A method, device and medium for stress distribution analysis of a defective pressure pipeline

Abstract

The application discloses a kind of stress distribution analysis methods, equipment and medium of pressure pipeline containing defect, it is related to structural health monitoring technical field, including the multi-source data of acquisition pipeline containing defect and mapping to pipe wall curved surface manifold coordinate system, construct dynamic deduction background stress non-injury reference stress field;Acoustic emission signal purification positioning is extracted to the internal energy release increment corrected by distance attenuation;Crack source is projected to the outer surface of pipeline, and in-plane displacement field and strain increment are calculated based on binocular vision, and energy assimilation coefficient is constructed, preliminary stress intensity factor is calculated based on surface displacement gradient, and it is modified using assimilation coefficient;With modified stress intensity factor as leading term, reconstruct modified stress distribution field, and carry out critical instability evaluation to pipeline.The application eliminates the interference of pipeline curvature, effectively avoids crack driving force underestimation.

Description

Technical Field

[0001] This invention relates to the field of structural health monitoring technology, and in particular to a method, equipment and medium for stress distribution analysis of pressure pipelines with defects. Background Technology

[0002] Pressure pipelines are critical infrastructure in the fields of oil and gas storage and transportation and petrochemicals. They operate under pressure for a long time, and their structural safety is of paramount importance. Under the combined effects of complex working conditions, alternating loads and corrosive environments, the pipeline body is prone to local defects such as cracks. These defects can cause significant local stress concentrations. If they are not monitored and assessed in time, they may lead to pipeline rupture and instability, resulting in catastrophic accidents.

[0003] Existing methods for analyzing stress distribution in defective pressure pipelines still have the following main limitations:

[0004] While existing methods combining digital image correlation technology and acoustic emission technology can complement surface and internal damage information, their fusion mechanism is imperfect. Furthermore, they have not yet quantitatively expressed the specific energy release process during pipeline stress corrosion cracking, nor have they effectively constructed an assimilation correction model between surface strain energy and internal energy release. Therefore, existing methods are unable to accurately assess the crack propagation driving force and predict the propagation trend, affecting the reliability and accuracy of long-term online monitoring of pipelines with defects. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, this invention provides a method, equipment, and medium for stress distribution analysis of pressure pipelines with defects. It solves the problems of existing methods combining digital image correlation technology and acoustic emission technology, which can complement surface and internal damage information, but still have imperfect fusion mechanisms, lack of quantitative characterization of specific energy release processes in pipeline stress corrosion cracking, and failure to effectively construct assimilation and correction models for surface strain energy and internal energy release. These problems lead to the inability to accurately assess crack propagation driving forces and predict propagation trends, thus limiting the reliability and accuracy of long-term online monitoring of pipelines with defects.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a method for analyzing the stress distribution of a pressure pipeline containing defects, comprising,

[0009] Collect multi-source data on the surface of a defective pipeline, establish a pipe wall manifold coordinate system with the pipeline axis as the reference, and construct a damage-free reference stress field;

[0010] Cepstral purification and localization of acoustic emission voltage signals were performed to obtain the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface, and the internal energy release increment corrected for distance attenuation was extracted.

[0011] The three-dimensional coordinates of the crack source are radially projected onto the outer surface of the pipe. Based on the synchronous binocular digital image pair, the in-plane displacement field and strain increment tensor are obtained by iterative optimization in the neighborhood of the projection point, and the energy assimilation coefficient is constructed.

[0012] Calculate the initial stress intensity factor, and then perform a product correction on the initial stress intensity factor based on the energy assimilation coefficient to obtain the corrected stress intensity factor;

[0013] Critical instability assessment of pipelines is conducted based on the modified stress intensity factor.

[0014] The multi-source data includes binocular digital image pairs, pipeline pressure signals, and acoustic emission voltage signals.

[0015] As a preferred embodiment of the stress distribution analysis method for defective pressure pipelines described in this invention, the step of performing cepstral purification and localization on the acoustic emission voltage signal to obtain the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface, and extracting the internal energy release increment corrected for distance attenuation, includes:

[0016] The acoustic emission voltage signal is filtered to obtain a filtered signal. A short-time Fourier transform is then performed on the filtered signal to obtain a complex spectrum, including the amplitude spectrum and the phase spectrum.

[0017] Calculate the power spectrum of the complex spectrum and obtain the output energy of each filter through a logarithmic filter bank;

[0018] Take the natural logarithm of the output energy to obtain the logarithmic energy. Apply a discrete cosine transform to the logarithmic energy and extract the first Q order coefficients to obtain the low-order cepstral coefficients, where Q is the number of low-order cepstral coefficients.

[0019] The low-order cepstral coefficients are inversely transformed to recover the energy of the logarithmic filter bank. The energy of the logarithmic filter bank is then subjected to exponential operation. Combined with the phase spectrum of the complex spectrum, the purified complex spectrum is constructed.

[0020] The purified complex spectrum is subjected to inverse Fourier transform to obtain the purified waveform in the time domain. The channel with the largest signal amplitude among all purified waveforms is selected as the reference channel. The cross-correlation function between the reference channel waveform and the waveforms of each of the other channels is calculated. The time shift corresponding to the global maximum value of the cross-correlation function is selected and defined as the arrival time difference.

[0021] Based on the time difference of arrival, the optimal source location is solved to obtain the estimated three-dimensional coordinates of the crack source, and the propagation distance from the optimal location to each sensor is recorded.

[0022] Set an event trigger threshold, filter out peak voltages greater than the event trigger threshold, define them as acoustic emission events, calculate the square integral of the voltage within the corresponding time window, define it as elastic wave energy, sum the elastic wave energy after correction, and obtain the energy release increment after attenuation correction.

