A High Signal-to-Noise Ratio Ultrasonic Imaging Method for Internal Defects in Materials

Abstract

This invention discloses a high signal-to-noise ratio (SNR) ultrasonic imaging method for internal defects in materials. By utilizing the spatial coherence of the received signals from each channel in the full matrix data and modifying the delay rule of the algorithm based on the surface configuration characteristics of PBX (partially filled composite material), high SNR imaging of crack defects with different orientations in PBXs with curved surface configurations is achieved. The quality of the reconstructed images is compared and evaluated. The results show that, compared with traditional linear synthetic focusing imaging, the algorithm proposed in this invention improves the SNR of ultrasonic reconstructed images of crack defects in particulate-filled composite material PBX by more than 10 dB, significantly improving the detection capability of crack defects while maintaining imaging efficiency.

Description

Technical Field

[0001] This invention relates to the field of testing and inspection technology, and in particular to a high signal-to-noise ratio ultrasonic imaging method for internal defects in materials. Background Technology

[0002] Particle-filled composite materials are two-phase or multi-phase composite materials composed of particulate materials and polymer-reinforced matrices. Due to their significant improvement and enhancement of mechanical properties (such as toughness and tensile strength), particle-filled composite materials are widely used in aerospace, chemical and petroleum, and construction fields. Polymer-bonded explosives (PBX) are a typical particle-filled composite material composed of highly filled (over 90%) elemental explosive crystals and a small amount of binder, possessing high mechanical properties and good processing and molding performance, and are widely used in domestic and international weaponry. Under environmental factors, PBX may develop macroscopic cracks, leading to degradation of the mechanical properties of explosive components and explosive sensitization, affecting system safety and reliability. Therefore, establishing accurate and effective methods for detecting and imaging internal cracks in PBX is crucial for revealing the fracture mechanics behavior of PBX and assessing its structural integrity.

[0003] Ultrasonic phased array detection and imaging methods demonstrate unique advantages and potential in the detection of internal defects. However, the material characteristics of PBX components, such as low sound velocity and strong attenuation of ultrasound waves, result in low signal-to-noise ratio (SNR) of defect echoes, thus affecting the quality of ultrasonic imaging. Furthermore, the curved surface features of PBX components pose a challenge to high SNR imaging of internal defects. Total Focusing Method (TFM) imaging algorithms based on the linear delay and sum (DAS) principle and their derivatives have been proven to achieve high SNR imaging and quantitative evaluation of internal defects. For example, vector TFM can be used to characterize crack direction, and multimode TFM can be used to improve defect characterization capabilities. In addition, Zhang Haiyan et al. proposed a TFM array imaging method with beam directivity function correction, achieving high SNR imaging of wrinkle defects in carbon fiber reinforced composite materials. However, experiments have shown that for complex-shaped workpieces such as PBX particle-filled composite materials, the signal processing algorithm based on the DAS principle has very limited effect on suppressing noise and clutter signals. Not only does the curved surface shape cause distortion of the reconstructed defect morphology, but the strong attenuation characteristics of the internal particle filling structure for ultrasonic waves also result in severe background noise in the image, affecting the identification and quantitative detection of internal defects.

[0004] Unlike imaging algorithms based on the DAS principle, the Delay Multiply And Sum (DMAS) nonlinear beamforming algorithm utilizes the spatial coherence of the received radio frequency signal to improve the signal-to-noise ratio, exhibiting excellent performance in suppressing noise and clutter signals. Matrone et al. studied the application of DMAS in medical ultrasound imaging and further improved its performance by combining it with techniques such as synthetic aperture imaging, plane wave imaging, and multi-line transmission imaging. Luo et al. applied DMAS to ultrasonic plane wave composite imaging of a wedge-shaped two-layer medium and detected internal defects in rails, achieving high signal-to-noise ratio imaging of the defects. Teng et al. applied DMAS to post-processing imaging of Full Matrix Capture (FMC) data, improving the imaging quality of ultrasonic phased array detection. Yu et al. used the DMAS algorithm combined with pseudo-color imaging technology to detect artificially prefabricated transverse through-hole defects in planar PBX materials, obtaining high signal-to-noise ratio imaging results. The proposed Baseband-DMAS (BB-DMAS) nonlinear beamforming algorithm avoids spectral component aliasing and high computational complexity caused by the multiplication of combined radio frequency (RF) signals. This algorithm demodulates the received RF signal and uses spatial coherence of the baseband signal instead of paired RF signal multiplication, achieving high signal-to-noise ratio (SNR) ultrasonic images while significantly reducing the computational load of the imaging algorithm. This characteristic makes BB-DMAS a promising candidate for high SNR ultrasonic imaging of internal defects in particle-filled composite materials. Summary of the Invention

