Nondestructive testing method and device for inclusions in titanium alloy bar

Abstract

The invention relates to the technical field of ultrasonic testing, in particular to a nondestructive testing method and device for inclusions in a titanium alloy bar. The method comprises the following steps: synchronously acquiring ultrasonic A scanning signals before and after multi-moment thermal imaging and thermal excitation by applying thermal pulse excitation; calculating a delay cooling factor based on the pixel temperature time sequence change, determining a hot spot existence rate in combination with the spatial distribution of the delay cooling factor in a multi-direction neighborhood, and screening a thermal abnormal region; in the region, a peak redundancy coefficient is determined by using time sequence volatility of ultrasonic signal amplitude, A scanning waveforms before and after thermal excitation are analyzed through window sliding cross-correlation to obtain a distortion degree, the two are fused to obtain an ultrasonic abnormal degree, and whether inclusions exist or not is judged according to the ultrasonic abnormal degree. According to the method, ultrasonic scattering noise caused by a titanium alloy coarse grain structure is effectively avoided, and the reliability of tiny inclusion detection is improved.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic testing technology, specifically to a method and apparatus for non-destructive testing of inclusions inside titanium alloy bars. Background Technology

[0002] Internal non-metallic inclusions (such as oxides and nitrides) are one of the common defects in titanium alloy bars. They may become crack sources during subsequent deformation or stress. Therefore, conducting efficient and reliable non-destructive testing on the bars before they are put into use to screen out and remove materials containing harmful inclusions is an indispensable part of ensuring the safety of the industrial chain.

[0003] In existing technologies, pulse-echo ultrasonic testing is commonly used to determine the presence of inclusions based on the reflected waveform. However, when ultrasonic waves pass through this coarse-grained structure, they generate strong scattering at the grain boundaries, creating high background "structural noise." This makes it difficult to reliably identify tiny inclusions, resulting in both missed detections and false alarms. In other words, it is impossible to reliably and accurately detect tiny internal inclusions against the background of strong structural noise unique to titanium alloys. Summary of the Invention

[0004] To address the technical problem of reliably and accurately detecting minute internal inclusions in titanium alloys under the unique strong structural noise background, this invention provides a non-destructive testing method and apparatus for internal inclusions in titanium alloy bars. The specific technical solution adopted is as follows: This invention proposes a non-destructive testing method for inclusions inside titanium alloy bars, the method comprising: A thermal pulse excitation is applied to the bar to obtain the thermal image at the imaging sampling time after thermal excitation and the signal amplitude of the reflective ultrasonic A-scan waveform at the ultrasonic sampling time before and after thermal excitation, and to determine the temperature value corresponding to each pixel in the thermal image. Based on the temporal change of the temperature value of each pixel after thermal excitation at different imaging sampling times, the delayed cooling factor is determined. Within a preset radius area centered on different pixels, the hot spot presence rate of the pixel in thermal imaging is determined based on the change of the delayed cooling factor between the center and other pixels in different directions. Thermally abnormal areas are screened based on the hot spot presence rate. Within the thermally abnormal region, the peak redundancy coefficient is determined by combining the distribution fluctuation of the signal amplitude of the pixel at different ultrasonic sampling times. Window sliding cross-correlation analysis is then performed between the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion. The degree of ultrasonic abnormality is determined by combining the peak redundancy coefficient and the A-scan waveform distortion. The degree of ultrasonic anomaly is used to determine whether the thermal anomaly area contains abnormal substances.

[0005] Further, determining the delayed cooling factor based on the temporal change of the temperature value of each pixel at different imaging sampling times includes: Determine the peak temperature value and the time interval corresponding to the temperature half-decay after thermal excitation of each pixel; The time interval is standardized to obtain the delayed cooling factor.

[0006] Furthermore, determining the hot spot presence rate of a pixel in thermal imaging based on the changes in the delayed cooling factor between the center of the circle and other pixels in different directions includes: Obtain the delayed cooling factor of pixels along the radius from the center of the circle to the contour in each preset direction, and sort them according to the order from the center of the circle to the contour to obtain the cooling sequence; Calculate the numerical difference of the delayed cooling factor between the center pixel and other pixels in the cooling sequence, and determine the thermal diffusion difference degree of the center pixel in the preset direction; Discreteness analysis was performed on the thermal diffusion differences in all preset directions to obtain the hot spot presence rate.

[0007] Furthermore, the discreteness analysis of the thermal diffusion differences in all preset directions to obtain the hot spot presence rate includes: Calculate the normalized standard deviation of the thermal diffusion difference in all preset directions, and use the difference between constant 1 and the normalized standard deviation as the directional analysis index. The normalized value of the average thermal diffusion difference in all preset directions is used as the numerical analysis index. The mean values ​​of the numerical analysis index and the directional analysis index are calculated to obtain the hot spot presence rate.

