes carbon steel, stainless steel, titanium alloy, or composite materials.

9. The method according to claim 1, characterized in that, In step S2, during the acquisition of the full matrix signal, the signal sampling time of the receiving array element is not less than the time required for the ultrasonic wave to travel from the probe to the maximum depth of the thick-walled structural component and back.

10. An ultrasonic nondestructive testing system, characterized in that, include: A phased array probe used to transmit and receive ultrasonic waves, with an array aperture L0 = 10-50 mm and an element spacing ≤ 0.5λ0; The data acquisition module is used to control the probe to acquire full matrix signals, with a sampling rate ≥20MHz; The signal processing module is configured to perform the method described in any one of claims 1-8; The display module is used to display the total focus detection image and imaging parameters.

Images