[0023] As a preferred embodiment of the stress distribution analysis method for defective pressure pipelines described in this invention, the method involves: radially projecting the three-dimensional coordinates of the crack source onto the outer surface of the pipeline; iteratively optimizing the in-plane displacement field and strain increment tensor within the neighborhood of the projection point based on synchronized binocular digital image pairs; and constructing an energy assimilation coefficient, including:

[0024] Calculate the radial projection points on the outer surface of the pipe. For each projection point, acquire the calculated principal strain gradient magnitude using a binocular digital image synchronized with the event time. Apply Gaussian smoothing to the principal strain gradient magnitude and identify four-connected regions where the principal strain gradient magnitude exceeds the identification threshold. Define these regions as the analysis window for the current event.

[0025] Within the image pixel region corresponding to the analysis window, the binocular digital image pair before the current event is used as the reference image, and the binocular digital image pair at the current moment is used as the target image. By performing sub-region iterative optimization within the window, the displacement field is obtained, the strain tensor field is calculated, and the surface strain energy increment is calculated based on the stress tensor and strain increment.

[0026] The energy release increment after attenuation correction is combined with the corresponding surface strain energy increment to construct the energy assimilation coefficient, as shown in the formula:

[0027] ;

[0028] in Let be the energy assimilation coefficient at time t. This represents the increment of energy released during crack propagation at time t after distance attenuation correction. Let t be the energy release increment of the acoustic emission event at time t. The background strain energy constant is given.

[0029] As a preferred embodiment of the stress distribution analysis method for defective pressure pipelines described in this invention, the calculation of the preliminary stress intensity factor, based on the energy assimilation coefficient, involves multiplying and correcting the preliminary stress intensity factor to obtain a corrected stress intensity factor, including:

[0030] Along the crack extension line, the relationship between the normal displacement and the radial distance from the crack tip can be rewritten in linear form as follows:

[0031] ;

[0032] ;

[0033] in This represents the normal displacement at a radial distance r from the crack tip. It is a type I stress intensity factor. Here, G is a constant related to the stress state, G is the shear modulus of the material, and E is the Young's modulus of the material. The Poisson's ratio of the material;

[0034] The least squares linear regression was performed on the data points extracted in the neighborhood of the crack tip to obtain the fitting slope.

[0035] The preliminary stress intensity factor is calculated based on the fitted slope, using the following formula:

[0036] ;

[0037] in Let S be the initial stress intensity factor calculated based on the surface displacement field at time t. Let be the measured slope obtained by fitting at time t;

[0038] Based on the energy assimilation coefficient, the initial stress intensity factor is corrected to obtain the corrected stress intensity factor.

[0039] As a preferred embodiment of the stress distribution analysis method for defective pressure pipelines described in this invention, the step of assessing the critical instability of the pipeline based on a modified stress intensity factor includes:

[0040] A fracture threshold is set, and the corrected stress intensity factor is compared with the fracture threshold to assess the critical instability of the pipeline.

[0041] As a preferred embodiment of the stress distribution analysis method for defective pressure pipelines described in this invention, the step of collecting multi-source data from the surface of the defective pipeline, establishing a pipe wall manifold coordinate system with the pipeline axis as the reference, and constructing a damage-free reference stress field includes:

[0042] Using smart devices to collect multi-source data from the surface of defective pipes;

[0043] The aforementioned intelligent devices include binocular cameras, internal pressure and broadband acoustic emission sensors;

[0044] The pixel coordinates of the binocular digital image are converted into the physical three-dimensional coordinates of the pipe wall. Pipe parameters are obtained from the pipe design document, and the curved manifold coordinate system of the pipe wall is defined. The installation position of the broadband acoustic emission sensor is then converted to the curved manifold coordinate system.

[0045] Construct a damage-free reference stress field.

[0046] In a second aspect, the present invention provides an apparatus for stress distribution analysis of a pressure pipeline with defects, comprising a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, implements any step of the stress distribution analysis method for a pressure pipeline with defects as described in the first aspect of the present invention.

[0047] Thirdly, the present invention provides a stress distribution analysis medium for a pressure pipeline with defects, wherein a computer program is stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the stress distribution analysis method for a pressure pipeline with defects as described in the first aspect of the present invention.

[0048] The beneficial effects of this invention are as follows: This invention breaks through the limitations of traditional Euclidean flat space in geometric measurement and acoustic positioning. By mapping the physical three-dimensional coordinates reconstructed by binocular vision and the spatial position of the acoustic emission sensor to the manifold coordinate system of the pipe wall surface, the nonlinear interference of the pipe shell curvature on the reconstruction of the three-dimensional displacement field and the propagation path of the acoustic wave is completely removed from the physical layer, and high-precision alignment at the spatial level is achieved.

[0049] It breaks through the bottleneck of traditional monitoring where local defect characteristics are easily masked by pressurized operation conditions. By introducing instantaneous internal pressure and pipeline geometric parameters in real time to construct a damage-free reference stress field, the uniform background stress is dynamically subtracted analytically in the strain energy area integral stage. This allows the fusion analysis to accurately focus on the tiny geometric work increment and acoustic energy mutation caused by real crack damage, eliminating the distortion in fracture mechanics parameter assessment caused by surface distortion and pressurized interference. Attached Figure Description

[0050] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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.

[0051] Figure 1 This is a flowchart of the stress distribution analysis method for a pressure pipeline with defects, as shown in the embodiment.