[0008] Furthermore, the step of screening thermally anomalous regions based on the presence rate of hot spots includes: Pixels with a hot spot presence rate greater than a preset presence threshold are considered as suspected pixels. The number of other suspected pixels within the eight-neighbor range of a suspected pixel is determined, and suspected pixels with a number greater than a preset number are designated as inclusion abnormal pixels; the inclusion abnormal pixels form a thermally abnormal region.

[0009] Furthermore, determining the peak redundancy coefficient by combining the distribution fluctuations of the signal amplitude of the combined pixels at different ultrasonic sampling times includes: Calculate the mean of the peak values ​​of the signal amplitudes of all pixels in the thermal anomaly region, and standardize the mean of the peak values ​​to obtain the peak amplitude characteristic value of the thermal anomaly region. Calculate the peak value and standard deviation of the signal amplitude of all pixels in the thermal anomaly region, and normalize them as peak discrete feature values; The peak amplitude characteristic value and the peak discrete characteristic value are combined to determine the peak redundancy coefficient.

[0010] Furthermore, the step of performing window sliding cross-correlation analysis on the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion includes: At different ultrasonic sampling times before thermal excitation, the pre-excitation amplitude sequence of the A-scan waveform is obtained based on a window of a preset length; At different ultrasonic sampling times after thermal excitation, the amplitude sequence of the A-scan waveform after excitation is obtained based on a window of a preset length. Based on window sliding with different phase lengths, Pearson correlation analysis is performed by combining the amplitude sequence before and after excitation of each pixel to determine the A-scan waveform distortion degree in the thermal anomaly region.

[0011] Furthermore, the determination of the degree of ultrasound abnormality by combining the peak redundancy coefficient and the A-scan waveform distortion includes: The product of the peak redundancy coefficient and the A-scan waveform distortion degree is calculated and normalized to represent the degree of ultrasound abnormality.

[0012] Furthermore, the determination of whether the thermally abnormal region contains abnormal objects based on the degree of ultrasonic anomaly includes: If the ultrasonic anomaly is greater than a preset anomaly threshold, the thermal anomaly region is determined to contain abnormal material; otherwise, it is determined that no abnormal material is contained.

[0013] On the other hand, a non-destructive testing device for inclusions inside titanium alloy bars is also provided. The device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the method as described in any of the foregoing.

[0014] The present invention has the following beneficial effects: This invention acquires thermal imaging and reflective ultrasonic A-scan signals simultaneously after thermal pulse excitation. First, it determines the delayed cooling factor based on the temporal temperature variation of pixels and calculates the presence rate of hot spots by combining their spatial distribution characteristics in multi-directional neighborhoods. This allows for the screening of thermal anomaly regions less affected by surface interference and more likely to correspond to internal defects. Then, within these regions, the peak redundancy coefficient is determined using the multi-moment distribution fluctuation of the signal amplitude, and the A-scan waveform distortion degree is obtained through window sliding cross-correlation analysis of the A-scan waveforms before and after thermal excitation. These two factors are then combined to form the degree of ultrasonic anomaly, ultimately determining the presence or absence of inclusions. This method effectively avoids the strong ultrasonic scattering noise interference caused by the coarse-grained structure of titanium alloys. Through a thermo-acoustic synergistic mechanism, it shifts the detection focus from the single ultrasonic echo amplitude, which is easily contaminated by noise, to a comprehensive judgment of the dynamic acoustic response under thermal modulation, significantly improving the reliability of identifying tiny internal inclusions in the context of strong structural noise. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the 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.

[0016] Figure 1 This is a flowchart of a non-destructive testing method for inclusions inside titanium alloy bars according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a typical temperature change comparison curve provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of a waveform time curve provided for an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a non-destructive testing method and apparatus for internal inclusions in titanium alloy bars according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following describes in detail, with reference to the accompanying drawings, a specific scheme for a non-destructive testing method for inclusions inside titanium alloy bars provided by the present invention.

[0020] Please see Figure 1 The diagram illustrates a flowchart of a non-destructive testing method for inclusions inside titanium alloy bars according to an embodiment of the present invention. The method includes: S101: Apply thermal pulse excitation to the bar, acquire the thermal image at the imaging sampling time after thermal excitation and the signal amplitude of the reflective ultrasonic A-scan waveform at the ultrasonic sampling time before and after thermal excitation, and determine the temperature value corresponding to the position of each pixel in the thermal image.