[0005] This invention addresses the problem of internal defect detection in curved, particle-filled composite materials (PBX). It proposes a high signal-to-noise ratio (SNR) ultrasonic imaging method for internal material defects by combining the BB-DMAS nonlinear beamforming method with synthetic focusing imaging principles. The method is based on a Baseband-Nonlinear Synthetic Focusing (BB-NSF) imaging algorithm model. It effectively suppresses noise and clutter in the received signal through the spatial coherence of the full matrix data. The algorithm's delay rule is modified based on the configuration characteristics of the test block. High-quality ultrasonic images are obtained by using a flexible transducer array to perform ultrasonic imaging on cracks in unequal-thickness curved PBX test blocks. Furthermore, the influence of the flexible transducer position on the imaging results of PBX cracks with different orientations is simulated and analyzed, identifying the key factors affecting crack defect detection in the signal. Finally, the imaging quality is quantitatively evaluated.

[0006] The present invention achieves the above objectives through the following technical solutions:

[0007] A high signal-to-noise ratio ultrasonic imaging method for internal defects in materials includes the following steps:

[0008] Step 1: Process the received FMC raw signal using a denoising method. The received FMC raw signal is processed by using Hilbert transform to filter the signal s. i (t)(i=1,2,3,…,N) is decomposed into its in-phase components and quadrature components;

[0009] Step 2: Calculate the delay time of the transmit-receive group signal based on the virtual focus point (x,z), and extract the baseband signal amplitude of the corresponding received signal at that position; while keeping the phase unchanged, scale the amplitude to the pth root, and then coherently sum the scaled values ​​to generate a new amplitude A(x,z);

[0010] Step 3: Power the magnitude p to restore its dimension, and obtain the pixel intensity of the final image virtual focus point (x,z) by calculating the magnitudes of the two components in the Hilbert transform.

[0011] Step 4, for a curved surface block with radius R, the angles corresponding to each element and the z-coordinate axis are:

[0012]

[0013] According to the parametric equation of a circle, the x-axis and z-axis coordinates of the flexible transducer array element can be expressed as x' m =Rsinθ m and z' m =R(1-cosθ) m );

[0014] Step 5, for the surface profile of the test block, the flight time after delay correction can be rewritten as:

[0015]

[0016] Substituting formula (5) into (1) yields the BB-NSF imaging algorithm for curved surface components.

[0017] The received signal in the time domain is demodulated into in-phase and quadrature components using the Hilbert transform. The magnitude and phase components are as follows:

[0018]

[0019] Where H[·] represents the Hilbert transform, I i (t) and Q i (t) represent the in-phase and quadrature components, respectively; X i (t) represents the modulus of the two components in the Hilbert transform, f i (t) represents the corresponding phase;

[0020] For the focal point P(x,z) of the imaging region, the flight time of the ultrasonic wave emitted from the t-th array element, passing through the focal point, and being received by the r-th array element is:

[0021]

[0022] Where i represents the i-th received signal in the FMC data, (z t ,x t ) and (z r ,x r The coordinates of the transmitting and receiving elements are r and t, respectively, with the subscript relationship being r = mod(i, M) and t = 1 + (ir) / M; L The longitudinal wave velocity is the speed at which ultrasound propagates in the test block.

[0023] A further approach is to extract the signal amplitude corresponding to the flight time of each transmit-receive group in step 3, and scale it using the p-th root. After coherently summing the scaled amplitudes of each channel, the signal dimension is recovered by raising the power of p; therefore, the BB-NSF algorithm based on FMC data is as follows:

[0024]

[0025] Where I(x,z) represents the intensity of any pixel in the ultrasound image, and the coefficient p represents the spatial coherence between received signals in the FMC data.