[0052] Figure 2 This is a schematic diagram of the stress distribution analysis method for a pressure pipeline with defects in the embodiment. Detailed Implementation

[0053] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0054] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0055] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0056] This embodiment refers to Figure 1 and Figure 2 A method for stress distribution analysis in a pressure pipeline with defects is provided, comprising the following steps:

[0057] S1. Collect multi-source data on the surface of the defective pipeline, establish a manifold coordinate system of the pipe wall surface with the pipeline axis as the reference, and construct a damage-free reference stress field;

[0058] Specifically, multi-source data on the surface of defective pipelines are collected, a pipe wall manifold coordinate system is established with the pipeline axis as the reference, and a damage-free reference stress field is constructed, including:

[0059] Using smart devices, multi-source data is collected on the surface of defective pipes, including binocular digital image pairs, pressure signals inside the pipe, and acoustic emission voltage signals.

[0060] The aforementioned intelligent device integrates a hardware phase-locked loop synchronization generator, providing a unified high-precision clock reference for binocular cameras, internal pressure and broadband acoustic emission sensors, ensuring that the collected multi-source data are synchronized in timestamps;

[0061] For binocular cameras, the stereo vision calibration technique of Zhang Zhengyou's method is used to obtain the intrinsic and extrinsic parameter matrices of the camera. Through the principle of triangulation, the pixel coordinates of the binocular digital image are converted into the physical three-dimensional coordinates of the pipe wall. These coordinates are located in the global coordinate system of the world coordinate system.

[0062] Obtain pipe parameters from the pipe design documents, including the pipe's average inner radius, wall thickness, and precise mounting position parameters of the binocular camera and broadband acoustic emission sensor relative to the pipe axis (e.g., the circumferential mounting angle of the sensor relative to a specified reference generatrix).

[0063] Define a coordinate system for the pipe wall surface with the pipe axis as the Z-axis. The axial coordinate Z is directly taken from the component along the pipe axis in the world coordinate system. The circumferential coordinate is derived from the reference generatrix by calculating the azimuth angle of the physical 3D coordinate points projected onto the ideal cylindrical surface, with values ​​ranging from 0 to... For example, the reference bus is 0;

[0064] Based on the above pipeline parameters and mapping relationships, the physical three-dimensional coordinates of all pipe wall surface points obtained through stereo vision reconstruction are uniformly mapped to the pipe wall curved manifold coordinate system. The installation position of the broadband acoustic emission sensor is transformed to this curved manifold coordinate system. This mapping eliminates the influence of pipeline curvature on geometric measurement and acoustic positioning, ensuring that the subsequent visual displacement field and acoustic emission source positioning results are within the same physical reference frame.

[0065] The complete mapping relationship from the pixel coordinates to the surface manifold coordinate system and from the sensor position to the surface manifold coordinate system is as follows: For any spatial point, its coordinates in the surface coordinate system can be calculated by the following steps: Project the physical three-dimensional coordinates to the nearest cylindrical surface point, calculate the circumferential angle through the four-quadrant arctangent function, and directly take the axial coordinates;

[0066] The standard stress calculation formula for thin-walled cylindrical shells is used to construct a damage-free reference stress field component. This formula originates from the principle of cross-sectional force balance and is a recognized analytical solution in the field of pressure piping design, specifically in GB / T 150 Pressure Vessel or the ASME Boiler and Pressure Vessel Code. This formula is derived through force balance analysis of a unit length of shell. The projected force of the internal pressure acting on the diameter is balanced with the circumferential stress on both sides, thus deriving the circumferential reference stress. The axial force of the internal pressure on the pipe end is balanced with the axial stress, thus deriving the axial reference stress. In a thin-walled cylindrical shell subjected only to uniform internal pressure, due to the axisymmetry of the load and geometry, there is no mechanical mechanism for generating macroscopic shear stress; therefore, the shear reference stress component is zero. The formula is as follows:

[0067] ;

[0068] ;

[0069] ;

[0070] in Let be the circumferential reference stress component at time t, where t is time. Let be the pressure inside the pipe at time t, R be the inner radius of the pipe, and D be the wall thickness of the pipe. Let be the axial reference stress component at time t. Let be the shear reference stress at time t. denoted as circumferential angular coordinates, and z as the z-component of the Cartesian coordinates;

[0071] The reference stress field dynamically subtracts the uniform background stress caused by the normal internal pressure operation of the pipeline, directly serving the subsequent analysis of surface geometric work increment and reconstruction of defect-induced stress distribution. This enables the fusion analysis to focus on the local disturbances caused by defects, adapting to the actual working conditions of pressurized oil, gas and chemical pipelines.

[0072] S2. Perform cepstral purification and localization on the acoustic emission voltage signal to obtain the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface and extract the internal energy release increment after distance attenuation correction.

[0073] Specifically, cepstral purification and localization are performed on the acoustic emission voltage signal to obtain the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface, and the internal energy release increment corrected for distance attenuation is extracted, including:

[0074] A zero-phase digital bandpass filter is used to filter the acoustic emission voltage signal to retain the main ultrasonic frequency components generated by material damage and to initially suppress low-frequency vibration and high-frequency electronic noise, thus obtaining the filtered signal.

[0075] For example, a Butterworth filter can be used as the zero-phase digital bandpass filter mentioned above, with an order of 4-8, in the 100kHz-800kHz frequency band.

[0076] Within the purified frequency band, a set of M center frequencies is constructed, uniformly distributed on a logarithmic scale, forming a triangular bandpass filter. The bandwidth of each filter is designed according to logarithmic frequency intervals, and the center frequencies of adjacent filters are approximately uniformly distributed on a logarithmic scale. A short-time Fourier transform is performed on the filtered signal, for example, with a window length of 2-5 ms, a Hanning window function, and an overlap rate of 50%-75%, to obtain a complex spectrum, including amplitude and phase spectra. The power spectrum of the complex spectrum is calculated and passed through a logarithmic filter bank to obtain the output energy of each filter, where M is the total number of triangular bandpass filter banks, for example, M=24.