[0021] Internal non-metallic inclusions (such as oxides and nitrides) are one of the common defects in titanium alloy bars. They may become crack sources during subsequent deformation or stress. Therefore, conducting efficient and reliable non-destructive testing on the bars before they are put into use to screen out and remove materials containing harmful inclusions is an indispensable part of ensuring the safety of the industrial chain.

[0022] In existing technologies, pulse-echo ultrasonic testing is commonly used to determine the presence of inclusions based on the reflected waveform. However, when ultrasonic waves pass through this coarse-grained structure, they generate strong scattering at the grain boundaries, creating high background "structural noise." This makes it difficult to reliably identify tiny inclusions, resulting in both missed detections and false alarms. In other words, it is impossible to reliably and accurately detect tiny internal inclusions against the background of strong structural noise unique to titanium alloys.

[0023] The basic principle of pulse-echo ultrasonic testing is as follows: a piezoelectric transducer emits high-frequency ultrasonic pulses into the interior of a rod. When the ultrasonic waves propagate in a homogeneous material, they maintain relatively stable acoustic energy. However, when they encounter inclusions with acoustic impedances different from the matrix, some of the sound waves are reflected at the interface and captured by the same probe or receiving probe. By analyzing the arrival time, amplitude, and waveform characteristics of the reflected echo (defect wave), the depth, location, and equivalent size of the inclusion defect can be inferred.

[0024] This invention achieves the acquisition of signal amplitudes for thermal imaging and reflective ultrasound A-scan waveforms through thermal pulse excitation, facilitating subsequent overall numerical analysis by combining the characteristics of these two dimensions.

[0025] Specifically, in this embodiment of the invention, a detection area can be pre-defined on the surface of the rod. Then, a short, uniform heat flux pulse with millisecond-level, controllable energy is applied to the surface of the detection area using a high-energy flash lamp (such as a xenon lamp) array or a laser scanning source. A high-speed infrared thermal imager is used to record the thermal image of the temperature field on the surface of the rod evolving over time after the heat pulse ends. This allows for the acquisition of thermal images at different imaging sampling times.

[0026] Before applying thermal pulse excitation to the rod, the reflective ultrasonic A-scan waveform of the area to be tested is obtained in advance; after applying thermal pulse excitation to the rod, the reflective ultrasonic A-scan waveform at different ultrasonic frame sampling times after thermal excitation is obtained.

[0027] The imaging sampling time and ultrasound sampling time can be preset periodic times. The imaging sampling time includes multiple preset delay time points after the end of the thermal pulse, and the thermal imaging field of view and the ultrasound focusing area are pre-calibrated and aligned in the spatial coordinate system. It should be noted that because the time change dimensions of thermal imaging and A-scan waveforms are different in subsequent analysis, their corresponding imaging sampling times and ultrasound sampling times are also inconsistent. The imaging sampling time for thermal imaging can be, for example, 0.1 seconds, while the A-scan waveform needs to be acquired every 1 microsecond, that is, the ultrasound sampling time is... Sampling is performed once per second. In the specific calculation process, the second is still used as the unit of measurement. However, during the data acquisition process, different imaging sampling times and ultrasound sampling times can be set according to the characteristics of different data.

[0028] In thermal imaging, the temperature value of each pixel can be analyzed in the corresponding thermal imager and recorded as the temperature value of the corresponding position on the surface of the rod.

[0029] S102: Based on the temporal change of the temperature value of each pixel after thermal excitation at different imaging sampling times, determine the delayed cooling factor. Within a preset radius area centered on different pixels, determine the hot spot presence rate of the pixel in thermal imaging based on the change of the delayed cooling factor of the center and other pixels in different directions. Screen thermally abnormal areas based on the hot spot presence rate.

[0030] In the thermal excitation testing of titanium alloy bars, if defects are identified solely based on the temperature at a single moment or a simple temperature difference, it is easy to misjudge shallow interferences such as surface contamination, uneven coupling, or environmental reflection as internal inclusions. This is because these interferences can also cause local temperature anomalies, but their thermal diffusion behavior is fundamentally different from that of deep defects.

[0031] Internal inclusions, due to their lower thermal conductivity than the substrate, hinder heat conduction to the interior, resulting in slower heat dissipation in the surface area directly above them, forming hot spots with delayed cooling and an approximately circular spatial distribution. Surface disturbances (such as oil stains and scratches) usually cause enhanced local heat absorption / dissipation, with hot spots spreading rapidly and having irregular shapes.

[0032] Therefore, in this embodiment of the invention, deep features are extracted from the spatiotemporal evolution of thermal response. By analyzing the cooling time sequence change of the temperature value of each pixel at multiple imaging sampling times, its delayed cooling factor is calculated to quantify the hysteresis degree of local heat change. Furthermore, within a preset radius neighborhood centered on the pixel, the spatial gradient change of the delayed cooling factor in different directions is examined. Based on this, it is determined whether the pixel is located at the center of a hot spot with isotropic and slow diffusion characteristics, and the hot spot presence rate is defined accordingly.