[0026] A further approach is that, in step 3, the coefficient p can be an integer or a non-integer. As the value of p increases, more signal coherence is introduced during the BB-NSF imaging process, thereby achieving better clutter suppression for ultrasound imaging.

[0027] Normally, the coefficient p is set to 2.

[0028] The beneficial effects of this invention are as follows:

[0029] 1) By combining the BB-NSF algorithm, the imaging characteristics of the flexible transducer for cracks with different orientations in unequal thickness PBX specimens are different. For defects with large orientation angles and a burial depth of 15mm, the crack tip, crack root and shape features can be fully presented.

[0030] 2) Experiments show that the surface-corrected BB-NSF algorithm can improve the signal-to-noise ratio of the ultrasonic reconstruction image of particle-filled composite material PBX crack defects by more than 10dB by utilizing the spatial coherence of the received signals of each channel in the full matrix data, which significantly improves the detection capability of crack defects while taking into account imaging efficiency.

[0031] 3) The selection of p value directly affects the degree of signal-to-noise ratio improvement and imaging efficiency of the BB-NSF algorithm. The modified BB-NSF algorithm improves the signal-to-noise ratio at the far field position while suppressing the noise of the whole field image, thereby enhancing the image detection capability of far field defects and reducing the possibility of false detection or missed detection. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of the BB-NSF algorithm of the present invention;

[0034] Figure 2 This is a schematic diagram of the synthetic focusing imaging of the present invention;

[0035] Figure 3 This is a schematic diagram illustrating the delay time calculation principle of the post-processing imaging algorithm for curved workpieces in this invention.

[0036] Figure 4 This is a schematic diagram of the crack defect geometric model of the PBX particulate composite material of the present invention;

[0037] Figure 5 The simulation imaging results of defects A, B, and C of the flexible transducer of the present invention at different detection positions are shown.

[0038] Figure 6 (a) TFM image; (b) BB-NSF image: pixel peak value in the defect region and root mean square of pixel intensity in the background noise region; (c) SNR index of TFM and BB-NSF defects;

[0039] Figure 7 Schematic diagram of an experiment on a PBX test block with artificially pre-fabricated crack defects;

[0040] Figure 8 Time-domain signals with T=5 and R=8: (a) Detection location of defect A, (b) defect B, (c) defect C;

[0041] Figure 9 (a) TFM without delay correction; (b) TFM, (c) F-DMAS, (d) BB-NSF algorithm imaging results for defect A; (e) BB-NSF algorithm imaging results for defects A, B and C after delay correction. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0043] In any embodiment, such as Figure 1-3 As shown, the present invention provides a high signal-to-noise ratio ultrasonic imaging method for internal defects in materials, the implementation process of which includes:

[0044] Based on the data acquisition principle of ultrasonic phased array (FMC), assuming that a one-dimensional linear phased array transducer has M array elements, by sequentially exciting the array elements to emit ultrasonic signals, and all array elements simultaneously receiving the echo signals, an echo matrix data of N = M × M can be obtained. Figure 1 This is a schematic diagram of the BB-NSF algorithm proposed in this invention. First, a zero-phase bandpass filter and wavelet denoising method are used to process the received FMC raw signal before the imaging algorithm to reduce the impact of low-frequency and high-frequency noise components and phase distortion caused by the filter on the imaging result. Then, Hilbert transform is used to transform the filtered signal s... i The signal (t) (i = 1, 2, 3, ..., N) is decomposed into its in-phase and quadrature components. Based on the TFM synthetic focusing imaging algorithm, the delay time of the transmit-receive group signal is calculated based on the virtual focal point (x, z), and the baseband signal amplitude of the corresponding received signal at that location is extracted. While maintaining the phase, this amplitude is scaled p-th root, and then the scaled values ​​are coherently summed to generate a new amplitude A(x, z). Finally, the amplitude is raised to the p-th power to restore its dimension, and the pixel intensity of the final image virtual focal point (x, z) can be obtained by calculating the magnitudes of the two components in the Hilbert transform.