[0077] To compress the dynamic range and highlight the spectral envelope features, the natural logarithm of the output energy is taken to obtain the logarithmic energy. A discrete cosine transform is applied to the logarithmic energy, and the first Q-order coefficients are extracted to obtain the low-order cepstral coefficients, where Q is the number of low-order cepstral coefficients retained after the discrete cosine transform, for example, Q=12. The low-order coefficients of the discrete cosine transform mainly reflect the smooth envelope of the signal's logarithmic power spectrum, which is related to the physical mechanism of the damage source. The higher-order coefficients mainly correspond to spectral details and noise.

[0078] An inverse transform is performed on the low-order cepstral coefficients to recover the approximate logarithmic filter bank energy. An exponential operation is then performed on the logarithmic filter bank energy to return a smooth power spectral envelope with a linear scale. Combined with the phase spectrum of the complex spectrum, the purified complex spectrum is constructed, as shown in the formula:

[0079] ;

[0080] in The purified complex spectrum has an amplitude equal to the reconstructed smooth envelope and a phase equal to the phase of the complex spectrum. s is the index of the acoustic emission sensor. To reconstruct the signal power spectrum envelope at frequency f, which is a smoothed power spectrum that filters out rapidly fluctuating noise components, where f is the frequency variable. Let be the phase spectrum of the filtered signal, and j be the imaginary unit;

[0081] Perform an inverse Fourier transform on the purified complex spectrum to obtain the purified waveform in the time domain. Select the waveform with the largest signal amplitude among all purified waveforms as the reference channel. Calculate the cross-correlation function between the reference channel waveform and the waveforms of each of the other channels. The formula is as follows:

[0082] ;

[0083] in The cross-correlation function of the waveforms of reference channel r and channel s, For reference, the purification waveform of sensor r, This is the acoustic emission voltage waveform at time t after purification. For time-shifted variables;

[0084] Choose the time shift corresponding to the global maximum value of the cross-correlation function. This value is the arrival time difference between the acoustic emission signal propagating from the crack source to the sensor and propagating to the reference sensor.

[0085] The three-dimensional coordinates of all sensors on the pipe manifold and the group velocity of the acoustic wave propagating in the pipe wall thickness direction are known. Their values ​​are accurately determined in advance by conducting a pencil lead breakage calibration test on an intact pipe section in the field. For example, using a standard acoustic emission source Hsu-Nielsen source, a 0.5mm diameter pencil lead, broken at a height of 2-3mm from the pipe wall surface, a signal at a known location is excited. The calibration frequency should cover the passband frequency range of 100kHz-800kHz.

[0086] Based on the time difference of arrival, and combining the known sensor locations and the group velocity of the ultrasonic waves in the pipe wall, the crack initiation localization problem is transformed into a nonlinear least squares optimization problem, as shown in the following formula:

[0087] ;

[0088] in Let X be the wave propagation distance from the assumed source location X to the sensor s. Let X be the wave propagation distance from the assumed source location X to the reference sensor r. The time difference between arrival of sensor s and reference sensor r. Let X be the group velocity of the acoustic wave in the pipe material, and let X be the coordinate variable of the crack initiation location to be determined.

[0089] The optimal source location is determined by constructing and minimizing the residual sum of squares function, as shown in the formula:

[0090] ;

[0091] in The optimal source location that minimizes the sum of squared residuals;

[0092] The Levenberg-Marquardt iterative algorithm is used to solve the above nonlinear least squares problem to obtain the estimated three-dimensional coordinates of the crack source and record the propagation distance from the optimal position to each sensor.

[0093] During an initial period after synchronous acquisition begins, such as 0.1 to 0.5 seconds, the pipeline is in a stable state after the pressure boost is completed. At this time, there should be no active damaging acoustic emission events. The acoustic emission voltage signals of all acoustic emission sensors within this period are captured, and the root mean square value of the acoustic emission voltage signals within this period is calculated as the background noise level of the sensor. The arithmetic mean of all background noise levels is taken as the background noise floor of the entire acoustic emission monitoring system.

[0094] Using a linear law, the event trigger threshold is set using the following formula:

[0095] ;

[0096] in H represents the event trigger threshold, where H is the threshold coefficient, for example, a value between 3 and 8. This is the background noise floor;

[0097] When the peak voltage of any purified waveform exceeds the event trigger threshold, it is determined to be a valid acoustic emission event;

[0098] For each acoustic emission event, within its corresponding time window Within this time window, the time frame is synchronized with the relevant acquisition frame interval of the binocular digital image. The start and end times of the acquisition frame are used to calculate the square integral of the voltage for the purified waveform of each sensor channel, which is defined as the elastic wave energy, using the following formula:

[0099] ;

[0100] in Let be the square integral of the voltage captured by the s-th sensor within the event time window, and represent the elastic wave energy received at sensor s. and These are the start and end times of the nth DIC acquisition frame, respectively;

[0101] To accurately represent the energy released at the crack initiation point, it is necessary to correct for energy attenuation during wave propagation. The amplitude attenuation of ultrasonic waves propagating in pipe steel follows an exponential law. ,in Let be the attenuation coefficient of the material, for example, 0.02-0.08 Np / m near the center frequency of 500 kHz. Since energy is proportional to the square of the amplitude, the energy attenuation correction factor is . The elastic wave energy of each sensor is corrected and then summed to obtain the energy release increment after attenuation correction, as shown in the formula:

[0102] ;

[0103] in The value represents the incremental energy release during crack propagation after distance attenuation correction, where N is the total number of acoustic emission sensors. Let be the propagation distance from the crack source to the s-th sensor. The attenuation coefficient of the material for ultrasonic waves is, for example, 0.02-0.08 Np / m near the center frequency of 500 kHz.