[0033] The delayed cooling factor reflects anomalies in the time dimension, while the hot spot presence rate characterizes the diffusion pattern in the spatial dimension. Together, they constitute an indirect criterion for the defect depth attribute. Finally, based on the hot spot presence rate, thermally anomalous regions that conform to the thermal characteristics of internal defects are selected.

[0034] The preset radius neighborhood can be, for example, a neighborhood region with a radius of 10 pixels, to achieve spatial thermal diffusion analysis.

[0035] Furthermore, in some embodiments of the present invention, the delayed cooling factor is determined based on the temporal change of the temperature value of each pixel at different imaging sampling times, including: determining the time interval corresponding to the peak temperature value and the temperature half-decay after thermal excitation of each pixel; and standardizing the time interval to obtain the delayed cooling factor.

[0036] Among them, see Figure 2 , Figure 2 This is a schematic diagram of a typical temperature change comparison curve provided in one embodiment of the present invention. The half-life effect is represented by half the difference between the decay peak and the room temperature. For example, if the peak temperature is 60 degrees Celsius and the room temperature is 20 degrees Celsius, then the half-life is the state corresponding to a temperature of 40 degrees Celsius, where the peak temperature decreases by 20 degrees Celsius. Because the decay to normal room temperature takes a long time, and the difference in temperature reduction between normal decay and abnormal decay becomes smaller after half-life, it will reduce the reliability of the analysis. Therefore, in this embodiment of the present invention, delayed cooling analysis is performed by using the time interval corresponding to the peak value and the half-life effect.

[0037] The standardization of the time interval primarily involves dimensionless processing, making the delayed cooling factor a dimensionless data point for easier subsequent calculations. Specifically, the standard time required for the surface temperature to stabilize and decay to room temperature can be used as the denominator, and the time interval as the numerator. Standardization is achieved through fractional calculations to obtain the delayed cooling factor. The standard time is the average time required for the surface temperature to decay from its peak value to room temperature, measured by conducting the same thermal excitation experiment on defect-free titanium alloy bars of the same material. This can be, for example, 10 seconds, with no specific limitation. A higher delayed cooling factor indicates a more pronounced abnormal thermal decay effect.

[0038] Thermal anomalies (hot spots) caused by internal inclusions exhibit a nearly isotropic circular diffusion pattern in space, while surface disturbances (such as scratches, oil stains, and uneven coupling) typically manifest as highly directional, irregularly shaped localized thermal disturbances. Relying solely on the magnitude of the single-point delayed cooling factor to determine anomalies fails to distinguish between deep defects and shallow artifacts.

[0039] In this embodiment of the invention, the hot spot presence rate of a pixel in thermal imaging can be determined based on the changes in the delayed cooling factor between the center of the circle and other pixels in different directions. This includes: obtaining the delayed cooling factor of the pixels along the radius from the center of the circle to the contour in each preset direction, and sorting them according to the order from the center of the circle to the contour to obtain a cooling sequence; calculating the numerical difference of the delayed cooling factor between the center pixel and other pixels in the cooling sequence to determine the thermal diffusion difference degree corresponding to the center pixel in the preset direction; and performing a dispersion analysis on the thermal diffusion difference degree in all preset directions to obtain the hot spot presence rate.

[0040] By extracting cooling sequences along multiple preset directions (such as 0°, 45°, 90°…315° ​​with the horizontal direction as the reference) within the neighborhood centered on the pixel, and analyzing their radial variation law, the diffusion consistency of hot spots can be quantified.

[0041] Specifically, the difference in the delayed cooling factor between the center pixel and other pixels in the cooling sequence is calculated to determine the thermal diffusion difference of the center pixel. This can be achieved by calculating the numerical difference in the delayed cooling factor between the center pixel and other pixels in the cooling sequence, averaging all the numerical differences, normalizing them using the maximum and minimum values, and setting the value range to [0,1] to obtain the thermal diffusion difference.

[0042] In the maximum and minimum value normalization process, the maximum and minimum values ​​can be set based on historical large sample data or preset global empirical values ​​measured by calibration test blocks. Adjusting, calibrating or optimizing the maximum and minimum values ​​does not constitute a limitation of this invention.

[0043] It should be noted that the analysis of the difference in delayed cooling factor between the center pixel and other surrounding pixels shows that if the internal structure of the center pixel is different from that of other pixels, it indicates the presence of internal inclusions, which will result in a large difference in thermal diffusion.