[0045] The received signal in the time domain is demodulated into in-phase and quadrature components using the Hilbert transform. The magnitude and phase components can be expressed as follows:

[0046]

[0047] Where H[·] represents the Hilbert transform, I i (t) and Q i (t) represents the in-phase and quadrature components, respectively. X i (t) represents the modulus of the two components in the Hilbert transform, f i (t) represents the corresponding phase. For example... Figure 2 As shown, for the focal point P(x,z) of the imaging region, the flight time of the ultrasonic wave emitted from the t-th array element, passing through the focal point, and being received by the r-th array element can be expressed as:

[0048]

[0049] Where i represents the i-th received signal in the FMC data, (z t ,x t ) and (z r ,x r ) represent the coordinate positions of the transmitting and receiving array elements, respectively, with the subscript relationship being r = mod(i, M) and t = 1 + (ir) / M. L The longitudinal wave velocity is the speed at which ultrasound propagates in the test block.

[0050] Extract the signal amplitude corresponding to the flight time of each transmit-receive group, and scale it using the p-th root. After coherently summing the scaled amplitudes of each channel, the signal dimension is recovered by raising the power of p. Therefore, the BB-NSF algorithm based on FMC data can be written as:

[0051]

[0052] Where I(x,z) represents the intensity of any pixel in the ultrasound image, and the coefficient p can be expressed as the spatial coherence between received signals in the FMC data, and p is not limited to an integer. As the value of p increases, more signal coherence is introduced during BB-NSF imaging, resulting in better clutter suppression for ultrasound imaging. Without loss of generality, this invention sets the value of p to 2.

[0053] Unlike imaging algorithms for planar workpieces, when a flexible ultrasonic array is directly coupled to the surface of a curved workpiece, the geometric coordinates of each array element will change, and the time for spatial pixels to receive ultrasonic waves will also change. Therefore, it is necessary to modify the delay rule to achieve accurate mapping of the imaging space and image reconstruction. Figure 3 The diagram shows the time-of-flight of ultrasonic signals from a flexible transducer array on a curved workpiece. A two-dimensional rectangular coordinate system Oxz is established, with the origin O set at the center of the flexible transducer array. The x-axis is along the tangent to the right of the surface contour of the test block, and the z-axis points to the test area along the normal direction of the surface contour of the test block.

[0054] If the length of the flexible array is not stretched or compressed when coupled to the contoured surface, the effective detection arc length of the transducer can be expressed as L = (M-1)d, where d is the element spacing and M is the number of elements. For a curved surface block with radius R, the angles corresponding to each element and the z-axis can be expressed as:

[0055]

[0056] According to the parametric equation of a circle, the x-axis and z-axis coordinates of the flexible transducer array element can be expressed as x' m=Rsinθ m and z' m =R(1-cosθ) m Considering the surface profile of the test block, the flight time after delay correction can be rewritten as:

[0057]

[0058] Substituting formula (5) into (1), we can obtain the BB-NSF imaging algorithm for curved surface components.

[0059] In any embodiment, such as Figure 4-6 As shown, the simulation analysis process of the high signal-to-noise ratio ultrasonic imaging method for internal defects of materials according to the present invention includes:

[0060] Build in CIVA software Figure 4 The model shown is a PBX workpiece made of particulate composite material with uneven thickness and cracks of different orientations. The outer surface radius R1 is 50 mm, the inner surface radius R2 is 40 mm, and the thickness varies from 6.23 to 26.23 mm at different locations. The binder content in the particulate composite is set to 5%, and the particulate content is 95%. The particulate composite model is simplified to a density of 1.895 g / cm³. 3 An isotropic material with a longitudinal wave velocity of 3010 m / s was used. Three bottom-opening groove defects, A, B, and C, with orientations of 60°, 30°, and 0° respectively, were established on the particulate composite model. The defect length was 5 mm, the width was 0.5 mm, and the orientation was defined as the angle between the crack and the normal to the specimen contour at the crack root. A 16-element flexible linear array transducer was used to directly couple with the outer contour of the model under test and acquire FMC data at a sampling frequency of 33.3 MHz. The center frequency of the flexible transducer was 2.5 MHz, the width and length of the array elements were 3.5 mm and 6 mm respectively, and the center-to-center spacing of the array elements was 1.5 mm.