[0104] S3. The three-dimensional coordinates of the crack source are radially projected onto the outer surface of the pipe. Based on the synchronous binocular digital image pair, the in-plane displacement field and strain increment tensor are obtained by iterative optimization in the neighborhood of the projection point, and the energy assimilation coefficient is constructed.

[0105] Specifically, the three-dimensional coordinates of the crack source are radially projected onto the outer surface of the pipe. Based on synchronized binocular digital image pairs, the in-plane displacement field and strain increment tensor are iteratively optimized in the neighborhood of the projection point to construct the energy assimilation coefficient, including:

[0106] For each acoustic emission source, the three-dimensional coordinates are defined in the manifold coordinate system of the pipe wall, and the radial projection points on the outer surface of the pipe are calculated.

[0107] In the manifold coordinate system of the pipe wall, the outer surface of the pipe is discretized into a uniform rectangular mesh, for example, the circumferential resolution of the mesh is... With an axial resolution of 1 mm, the radial projection point is defined as the grid node in the discrete surface grid that is closest to the Euclidean distance of the estimated three-dimensional coordinates of the acoustic emission source. The nearest point search can be achieved by traversing the grid to calculate the distance. This mapping transforms the estimated location of internal damage into a clear and unique reference point on the outer surface of the pipe, which can be directly observed by the digital image correlation system.

[0108] For each projection point, the principal strain gradient magnitude calculated from the binocular digital image pair synchronized with the event time is obtained. The principal strain gradient magnitude is defined as the absolute value of the difference between the two principal strains.

[0109] Within the neighborhood of the projection point, Gaussian smoothing is applied to the principal strain gradient magnitude to suppress noise. For example, the standard deviation of the Gaussian kernel is 1 to 2 times the axial resolution of the grid. Four-connected regions where the principal strain gradient magnitude exceeds the identification threshold are identified. The identification threshold is the 95th quantile of the statistical distribution of the principal strain gradient magnitude under the no-damage state. This statistical distribution is calculated based on the gradient value histogram of all effective calculation points in the stable state in the first 10 to 20 frames after the start of synchronous acquisition. A region enclosed by a fixed boundary is extended outward from this connected region as the core. For example, the extension width is 1 to 1.5 times the side length of the digital image correlation calculation sub-region, which is defined as the analysis window of the current event.

[0110] Within the image pixel region corresponding to the analysis window, the binocular digital image pair before the acoustic emission event and when the pipeline is in a stable state is used as the reference image (image before deformation), and the binocular digital image pair at the current moment is used as the target image. By performing sub-region iterative optimization within the window, the two-dimensional in-plane displacement field with sub-pixel accuracy is obtained.

[0111] In the above sub-region iterative optimization, within the analysis window of the reference image, a square reference sub-region of size (2M+1)×(2M+1) pixels is selected with a pre-defined regular grid point (e.g., a spacing of 5-10 pixels) as the center, where M is a positive integer, such as 10 to 20, corresponding to a sub-region size of 21×21 to 41×41 pixels.

[0112] In the target image, a search area of ​​size (2S+1)×(2S+1) pixels is set with the initial projection position of the center point of the reference sub-region as the center, where S is the search radius, such as 20 to 50 pixels. Within this search area, the target sub-region of the same size as the reference sub-region is translated pixel by pixel. For each translation position, the normalized cross-correlation coefficient between the reference sub-region and the current target sub-region is calculated (using the zero-mean normalized cross-correlation criterion). After traversing the entire search area, the center position of the target sub-region corresponding to the maximum value of the correlation coefficient is taken as the integer pixel-level displacement vector of the grid point.

[0113] To achieve sub-pixel accuracy and accommodate possible deformations of the sub-region, it is assumed that the mapping relationship from the reference sub-region to the target sub-region can be described by an affine shape function that includes translation, rotation, stretching, and shearing. The initial values ​​of the shape function parameter vector are determined by the integer pixel displacement obtained in the previous step. A correlation error function, such as the zero-mean normalized sum of squared differences, is defined to measure the grayscale difference between the target sub-region and the reference sub-region after the shape function deformation. The shape function parameters are adjusted through an iterative optimization algorithm (such as the Levenberg-Marquardt algorithm) to minimize the correlation error function.

[0114] Set the iteration convergence condition to reach the maximum number of iterations (e.g., 50 times). When the convergence condition is met, the iteration stops, generating a continuous and high-precision two-dimensional in-plane displacement field covering the entire analysis window.

