A method and system for digital modeling of downhole threads

Abstract

The application relates to the field of ultrasonic nondestructive testing, and discloses a downhole thread digital modeling method and system. The method comprises the following steps: firstly, wrapping a thread area with an adaptive coupling module; secondly, fixing an ultrasonic phased array probe on the adaptive coupling module, performing a compound motion, and acquiring full matrix data; then, performing imaging processing on the full matrix data to obtain a two-dimensional image; then, extracting a thread three-dimensional feature point cloud based on the two-dimensional image; and finally, reconstructing a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud. The method has the sound beam deflection and focusing capacity of the phased array technology, can better adapt to threads of different specifications and surface states in combination with flexible coupling, has strong universality, and in addition, the method can not only judge whether a defect exists, but also accurately reconstruct a three-dimensional form of the defect, thereby providing unprecedented data support for safety evaluation and service life prediction.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of ultrasonic nondestructive testing, and particularly relates to a downhole thread digital modeling method and system. BACKGROUND

[0002] In existing oil and gas drilling, workover and other operations, the connection of downhole tools (such as drill pipes, casings, tubing, etc.) generally adopts a threaded structure. The processing quality of the thread, the wear, corrosion, cracks and deformation defects after service are directly related to the integrity of the entire wellbore and the safety of the operation.

[0003] With the development of nondestructive testing technology, ultrasonic testing has been tried to be applied to thread testing due to its sensitivity to internal defects and strong penetration. Chinese patent (CN114324598A) discloses a high-quality imaging method and system for bolt ultrasonic testing. The imaging method takes the bottom thread and bolt sidewall echo in the phased array ultrasonic fan-shaped scan image as a reference to determine the approximate position of the bolt sidewall crack defect. After the defect is found, the position of the probe is rotated and moved to adjust the amplitude of the defect and weaken the deformation wave caused by the reflection and refraction of the bolt sidewall. The original radio frequency signal is collected using the phased array ultrasonic fan-shaped scanning method, and the radio frequency signal is weighted to suppress the interference of the deformation wave on the defect echo. Thus, the traditional ultrasonic thickness measurement or A-scan detection can only provide single-point thickness information or waveform information on a line, and cannot form a comprehensive and intuitive image of the entire thread profile.

[0004] In addition, the detection and evaluation of threads also mainly face the following technical problems: low detection efficiency and strong subjectivity: traditional manual visual detection and drift diameter gauge detection methods are inefficient and seriously dependent on the experience of operators, cannot quantify defect size and are prone to missed detection. Insufficient internal defect detection capability: conventional nondestructive testing methods (such as magnetic powder and penetration) can only detect surface or near-surface defects and are powerless against hidden defects such as fatigue cracks and stress corrosion cracks in the thread root. Low degree of digitization: existing detection means cannot obtain complete three-dimensional geometric profile data of the thread, cannot establish a high-precision digital model, and thus cannot perform accurate stress analysis, life prediction or comparison with the original CAD model. Environmental adaptability challenge: downhole threads are often in complex environments such as residual oil stains, mud and slight rust, and optical detection methods are easily disturbed. SUMMARY

[0005] In view of the above problems, the present application provides a downhole thread digital modeling method and system, which can better adapt to threads of different specifications and surface states and has strong versatility.

[0006] The present application aims to provide a downhole thread digital modeling method, which comprises,

[0007] The adaptive coupling module is wrapped around the threaded region;

[0008] The ultrasonic phased array probe is fixed on the adaptive coupling module, and full matrix data is acquired by making a compound motion;

[0009] The full matrix data is subjected to imaging processing to obtain a two-dimensional image;

[0010] Three-dimensional feature point clouds of the thread are extracted based on the two-dimensional image;

[0011] A three-dimensional digital model of the thread is reconstructed based on the obtained three-dimensional feature point clouds.

[0012] Further, the ultrasonic phased array probe is fixed on the adaptive coupling module, and full matrix data is acquired by making a compound motion, including,

[0013] The mechanical driving device drives the ultrasonic phased array probe to move at a constant speed along the thread axis, while the ultrasonic phased array probe itself rotates at a preset angle;

[0014] At each acquisition position, all elements of the ultrasonic phased array probe sequentially emit ultrasonic waves, and all elements simultaneously receive echo signals to obtain N x N x T full matrix data, where N is the number of elements, and T is the number of time sampling points.

[0015] Further, the full matrix data is subjected to imaging processing to obtain a two-dimensional image, including,

[0016] The motion parameters of the ultrasonic phased array probe are determined: axial: Z coordinate, radial: R coordinate, and circumferential: θ coordinate;

[0017] Based on the determined motion parameters of the ultrasonic phased array probe, an imaging region is acquired, which is a cylindrical space;

[0018] For each pixel point P(x, y, z) in the imaging region, p , θ the echo signals contributed by all acoustic paths to the pixel point are calculated and superimposed to obtain a two-dimensional image.