[0061] Figure 5 The results show the BB-NSF imaging simulation of the flexible transducer at different positions. It can be seen that the proposed algorithm can effectively detect defect A with an orientation of 60° and a burial depth of approximately 11 mm. As the flexible transducer moves from left to right, before the transducer centerline crosses the center of defect A, the crack surface exhibits significant echo amplitude, the defect shape characteristics are fully presented, and the defect type can be effectively identified (see...). Figure 5 (ab). As the transducer centerline gradually crosses the defect location, the crack root and crack tip can still be clearly distinguished, but the crack outline information begins to be partially lost until it disappears completely (see...). Figure 5Therefore, for the detection of crack defects in curved PBX components, the best imaging effect is achieved when the transducer centerline is at a certain angle to the crack defect outline and the transducer center is located at the root of the crack defect, resulting in complete information on the defect and the bottom surface outline of the test block. Figure 5 The image shown in ef shows the imaging results of defects B and C. Due to the small crack orientation, almost all the signals received by the transducer array elements are reflected echo signals from the crack tip. The crack profile echo signal is weak, resulting in the loss of crack profile information, but the crack tip information is preserved and presented.

[0062] To objectively evaluate the improvement of BB-NSF in imaging noise suppression compared to traditional TFM, the signal-to-noise ratio (SNR) of defect imaging was obtained by comparing different imaging algorithms using simulated FMC data. The SNR was calculated as SNR = 20lg[I...]. max / RMS(I b )],(I max To define the pixel peak intensity of the defect region, RMS(I b The root mean square (RMS) index, representing the pixel intensity within a defect-free background noise region, is used to quantitatively evaluate image quality. For example... Figure 6 a and Figure 6 As shown in b, compared to TFM, the BB-NSF proposed in this invention significantly enhances the pixel peak intensity I in the defect region. max The root mean square (RMS) of pixel intensity within a defect-free background noise region. b Only a slight increase was observed. (For example...) Figure 6 As shown in Figure c, the signal-to-noise ratio (SNR) of traditional TFM imaging is in the range of 34dB to 52dB. The BB-NSF proposed in this invention has an SNR of 76dB to 114dB. The SNR of defects A, B, and C is improved by approximately 42dB, 69dB, and 61dB, respectively, demonstrating a significant improvement effect. This reflects the advantages of the BB-NSF algorithm in suppressing clutter and improving image SNR.

[0063] In one specific embodiment, such as Figure 7-9 As shown, the present invention provides a high signal-to-noise ratio ultrasonic imaging method for internal defects in materials, and the specific implementation process and result analysis include:

[0064] A PBX test block with artificially created pre-existing cracks was selected as the test object. Its external dimensions and the size and location of the defects were the same as those in the CIVA simulation model. The experimental schematic diagram is shown below. Figure 7 As shown. The PBX test specimen was isostatically pressed from 95% HMX crystals and 5% binder, and the inner and outer surfaces and artificially pre-fabricated crack defects were obtained through machining. The calibrated sound velocity of the specimen was 3010 m / s. During the experiment, a flexible PMUT was bent and tightly adhered to the surface of the PBX specimen, and Vaseline was used as the coupling agent.

[0065] The ultrasonic imaging system used in the experiment consisted of a piezoelectric micromachined ultrasonic transducer (PMUT) array, an FPGA-based lower-level FMC data acquisition module, and a host computer (PC). The FPGA data acquisition module could support up to 64 parallel transmit / receive channels. During the experiment, a one-cycle sinusoidal signal was used for excitation at a voltage of 20V. The flexible transducer acquired FMC data at different detection locations via the FPGA and uploaded it to the host computer. The proposed algorithm was then used to post-process the FMC data to obtain the defect reconstruction image. The parameter configuration of the FMC data acquisition system and the location of the crack tip depth for each defect are shown in Table 1.

[0066] Table 1 Experimental parameters and crack depth

[0067] Transducer parameters value Defect crack tip embedment depth / mm value Number of array elements 16 Defect A 16 Array element center distance / mm 1.5 Defect B 23 Center frequency / MHz 2.5 Defect C 24

[0068] Analyze a typical time-domain signal from the FMC data corresponding to defects A, B, and C. The ultrasonic signal transmitted by the 6th element and received by the 8th element is as follows: Figure 8 As shown, large-amplitude structural and electrical noise can be observed, which is due to the complex composition of the echo signal caused by acoustic scattering from the particulate composite material. Furthermore, a large-amplitude initial wave was also found in the extracted signal. It can be predicted that if these signals are used for imaging, there will be severe background noise in defect-free areas, and a near-field blind zone will exist in the near-surface region of the test block, which will easily cause serious interference to defect judgment.