[0115] Based on the displacement field and according to the Cauchy strain definition in continuum mechanics, the strain tensor field is calculated in cylindrical coordinates. Its components are calculated using the spatial partial derivatives of the displacement field. The partial derivatives are obtained within a regular mesh using a central difference scheme, and at the mesh boundaries, forward or backward differences are used to convert it into an equivalent strain increment field. The formula is as follows:

[0116] ;

[0117] ;

[0118] ;

[0119] ;

[0120] in For an equivalent variable increment field, , and The circumferential normal strain increment, axial normal strain increment, and shear strain increment represent the linear strain change along the pipe circumference, the linear strain change along the pipe axis, and half of the angular change between the circumferential and axial directions, respectively. It is the strain caused by the circumferential displacement gradient. and The displacement vectors are in the circumferential direction. The components of the displacement vector in the z-axis and the components of the displacement vector in the z-axis, and These are the partial derivatives of the circumferential displacement with respect to the circumferential coordinates and the axial displacement with respect to the axial coordinates. and Let be the partial derivatives of the circumferential displacement with respect to the axial coordinates and the partial derivatives of the axial displacement with respect to the circumferential coordinates. This is a geometric addition in the cylindrical surface coordinate system, which reflects how axial displacement causes changes in the circumferential length;

[0121] Calculate the gradient magnitude of the equivalent strain increment field, set a dynamic threshold, which is 0 times the average gradient magnitude (e.g., 0 = 3~5), and mark the connected pixel regions that satisfy the gradient magnitude greater than the dynamic threshold as high strain gradient anomaly regions.

[0122] For each abnormal region, a thinning algorithm (such as the Zhang-Suen parallel thinning algorithm) is used to iteratively delete boundary points until a central skeleton line with a single pixel width is obtained. On the skeleton line, the local curvature of each skeleton point is calculated. The curvature can be approximated by the angle between the vectors formed by the skeleton points before and after the point. Points on the central skeleton line with local curvature greater than the curvature threshold, as well as the two endpoints of the skeleton line, are selected to form a set of potential candidate points for the initial position of the crack tip. For example, the curvature threshold is set to twice the average curvature of the skeleton line.

[0123] Taking each candidate point in the initial location set of crack tips as the center, a circular analysis region with a radius of 3 to 5 times the size of the sub-region is selected. The strain tensor of all points in the region is extracted, and the maximum principal strain direction angle of each point is calculated. Cluster analysis is performed on all maximum principal strain direction angles (e.g., mean-shift clustering). The center direction angle of the maximum cluster is determined as the initial extension line direction of the crack at the candidate point.

[0124] The local segment of the skeleton line containing the candidate point is unfolded onto a two-dimensional plane according to the cylindrical coordinate transformation formula, within a small neighborhood (such as a circular region with a radius of 5-10 pixels) centered on the candidate point:

[0125] Extract the spatial geometric coordinate data of all pixels belonging to the central skeleton line in the neighborhood, perform principal component analysis on these spatial coordinate data, and extract the first principal component vector, which represents the direction with the largest variance of the spatial distribution of the skeleton line in the neighborhood of the point, and is defined as the local geometric direction of the crack (i.e. the direction of the crack extension line).

[0126] Along the direction of the extension line determined by the first principal component, trace forward a certain distance on the skeleton line, fit a straight line, extend the fitted line in the opposite direction to the candidate point, calculate the orthogonal projection of the candidate point to the fitted line, and use the projection point as the new pole after spatial fine-tuning, so that it falls more accurately on the line, and obtain the coordinates of the crack tip with precise positioning.

[0127] The method for determining the above-mentioned distance is as follows: starting from the current point, move forward along the skeleton line until you encounter any of the following conditions and stop, the cumulative tracking length reaches the preset maximum fitting length (e.g., 20 pixels), and you track to the branch point or endpoint of the skeleton line;

[0128] A local polar coordinate system is established with the precisely located crack tip as the origin and the direction of the crack extension line as the polar axis.

[0129] In the local polar coordinate system, along the direction of the polar angle (equal to 0, i.e., the direction of the crack extension line), starting from the origin at the crack tip, within the interval of radial distance... Radial position points are selected at equal intervals, with the interval set to the step size of the computational grid in the digital image, for example, 0.05 mm. For each point, its displacement vector in the global digital image correlation (DIC) coordinate system is obtained. The displacement vector is projected onto a direction perpendicular to the crack extension line, and the normal displacement component of that point is calculated using the following formula:

[0130] ;

[0131] in The displacement component of the k-th sampling point along the local normal direction of the crack is called the normal displacement. and Let k be the displacement components along the x and y axes of the k-th sampling point in the global digital image correlation (DIC) coordinate system. The angle between the direction of the crack extension line and the positive x-axis of the global coordinate system (determined through principal component analysis). and For the minimum and maximum values ​​of the radial distance, to avoid the near-end plastic singularity region and ensure that the linear elastic approximation holds, for example... =0.5mm, =3mm;

[0132] The meaning of DIC mentioned above: Digital Image Correlation (DIC) is an optical measurement technique that measures the displacement and strain of an object's surface in a full-field, non-contact manner by comparing a series of digital images before and after structural deformation. In this invention, the DIC system is used to obtain the displacement field of the pipe surface.

[0133] The acoustic emission voltage signal is corrected using an exponential decay model to obtain the energy value released at the source during the event, as shown in the formula:

[0134] ;

[0135] in Let be the energy value released at the source by the i-th acoustic emission event. Let be the acoustic emission voltage signal of the i-th event detected by the acoustic emission sensor, and represent the original signal energy. The attenuation coefficient of the acoustic emission signal in the pipe material was determined through experimental calibration. Let be the straight-line propagation distance from the location of the i-th acoustic emission event source to the sensor;

[0136] For all acoustic emission events occurring within the time window, sum the corrected energy values ​​to obtain the internal damage energy release increment corresponding to that time window;

[0137] Based on the stress tensor and strain increment, the surface strain energy increment at time t is calculated by discrete integration, using the following formula:

[0138] ;

[0139] Among them Let be the increment of elastic strain energy in the monitored area on the pipe surface at time t. This represents the total number of computational grids divided within the monitored area. The double dot product operator for tensors represents the sum of the products of corresponding components of two tensors. Let t be the stress tensor of the k-th grid point. Let t be the surface strain increment tensor at the k-th grid point. The material volume represented by the k-th grid point is the surface area of ​​that grid point multiplied by the tube wall deformation thickness;