[0019] Further, for each pixel point P(x, y, z) in the imaging region, p , θ the echo signals contributed by all acoustic paths to the pixel point are calculated and superimposed to obtain a two-dimensional image, including,

[0020] Based on each pixel point P(i, j) and each transmit-receive event (i, j) in the full matrix data, the sound path time T_ij is calculated, where i represents the element i transmitting, and j represents the element j receiving; specifically including,

[0021] calculating the time T_ix of the sound wave propagating from the transmitting element i to the pixel point P;

[0022] calculating the time T_xj of the sound wave propagating from the pixel point P to the receiving element j after being scattered by the pixel point P;

[0023] the total sound path time satisfies: T_ij = T_ix + T_xj;

[0024] in the recorded A-scan signal of the transmitting element i and the receiving element j, finding the signal amplitude A_ij corresponding to the time T_ij, and superimposing the signal amplitudes A_ij obtained by all N×N times of transmitting-receiving events:

[0025]

[0026] the absolute value of the superimposed amplitude I(P) represents the possibility strength of the existence of defects or interfaces at the pixel point P, and the absolute value of the amplitude I(P) is assigned to the pixel point P;

[0027] repeating the above process for each pixel point in the imaging area to obtain a two-dimensional image.

[0028] Further, extracting a three-dimensional feature point cloud of the thread based on the two-dimensional image comprises,

[0029] setting an intensity threshold, and obtaining pixel points in the two-dimensional image with intensity higher than the intensity threshold;

[0030] using an image segmentation algorithm to separate the highlight area composed of the pixel points with intensity higher than the intensity threshold from the background;

[0031] extracting three-dimensional coordinates (x, y, z) of all pixel points with intensity higher than the intensity threshold to form a three-dimensional point cloud. p , θ

[0032] Further, reconstructing a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud comprises,

[0033] constructing an indicator function;

[0034] extracting an isosurface of the indicator function;

[0035] generating a three-dimensional digital model based on the extracted isosurface.

[0036] Further, constructing an indicator function comprises,

[0037] ​Transforming point cloud to vector field: take each three-dimensional feature point cloud data point q and its estimated unit normal vector n(q) as a sample of a vector field V, i.e. V(q)=n(q), the vector field V is an approximation of the gradient field of the ideal indicator function χ at the point cloud position, satisfying:

[0038] V≈▽χ

[0039] Solving Poisson equation:

[0040] ▽ 2 χ=▽V

[0041] Where, ▽ 2 is the Laplace operator, and ▽ is the divergence operator;

[0042] Numerical solution: discretize the space by octree, represent χ by cubic B-spline basis function, convert the Poisson equation into a linear equation system Ax=b, b is each data point, and use the conjugate gradient method to solve iteratively, get the indicator function value χ on the entire space grid node, the indicator function value χ is the indicator function defined on the entire space grid.

[0043] Further, extracting the isosurface of the indicator function includes,

[0044] Traverse the voxel: traverse each cubic unit composed of octree leaf nodes;

[0045] Vertex classification: check the indicator function value of each cubic unit 8 vertices, compare it with the isosurface threshold value, if the indicator function value is greater than or equal to 0, mark the vertex as "inside"; otherwise, mark it as "outside";

[0046] Configuration lookup: according to the marking of each cubic unit 8 vertices, determine the configuration of the isosurface in the current cubic unit through a predefined lookup table;

[0047] Triangle patch generation: according to the determined configuration of the isosurface in the current cubic unit, calculate the intersection points of the isosurface and the edge through linear interpolation on the edge, connect the calculated intersection points of the isosurface and the edge into one or more triangular patches as the local surface in the cubic unit;

[0048] Traversal completion: after executing the above steps on all cubic units, extract the isosurface of all cubic units as a continuous grid composed of triangular patches.

[0049] Further, based on the extracted isosurface, generating a three-dimensional digital model includes,

[0050] Through preprocessing, the continuous grid composed of triangular patches forms a threaded digital surface, which is a three-dimensional digital model.

[0051] Further, it also includes comparing the three-dimensional digital model with a standard thread CAD model, specifically including,

[0052] Model alignment of the three-dimensional digital model with the standard thread CAD model;

[0053] Identifying the difference area between the three-dimensional digital model and the standard thread CAD model;

[0054] Classifying the identified difference area by type and defect type.

[0055] Further, it also includes automatically calculating and labeling quantitative indicators, including wear amount, corrosion depth, and crack length.

[0056] Another object of the present application is to provide a downhole thread digital modeling system, comprising,

[0057] An ultrasonic phased array probe is used to obtain full matrix data when making a compound motion relative to the threaded part;

[0058] An imaging module is used to image process the full matrix data to obtain a two-dimensional image;

[0059] A point cloud generation module is used to extract a thread three-dimensional feature point cloud based on the two-dimensional image;

[0060] A model establishment module is used to reconstruct a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud.