[0069] Based on the results of CIVA simulation, a suitable detection location was selected, and the FMC data of the PBX test block was collected for post-processing imaging. Figure 9 (a) shows the TFM imaging result without surface time-delay correction. It can be seen that not only is there severe background noise in the image, but the crack defect A and the bottom contour of the tested block are also severely distorted, making it impossible to determine the shape and true location of the defect. In contrast, after introducing surface time-delay correction, the bottom contour and defect information of the tested block are presented, but severe background noise still exists in the image, easily leading to misjudgment (see...). Figure 9 b). Figure 9cd represents F-DMAS with surface delay correction and BB-NSF proposed in this invention, respectively. It can be seen that the signal-to-noise ratio (SNR) in both the near-field and defect regions is significantly improved. This is because nonlinear operations are performed on the full matrix data, enhancing the coherent signal (defect echo) and weakening the incoherent signal (noise). Therefore, the pixel intensity at the defect location increases, while the pixel intensity at non-defect locations decreases, resulting in a significant improvement in the overall SNR. Considering the curved contour characteristics of the tested object, the BB-NSF algorithm with surface delay correction is used to reconstruct the images of three crack defects, as shown below. Figure 9 The full-field image shown in e is obtained from the imaging results. The shape and location features of defect A, as well as the bottom and sidewall contours of the tested block, can be obtained. For defects B and C, the background noise of the BB-NSF image is significantly reduced, and the intensity of the crack tip signal of defects B and C is increased. This trend is consistent with... Figure 5 The simulation results show good consistency.

[0070] Table 2 shows the SNR and computation time of different algorithms for imaging defects A, B, and C. Introducing a surface delay correction rule into the imaging algorithms improves the SNR of defects A, B, and C compared to the classic TFM. The TFM based on directivity correction and the TFM based on coherence factor improve the SNR by an average of ~3.65dB and ~6.85dB, respectively, while increasing the computation time for one frame by 0.2s and 0.03s, respectively. The F-DMAS imaging algorithm significantly improves the SNR of defects A, B, and C by an average of ~10.64dB, allowing for clear observation of defect location information. However, the F-DMAS algorithm is extremely time-consuming to acquire one frame, requiring approximately 142.79s, which limits real-time detection requirements. In contrast, the BB-NSF proposed in this invention can dynamically adjust the signal-to-noise ratio of the imaging results by setting the spatial coherence coefficient p between the received signals in the FMC data. When p=1, the SNR of each defect is comparable to that of TFM. As the p value increases, the SNR of the defects improves. When p=2, an SNR comparable to that of F-DMAS can be achieved, while the computation time is only 0.41s, significantly improving the computational burden of the F-DMAS imaging algorithm. Compared to the classic TFM imaging algorithm, the computation time is only 0.12s longer. Therefore, the BB-NSF algorithm demonstrates a significant advantage in computational efficiency while achieving a high signal-to-noise ratio.

[0071] Furthermore, as shown in Table 2, for any algorithm, the signal-to-noise ratio (SNR) decreases significantly with increasing detection depth of defects A, B, and C. In particular, the SNR of the unoptimized TFM imaging drops to as low as 5.02 dB (defect C), at which point the defect signal is completely submerged by background noise. This is mainly due to the strong attenuation characteristics of the PBX material; the echo signal from far-field defects may be submerged in electrical or structural noise, leading to false detections or missed detections. The BB-NSF algorithm proposed in this invention improves the overall SNR of the image, increasing the SNR of the deeper defect C to 15.64 dB, suppressing the influence of background noise, and achieving effective identification of defect information. This demonstrates that the algorithm improves the image detection capability of far-field defects to a certain extent and reduces the possibility of false detections or missed detections.