[0140] The energy release increment after attenuation correction is combined with the corresponding surface strain energy increment to define an energy assimilation coefficient, which is used to quantify the proportional relationship between the energy release from internal damage and the change in observed surface strain energy. The formula is as follows:

[0141] ;

[0142] in The energy assimilation coefficient at time t is used to quantify the proportional relationship between the energy release from internal damage and the change in observed surface strain energy. This represents the increment of energy released during crack propagation at time t after distance attenuation correction. Let t be the energy release increment of the acoustic emission event at time t. The background strain energy constant is defined as the arithmetic mean of the surface strain energy increments during the initial elastic stage of pipeline loading (after confirming no damage). This arithmetic mean is calculated over S consecutive time frames (e.g., S=15).

[0143] S4. Calculate the preliminary stress intensity factor, and perform a product correction on the preliminary stress intensity factor based on the energy assimilation coefficient to obtain the corrected stress intensity factor.

[0144] Specifically, the initial stress intensity factor is calculated, and a product correction is performed on the initial stress intensity factor based on the energy assimilation coefficient to obtain the corrected stress intensity factor, including:

[0145] According to linear elastic fracture mechanics, for a type I crack, the relationship between the normal displacement at the crack tip and the stress intensity factor along the crack extension direction (equal to 0) is given by the following formula:

[0146] ;

[0147] ;

[0148] in This represents the normal displacement at a radial distance r from the crack tip. The Type I (opening) stress intensity factor is a key parameter representing the stress field intensity at the crack tip. For plane strain states, the constant is related to the stress state. For plane stress state G is the shear modulus of the material, and E is the Young's modulus of the material. The Poisson's ratio of the material;

[0149] To facilitate fitting, the above theoretical relationship is rewritten in linear form, as shown in the formula:

[0150] ;

[0151] ;

[0152] in This is the rational proportion coefficient;

[0153] Construct data pairs for linear fitting, calculate the square root of each radial distance to obtain new independent variables, define the theoretical normal displacement as the dependent variable, and obtain the data sequence;

[0154] The formula for least squares linear regression on a data sequence is:

[0155] ;

[0156] Where S is the measured slope obtained by fitting, I is the fitting intercept (used to compensate for possible rigid body displacement), X is the independent variable, and Y is the dependent variable;

[0157] The measured slope obtained from the fitting is regarded as a rational proportionality coefficient. Algebraic operations are performed to calculate the preliminary stress intensity factor based on the surface displacement field. The formula is as follows:

[0158] ;

[0159] in This is the preliminary stress intensity factor calculated based on the surface displacement field at time t;

[0160] Based on the energy assimilation coefficient, the initial stress intensity factor is corrected to obtain the corrected stress intensity factor, as shown in the formula:

[0161] ;

[0162] in is the modified stress intensity factor at time t, and is the final crack driving force parameter that integrates surface displacement and internal acoustic emission information.

[0163] S5. Based on the modified stress intensity factor, conduct a critical instability assessment of the pipeline;

[0164] Set a fracture threshold, for example, 85% of the material’s plane strain fracture toughness, and obtain the plane strain fracture toughness of the pipe material using the ASTM E399 standard test method.

[0165] The corrected stress intensity factor is compared with the fracture threshold. If the corrected stress intensity factor is greater than the fracture threshold, it is marked as an instantaneous over-limit flag; otherwise, it is marked as a stable flag.

[0166] Set a sampling period, for example, 5. If all the sampling periods are instantaneous over-limit indicators, the pipeline is determined to be in a critical unstable state, and an alarm signal is generated and triggered. Otherwise, the pipeline is determined to be in a stable damage stage, and monitoring continues.

[0167] This embodiment also provides an apparatus for stress distribution analysis of pressure pipelines with defects, applicable to the stress distribution analysis method for pressure pipelines with defects, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the stress distribution analysis method for pressure pipelines with defects as proposed in the above embodiment.

[0168] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0169] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the stress distribution analysis method for defective pressure pipelines as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0170] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing stress distribution in a pressure pipeline with defects, characterized in that: include, Collect multi-source data on the surface of a defective pipeline, establish a pipe wall manifold coordinate system with the pipeline axis as the reference, and construct a damage-free reference stress field; The acoustic emission voltage signal is subjected to cepstral purification and localization to obtain the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface. The internal energy release increment corrected for distance attenuation is extracted using the following formula: ； in The value represents the incremental energy release during crack propagation after distance attenuation correction, where N is the total number of acoustic emission sensors. Let be the square integral of the voltage captured by the s-th sensor within the event time window. Let be the propagation distance from the crack source to the s-th sensor. The attenuation coefficient of the material for ultrasonic waves; The three-dimensional coordinates of the crack initiation are radially projected onto the outer surface of the pipe. Based on synchronized binocular digital image pairs, the in-plane displacement field and strain increment tensor are obtained through iterative optimization in the neighborhood of the projection point. The energy assimilation coefficient is then constructed, and the formula is as follows: ； in Let be the energy assimilation coefficient at time t. This represents the increment of energy released during crack propagation at time t after distance attenuation correction. Let be the elastic strain energy increment at time t during the acoustic emission event. The background strain energy constant is defined as the arithmetic mean of the surface strain energy increments calculated over S consecutive time frames during the initial elastic stage of pipeline loading. Calculate the initial stress intensity factor, and then perform a product correction on the initial stress intensity factor based on the energy assimilation coefficient to obtain the corrected stress intensity factor; Critical instability assessment of pipelines is conducted based on the modified stress intensity factor. The multi-source data includes binocular digital image pairs, pipeline pressure signals, and acoustic emission voltage signals.