[0044] In this embodiment of the invention, the dispersion analysis of the thermal diffusion difference in all preset directions is performed to obtain the hot spot presence rate. The normalized standard deviation of the thermal diffusion difference in all preset directions is calculated, and the difference between the constant 1 and the normalized standard deviation is used as the directional analysis index. The normalized mean of the thermal diffusion difference in all preset directions is used as the numerical analysis index. The mean of the numerical analysis index and the directional analysis index is calculated to obtain the hot spot presence rate.

[0045] Among them, the standard deviation is normalized by the maximum and minimum values, the value range is set to [0,1], and the difference between the constant 1 and the normalized value is used as the direction analysis index. The larger the value of the direction analysis index, the more the center pixel is in an isotropic abnormal region where the center cools the slowest and the surrounding area decays uniformly, and the more likely it is caused by internal inclusions.

[0046] Internal inclusions exhibit isotropic properties, while surface scratches exhibit anisotropic properties. The larger the value of the directional analysis index, the smaller the dispersion, which means the more obvious the isotropic effect and the more likely it is to contain internal inclusions.

[0047] The larger the mean value of thermal diffusion difference, the more likely the internal structure of the center pixel is different from that of other pixels, and the more likely it is to have internal inclusions. Therefore, the mean values ​​of numerical analysis index and directional analysis index are directly calculated to obtain the hot spot presence rate.

[0048] Because ultrasound is susceptible to interference from grain boundary scattering, while thermal fields are not affected by this, but surface artifacts need to be eliminated, the presence rate of hot spots can provide morphological criteria for spatial thermal response, upgrading traditional amplitude-based thermal imaging to intelligent recognition based on spatiotemporal morphological joint features, effectively filtering out anisotropic noise, and significantly improving the specificity of identifying real internal defects.

[0049] Furthermore, in some embodiments of the present invention, screening thermally abnormal regions based on the hot spot presence rate includes: identifying pixels with a hot spot presence rate greater than a preset presence threshold as suspected pixels; determining the number of other suspected pixels within the eight-neighborhood of the suspected pixels, and identifying suspected pixels with a number greater than a preset number as inclusion abnormal pixels; the inclusion abnormal pixels constitute a thermally abnormal region.

[0050] In this embodiment of the invention, the preset existence threshold is a threshold value of the hot spot existence rate. Specifically, the preset existence threshold can be, for example, 0.75, and the preset number can be, for example, 3. That is, pixels with a hot spot existence rate greater than 0.75 are regarded as suspected pixels, thereby realizing the thermal diffusion anomaly analysis of each pixel.

[0051] Since inclusions typically affect a continuous area, by selecting more than 3 suspected pixels within an eight-neighbor range as inclusion anomalous pixels, the influence caused by single data point anomalies is eliminated, resulting in a more accurate and reliable thermal anomaly region.

[0052] S103: Within the thermally abnormal region, the peak redundancy coefficient is determined by combining the distribution fluctuation of the signal amplitude of the pixel at different ultrasonic sampling times. Window sliding cross-correlation analysis is performed between the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion. The degree of ultrasonic abnormality is determined by combining the peak redundancy coefficient and the A-scan waveform distortion.

[0053] In ultrasonic testing of titanium alloy bars, even if the detection range is limited to the thermally abnormal region, it is still difficult to reliably distinguish the coupling fluctuations between real inclusions and residual structural noise by simply relying on the amplitude of the traditional A-scan waveform.

[0054] A dynamic ultrasound response model sensitive to thermal excitation is constructed. Within the thermally abnormal region, on the one hand, the distribution fluctuation of the signal amplitude of the ultrasound channel corresponding to the pixel is analyzed at multiple ultrasound sampling times to quantify its energy anomaly degree and obtain the peak redundancy coefficient; on the other hand, window sliding cross-correlation analysis is performed on the A-scan waveforms before and after thermal excitation to extract the waveform structure changes caused by the nonlinear effect of thermal excitation and obtain the A-scan waveform distortion degree; finally, the two are fused to determine the degree of ultrasound anomaly.

[0055] Furthermore, in some embodiments of the present invention, the peak redundancy coefficient is determined by combining the distribution fluctuation of the signal amplitude of the pixel at different ultrasonic sampling times, including: calculating the mean of the peak maximum values ​​of the signal amplitude of all pixels in the thermal anomaly region, standardizing the mean of the peak maximum values ​​to obtain the peak amplitude characteristic value of the thermal anomaly region; calculating the standard deviation of the peak maximum values ​​of the signal amplitude of all pixels in the thermal anomaly region, normalizing it as the peak discrete characteristic value; and fusing the peak amplitude characteristic value and the peak discrete characteristic value to determine the peak redundancy coefficient.