[0072] Table 2 Performance metrics of imaging algorithms based on experimental data

[0073]

[0074] This invention addresses the problem of high signal-to-noise ratio imaging of crack defects in particle-filled composite materials. It proposes a post-processing imaging algorithm based on baseband nonlinear synthetic focusing (BB-NSF), using a flexible transducer array to perform ultrasonic imaging of cracks in unequal-thickness curved PBX specimens. This yields high-quality ultrasonic images, achieving the following results:

[0075] 1) By combining the BB-NSF algorithm, the imaging characteristics of the flexible transducer for cracks with different orientations in unequal thickness PBX specimens are different. For defects with large orientation angles and a burial depth of 15mm, the crack tip, crack root and shape features can be fully presented.

[0076] 2) Experiments show that the surface-corrected BB-NSF algorithm can improve the signal-to-noise ratio of the ultrasonic reconstruction image of particle-filled composite material PBX crack defects by more than 10dB by utilizing the spatial coherence of the received signals of each channel in the full matrix data, which significantly improves the detection capability of crack defects while taking into account imaging efficiency.

[0077] 3) The selection of p value directly affects the degree of signal-to-noise ratio improvement and imaging efficiency of the BB-NSF algorithm. The modified BB-NSF algorithm improves the signal-to-noise ratio at the far field position while suppressing the noise of the whole field image, thereby enhancing the image detection capability of far field defects and reducing the possibility of false detection or missed detection.

[0078] Further experiments revealed that the method of the present invention is applicable to both homogeneous and heterogeneous materials, and is particularly effective in detecting internal defects (cracks, inclusions, pores, and delamination) in metal or metal matrix composites and two-phase or multi-phase composites (such as particle-filled composites).

[0079] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately. Furthermore, various different embodiments of the present invention can also be arbitrarily combined, as long as they do not violate the spirit of the present invention, they should also be considered as the content disclosed in the present invention.

Claims

1. A high signal-to-noise ratio ultrasonic imaging method for internal defects in materials, characterized in that, Includes the following steps: Step 1: Process the received FMC raw signal using a denoising method, and use Hilbert transform to filter the signal. (i=1, 2, 3, …, N) is decomposed into its in-phase components and quadrature components: (1) in Represents the Hilbert transform. and These represent in-phase and quadrature components, respectively. Let be the modulus of the two components in the Hilbert transform. For the corresponding phase; For the focal point of the imaging region The flight time of the ultrasonic wave emitted from the t-th array element, passing through the focal point, and being received by the r-th array element is: (2) Where i represents the i-th received signal in the FMC data. and These represent the coordinates of the transmitting and receiving array elements, respectively, with the following subscript relationship: and M represents the number of array elements. The longitudinal wave velocity of the ultrasound propagation in the test block; Step 2, based on virtual focus point Calculate the delay time of the transmit-receive group signal and extract the baseband signal amplitude of the corresponding received signal at that location; while keeping the phase unchanged, scale this amplitude by the p-th root, and then coherently sum the scaled values ​​to generate a new amplitude. ; Step 3: Power the amplitude to the p-th power to recover its dimension, and obtain the final virtual focus point of the image by calculating the magnitudes of the two components in the Hilbert transform. Pixel intensity; In step 3, the signal amplitude corresponding to the flight time of each transmit-receive group is extracted, and it is scaled using the p-th root. After coherently summing the amplitudes of each scaled channel, the signal dimension is recovered by power p; therefore, the BB-NSF algorithm based on FMC data is as follows: (3) in denoted as the intensity of any pixel in the ultrasound image, and coefficient p represents the spatial coherence between received signals in the FMC data; Step 4, for a curved surface block with radius R, the angles corresponding to each element and the z-coordinate axis are: (4) According to the parametric equation of a circle, the x-axis and z-axis coordinates of the flexible transducer array element can be expressed as follows: and ; Step 5, for the surface profile of the test block, the flight time after delay correction can be rewritten as: (5) Substituting formula (5) into (3) yields the BB-NSF imaging algorithm for curved surface components.

2. The high signal-to-noise ratio ultrasonic imaging method for internal defects of materials as described in claim 1, characterized in that, In step 3, the coefficient p can be an integer or a non-integer. As the value of p increases, more signal coherence is introduced during the BB-NSF imaging process, thereby achieving better clutter suppression for ultrasound imaging.

3. The high signal-to-noise ratio ultrasonic imaging method for internal defects of materials as described in claim 1, characterized in that, In step 3, the coefficient p is set to 2.

Images