2. The method for analyzing stress distribution in a defective pressure pipeline as described in claim 1, characterized in that: The cepstral purification and localization of the acoustic emission voltage signal, obtaining the three-dimensional coordinates of the crack source in the manifold coordinate system of the pipe wall surface, and extracting the internal energy release increment corrected for distance attenuation include: The acoustic emission voltage signal is filtered to obtain a filtered signal. A short-time Fourier transform is then performed on the filtered signal to obtain a complex spectrum, including the amplitude spectrum and the phase spectrum. Calculate the power spectrum of the complex spectrum and obtain the output energy of each filter through a logarithmic filter bank; Take the natural logarithm of the output energy to obtain the logarithmic energy. Apply a discrete cosine transform to the logarithmic energy and extract the first Q order coefficients to obtain the low-order cepstral coefficients, where Q is the number of low-order cepstral coefficients. The low-order cepstral coefficients are inversely transformed to recover the energy of the logarithmic filter bank. The energy of the logarithmic filter bank is then subjected to exponential operation. Combined with the phase spectrum of the complex spectrum, the purified complex spectrum is constructed. The purified complex spectrum is subjected to inverse Fourier transform to obtain the purified waveform in the time domain. The channel with the largest signal amplitude among all purified waveforms is selected as the reference channel. The cross-correlation function between the reference channel waveform and the waveforms of each of the other channels is calculated. The time shift corresponding to the global maximum value of the cross-correlation function is selected and defined as the arrival time difference. Based on the time difference of arrival, the optimal source location is solved to obtain the estimated three-dimensional coordinates of the crack source, and the propagation distance from the optimal location to each sensor is recorded. Set an event trigger threshold, filter out peak voltages greater than the event trigger threshold, define them as acoustic emission events, calculate the square integral of the voltage within the corresponding time window, define it as elastic wave energy, sum the elastic wave energy after correction, and obtain the energy release increment after attenuation correction.

3. The method for analyzing stress distribution in a pressure pipeline with defects as described in claim 2, characterized in that: The process involves radially projecting the three-dimensional coordinates of the crack source onto the outer surface of the pipe, and based on synchronized binocular digital image pairs, iteratively optimizing the in-plane displacement field and strain increment tensor within the neighborhood of the projection point to construct an energy assimilation coefficient, including: Calculate the radial projection points on the outer surface of the pipe. For each projection point, acquire the calculated principal strain gradient magnitude using a binocular digital image synchronized with the event time. Apply Gaussian smoothing to the principal strain gradient magnitude and identify four-connected regions where the principal strain gradient magnitude exceeds the identification threshold. Define these regions as the analysis window for the current event. Within the image pixel region corresponding to the analysis window, the binocular digital image pair before the current event is used as the reference image, and the binocular digital image pair at the current moment is used as the target image. By performing sub-region iterative optimization within the window, the displacement field is obtained, the strain tensor field is calculated, and the surface strain energy increment is calculated based on the stress tensor and strain increment. The energy release increment after attenuation correction is combined with the corresponding surface strain energy increment to construct the energy assimilation coefficient.

4. The method for analyzing stress distribution in a defective pressure pipeline as described in claim 3, characterized in that: The calculation of the preliminary stress intensity factor involves multiplying and correcting the preliminary stress intensity factor based on the energy assimilation coefficient to obtain the corrected stress intensity factor, including: Along the crack extension line, the relationship between the normal displacement and the radial distance from the crack tip can be rewritten in linear form as follows: ； ； in This represents the normal displacement at a radial distance r from the crack tip. It is a type I stress intensity factor. Here, G is a constant related to the stress state, G is the shear modulus of the material, and E is the Young's modulus of the material. The Poisson's ratio of the material; The least squares linear regression was performed on the data points extracted in the neighborhood of the crack tip to obtain the fitting slope. The preliminary stress intensity factor is calculated based on the fitted slope, using the following formula: ； in Let S be the initial stress intensity factor calculated based on the surface displacement field at time t. Let be the measured slope obtained by fitting at time t; Based on the energy assimilation coefficient, the initial stress intensity factor is corrected to obtain the corrected stress intensity factor.

5. The method for analyzing stress distribution in a defective pressure pipeline as described in claim 4, characterized in that: The critical instability assessment of the pipeline includes: A fracture threshold is set, and the corrected stress intensity factor is compared with the fracture threshold to assess the critical instability of the pipeline.

6. The method for analyzing stress distribution in a defective pressure pipeline as described in claim 1, characterized in that: The process of collecting multi-source data from the surface of the defective pipeline, establishing a pipe wall manifold coordinate system based on the pipeline axis, and constructing a damage-free reference stress field includes: Using smart devices to collect multi-source data from the surface of defective pipes; The aforementioned intelligent devices include binocular cameras, internal pressure and broadband acoustic emission sensors; The pixel coordinates of the binocular digital image are converted into the physical three-dimensional coordinates of the pipe wall. Pipe parameters are obtained from the pipe design document, and the curved manifold coordinate system of the pipe wall is defined. The installation position of the broadband acoustic emission sensor is then converted to the curved manifold coordinate system. Construct a damage-free reference stress field.

7. An apparatus for analyzing stress distribution in a pressure pipeline with defects, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the stress distribution analysis method for defective pressure pipelines as described in any one of claims 1 to 6.

8. A computer storage medium storing a computer program thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the stress distribution analysis method for defective pressure pipelines as described in any one of claims 1 to 6.

Images