[0056] After the thermal pulse ends, a focused ultrasound pulse is immediately emitted synchronously to each thermal anomaly region. For each thermal anomaly region, a set of A-scan waveforms is obtained, and the signal contains time axis and amplitude information. See [link to relevant documentation]. Figure 3 , Figure 3 This is a schematic diagram of a waveform time curve provided in one embodiment of the present invention. The A-scan waveform is a waveform curve showing the change of ultrasonic echo amplitude over time. The horizontal axis represents time (µs), and the vertical axis represents the echo amplitude. In areas with normal, homogeneous material, the peak value is stable and close to the reference echo peak amplitude; at depths where strong scatterers (such as defective materials containing inclusions) exist, a larger peak value will exist.

[0057] Therefore, by averaging the maximum peak values ​​of all pixels in the thermal anomaly region and standardizing the average value, the peak amplitude characteristic value is obtained. The larger the peak amplitude characteristic value, the higher the degree of abnormal loss of ultrasonic energy, and the more likely there are strong scattering body inclusions. The thermal anomaly region as a whole has a large peak characteristic, which means that it is more likely to have inclusions.

[0058] First, the average peak value of the bar without inclusions in the A-scan waveform can be determined and used as the reference value. Then, the ratio of the average peak value of all pixels in the thermal anomaly region to the reference value can be calculated, and the ratio can be normalized by the maximum and minimum values ​​to obtain the peak amplitude characteristic value.

[0059] In conventional bar stock, the peak value of the A-scan waveform tends to be stable. However, due to the unstable and irregular state of the inclusions themselves, the peak value of the A-scan waveform containing inclusions exhibits large numerical fluctuations. Analyzing these numerical fluctuations can help determine whether inclusions are present.

[0060] The standard deviation of the peak value of the signal amplitude of all pixels in the thermal anomaly region is calculated and normalized to obtain the peak discrete characteristic value. The larger the value of the peak discrete characteristic value, the larger the numerical fluctuation is, that is, the uneven distribution of echo energy due to the presence of inclusions, and the more likely there are inclusions.

[0061] In summary, the embodiments of the present invention calculate the product of the peak amplitude characteristic value and the peak discrete characteristic value to determine the peak redundancy coefficient. Combining the two dimensions of peak amplitude characteristic and discrete characteristic, the peak redundancy coefficient is obtained, which characterizes the waveform anomaly effect of the thermal anomaly region in the ultrasonic echo.

[0062] In this embodiment of the invention, window sliding cross-correlation analysis means using different sliding windows to perform correlation analysis. In this embodiment of the invention, the Pearson correlation algorithm can be used to achieve this correlation analysis effect.

[0063] Furthermore, in some embodiments of the present invention, a window sliding cross-correlation analysis is performed on the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion degree. This includes: obtaining the pre-excitation amplitude sequence of the A-scan waveform based on a window of a preset length at different ultrasonic sampling times before thermal excitation; obtaining the post-excitation amplitude sequence of the A-scan waveform based on a window of a preset length at different ultrasonic sampling times after thermal excitation; and performing Pearson correlation analysis based on window sliding of different lengths, combining the pre-excitation amplitude sequence and the post-excitation amplitude sequence of each pixel, to determine the A-scan waveform distortion degree in the thermally abnormal region.

[0064] It should be noted that, regardless of whether it is before or after thermal excitation, the A-scan waveform exhibits periodic changes in timing. Therefore, this embodiment of the invention uses a sliding window to perform correlation analysis under different sliding step sizes, thereby determining the distortion degree of the A-scan waveform.

[0065] The preset length can be, for example, 50. That is, at different ultrasonic sampling times before thermal excitation, 50 consecutive A-scan waveform signal amplitudes are obtained in time sequence and arranged in time sequence to obtain the pre-excitation amplitude sequence. Similarly, the post-excitation amplitude sequence is obtained. Since direct correlation analysis will have the effect of phase mismatch, the correlation analysis will be inaccurate.

[0066] Therefore, in this embodiment of the invention, the absolute value of the Pearson correlation coefficient between the amplitude sequence before and after excitation is first calculated. Then, the window of the amplitude sequence before excitation is slid forward according to a step size of one ultrasonic sampling moment to obtain a new amplitude sequence before excitation. The absolute value of the Pearson correlation coefficient between the new amplitude sequence before excitation and the original amplitude sequence after excitation is calculated. This process continues until the time corresponding to the sliding step size exceeds a preset period (e.g., 20 microseconds; if 1 microsecond is considered one ultrasonic sampling moment, then 20 steps are required). The correlation analysis is achieved by finding the maximum value of the absolute value of the Pearson correlation coefficient under all sliding step sizes. The purpose of window sliding is to find the optimal matching phase, thereby determining the maximum value of the absolute value of the Pearson correlation coefficient as the accurate correlation analysis result.

[0067] It should be noted that a larger absolute value of the Pearson correlation coefficient indicates a higher correlation and lower distortion before and after thermal excitation. Therefore, to analyze the distortion, a negative correlation analysis of the absolute value of the Pearson correlation coefficient is needed. This involves calculating the difference between the constant 1 and the maximum absolute value of the Pearson correlation coefficient to obtain the distortion analysis value of each pixel. Then, the mean of the distortion analysis values ​​of all pixels in the thermal anomaly region is calculated to obtain the A-scan waveform distortion degree of the thermal anomaly region. Since real defects (such as cracks and delamination) interfaces are sensitive to thermal stress, they produce significant nonlinear modulation (large waveform changes); while structural noise (grain boundaries) is usually linear scattering and insensitive to thermal stress (small waveform changes). Therefore, a larger A-scan waveform distortion degree corresponds to a higher probability of real internal inclusions.

[0068] Furthermore, in some embodiments of the present invention, the degree of ultrasound abnormality is determined by combining the peak redundancy coefficient and the A-scan waveform distortion, including: calculating the product of the peak redundancy coefficient and the A-scan waveform distortion, and normalizing it as the degree of ultrasound abnormality.

[0069] Since the peak redundancy coefficient characterizes the waveform anomaly effect of the thermally abnormal region in the ultrasonic echo, and the greater the A-scan waveform distortion, the higher the probability of corresponding real internal inclusions, and both are dimensionless parameters with the same value range, the product of the peak redundancy coefficient and the A-scan waveform distortion can be directly calculated. The product value can then be normalized to obtain the degree of ultrasonic anomaly.

[0070] S104: Determine whether the thermally abnormal area contains abnormal substances based on the degree of ultrasonic anomaly.

[0071] Among them, the degree of ultrasonic anomaly characterizes the effect of the A-scan waveform on the abnormality of the waveform itself and the abnormality before and after thermal excitation. That is, the larger the value of the degree of ultrasonic anomaly, the higher the confidence that there are internal inclusions in the thermal anomaly area. The presence of abnormal objects in the thermal anomaly area can be determined based on the value of the degree of ultrasonic anomaly.

[0072] Furthermore, in some embodiments of the present invention, it is determined that the thermal anomaly region with an ultrasonic anomaly degree greater than a preset anomaly threshold contains abnormal objects; otherwise, it is determined that no abnormal objects are contained.

[0073] The preset anomaly threshold is a threshold value for the degree of ultrasonic anomaly. In this embodiment of the invention, the preset anomaly threshold can be, for example, 0.7. That is, when the degree of ultrasonic anomaly is greater than 0.7, the corresponding thermal anomaly region is determined to contain abnormal objects; when the degree of ultrasonic anomaly is less than or equal to 0.7, the corresponding thermal anomaly region is determined to not contain abnormal objects, thereby achieving further screening of thermal anomaly regions. Through this fusion, false signals caused by structural noise in the ultrasonic channel but with no response in the thermal channel can be eliminated; it can also confirm real defects with weak ultrasonic signals but obvious abnormalities in the thermal channel.

[0074] This invention acquires thermal imaging and reflective ultrasonic A-scan signals simultaneously after thermal pulse excitation. First, it determines the delayed cooling factor based on the temporal temperature variation of pixels and calculates the presence rate of hot spots by combining their spatial distribution characteristics in multi-directional neighborhoods. This allows for the screening of thermal anomaly regions less affected by surface interference and more likely to correspond to internal defects. Then, within these regions, the peak redundancy coefficient is determined using the multi-moment distribution fluctuation of the signal amplitude, and the A-scan waveform distortion degree is obtained through window sliding cross-correlation analysis of the A-scan waveforms before and after thermal excitation. These two factors are then combined to form the degree of ultrasonic anomaly, ultimately determining the presence or absence of inclusions. This method effectively avoids the strong ultrasonic scattering noise interference caused by the coarse-grained structure of titanium alloys. Through a thermo-acoustic synergistic mechanism, it shifts the detection focus from the single ultrasonic echo amplitude, which is easily contaminated by noise, to a comprehensive judgment of the dynamic acoustic response under thermal modulation, significantly improving the reliability of identifying tiny internal inclusions in the context of strong structural noise.

[0075] On the other hand, a non-destructive testing device for inclusions inside titanium alloy bars is also provided. The device includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the methods described above.

[0076] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0077] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A non-destructive testing method for inclusions inside titanium alloy bars, characterized in that, The method includes: A thermal pulse excitation is applied to the bar to obtain the thermal image at the imaging sampling time after thermal excitation and the signal amplitude of the reflective ultrasonic A-scan waveform at the ultrasonic sampling time before and after thermal excitation, and to determine the temperature value corresponding to each pixel in the thermal image. Based on the temporal change of the temperature value of each pixel after thermal excitation at different imaging sampling times, the delayed cooling factor is determined. Within a preset radius area centered on different pixels, the hot spot presence rate of the pixel in thermal imaging is determined based on the change of the delayed cooling factor between the center and other pixels in different directions. Thermally abnormal areas are screened based on the hot spot presence rate. Within the thermally abnormal region, the peak redundancy coefficient is determined by combining the distribution fluctuation of the signal amplitude of the pixel at different ultrasonic sampling times. Window sliding cross-correlation analysis is then performed between the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion. The degree of ultrasonic abnormality is determined by combining the peak redundancy coefficient and the A-scan waveform distortion. The degree of ultrasonic anomaly is used to determine whether the thermal anomaly area contains abnormal substances.

2. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The step of determining the delayed cooling factor based on the temporal change of the temperature value of each pixel at different imaging sampling times after thermal excitation includes: Determine the peak temperature value and the time interval corresponding to the temperature half-decay after thermal excitation of each pixel; The time interval is standardized to obtain the delayed cooling factor.

3. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The step of determining the hot spot presence rate of a pixel in thermal imaging based on the change in the delayed cooling factor between the center of the circle and other pixels in different directions includes: Obtain the delayed cooling factor of pixels along the radius from the center of the circle to the contour in each preset direction, and sort them according to the order from the center of the circle to the contour to obtain the cooling sequence; Calculate the numerical difference of the delayed cooling factor between the center pixel and other pixels in the cooling sequence, and determine the thermal diffusion difference degree of the center pixel in the preset direction; Discreteness analysis was performed on the thermal diffusion differences in all preset directions to obtain the hot spot presence rate.

4. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 3, characterized in that, The discrete analysis of the thermal diffusion differences in all preset directions to obtain the hot spot presence rate includes: Calculate the normalized standard deviation of the thermal diffusion difference in all preset directions, and use the difference between constant 1 and the normalized standard deviation as the directional analysis index. The normalized value of the average thermal diffusion difference in all preset directions is used as the numerical analysis index. The mean values ​​of the numerical analysis index and the directional analysis index are calculated to obtain the hot spot presence rate.

5. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The process of screening thermally anomalous regions based on hot spot presence rate includes: Pixels with a hot spot presence rate greater than a preset presence threshold are considered as suspected pixels. The number of other suspected pixels within the eight-neighbor range of a suspected pixel is determined, and suspected pixels with a number greater than a preset number are designated as inclusion abnormal pixels; the inclusion abnormal pixels form a thermally abnormal region.

6. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The determination of the peak redundancy coefficient based on the distribution fluctuations of the signal amplitude of the combined pixels at different ultrasonic sampling times includes: Calculate the mean of the peak values ​​of the signal amplitudes of all pixels in the thermal anomaly region, and standardize the mean of the peak values ​​to obtain the peak amplitude characteristic value of the thermal anomaly region. Calculate the peak value and standard deviation of the signal amplitude of all pixels in the thermal anomaly region, and normalize them as peak discrete feature values; The peak amplitude characteristic value and the peak discrete characteristic value are combined to determine the peak redundancy coefficient.

7. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The step of performing window sliding cross-correlation analysis on the A-scan waveforms before and after thermal excitation to determine the A-scan waveform distortion includes: At different ultrasonic sampling times before thermal excitation, the pre-excitation amplitude sequence of the A-scan waveform is obtained based on a window of a preset length; At different ultrasonic sampling times after thermal excitation, the amplitude sequence of the A-scan waveform after excitation is obtained based on a window of a preset length. Based on window sliding with different phase lengths, Pearson correlation analysis is performed by combining the amplitude sequence before and after excitation of each pixel to determine the A-scan waveform distortion degree in the thermal anomaly region.

8. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The determination of the degree of ultrasound abnormality by combining the peak redundancy coefficient and the A-scan waveform distortion includes: The product of the peak redundancy coefficient and the A-scan waveform distortion degree is calculated and normalized to represent the degree of ultrasound abnormality.

9. The non-destructive testing method for inclusions inside titanium alloy bars as described in claim 1, characterized in that, The method of determining whether a thermally abnormal region contains abnormal substances based on the degree of ultrasonic anomaly includes: If the ultrasonic anomaly is greater than a preset anomaly threshold, the thermal anomaly region is determined to contain abnormal material; otherwise, it is determined that no abnormal material is contained.

10. A non-destructive testing device for inclusions inside titanium alloy bars, the device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 9.

Images