[0061] Another object of the present application is to provide a downhole thread digital modeling system, comprising an ultrasonic phased array probe, an adaptive coupling module, a mechanical rotary drive device, an ultrasonic emission / reception acquisition card, and an upper computer processing module, wherein,

[0062] The adaptive coupling module is used to wrap the thread area;

[0063] The ultrasonic phased array probe is used to be fixed on the adaptive coupling module and make a compound motion under the driving of the mechanical rotary drive device, and the ultrasonic emission / reception acquisition card obtains full matrix data;

[0064] The upper computer processing module is used to image process the full matrix data to obtain a two-dimensional image, extract a thread three-dimensional feature point cloud based on the two-dimensional image, and reconstruct a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud.

[0065] The method in the application has the sound beam deflection and focusing ability of the phased array technology, combines with the flexible coupling, can better adapt to threads of different specifications and surface states, and has strong versatility; in addition, the method can not only determine whether there is a defect, but also accurately reconstruct the three-dimensional form of the defect, thereby providing unprecedented data support for safety evaluation and life prediction.

[0066] Other features and advantages of the present application will be set forth in the following description, and in part will be apparent from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0068] Figure 1 A downhole thread digital modeling method flowchart in an embodiment of the present application is shown;

[0069] Figure 2 Another downhole thread digital modeling method flowchart in an embodiment of the present application is shown;

[0070] Figure 3 A phased array focusing principle diagram in an embodiment of the present application is shown;

[0071] Figure 4 A downhole thread digital modeling system structure diagram in an embodiment of the present application is shown;

[0072] Figure 5 Another downhole thread digital modeling system structure diagram in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0073] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0074] As Figure 1As shown in the embodiment of the present application, a downhole thread digital modeling method is introduced, which comprises the following steps: first, wrapping the thread area with an adaptive coupling module; second, fixing an ultrasonic phased array probe on the adaptive coupling module, making a compound motion, and acquiring full matrix data; third, performing imaging processing on the full matrix data to obtain a two-dimensional image; fourth, extracting a thread three-dimensional feature point cloud based on the two-dimensional image; and fifth, reconstructing a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud. The beam deflection and focusing ability of the phased array technology, combined with flexible coupling, can better adapt to threads of different specifications and surface states, and has strong versatility. In addition, the above method can not only determine whether there is a defect, but also accurately reconstruct the three-dimensional morphology of the defect, providing unprecedented data support for safety assessment and life prediction.

[0075] As shown in the embodiment of the present application, Figure 2 The method comprises the following steps: first, determining the positions of three probes. The method further comprises the following steps: wrapping the thread area with an adaptive coupling module, fixing an ultrasonic phased array probe on the adaptive coupling module, making a compound motion, and acquiring full matrix data. Further, the adaptive coupling module includes but is not limited to an inflatable air bag or a flexible film filled with coupling agent. Wrapping the thread area with the adaptive coupling module ensures that the ultrasonic wave can be effectively coupled to the irregular thread surface. Then, the ultrasonic phased array probe is fixed on the adaptive coupling module, and a mechanical driving device drives the ultrasonic phased array probe to move uniformly along the thread axis, while the probe itself rotates at a preset angle, which includes but is not limited to 360°. Then, when the ultrasonic phased array probe makes a compound motion relative to the threaded part, the full matrix data is acquired, which includes that at each acquisition position, all elements of the ultrasonic phased array probe emit ultrasonic waves in turn, and all elements receive echo signals at the same time to obtain N × N × T full matrix data, wherein N is the number of elements, and T is the number of time sampling points. The phased array focusing principle is as follows: Figure 3As shown. Full Matrix Acquisition (FMC) principle: At a fixed probe position, N transmission events are performed: First transmission: Array element 1 emits ultrasonic waves, and all N array elements (including element 1 itself) simultaneously receive and record the echo signal (A-scan signal). Second transmission: Array element 2 emits ultrasonic waves, and all N array elements again simultaneously receive and record the echo signal. This continues until the Nth transmission: Array element N emits ultrasonic waves, and all N array elements receive and record. Finally, a three-dimensional data matrix of N (emitting array elements) × N (receiving array elements) × T (number of time sampling points) is obtained, i.e., the full matrix data (FMC). The full matrix data (FMC) contains all possible acoustic path information that the probe can acquire at that position. During the movement and rotation of the ultrasonic phased array probe along the thread, data is acquired using FMC mode, obtaining the most comprehensive information at once. Preferably, laser line scanning or structured light 3D vision technology can also be used for measurement and data acquisition, which offers extremely fast measurement speed and very high accuracy for surface morphology. Furthermore, plane wave imaging or phase-controlled S-scan mode can be used, which has a small data volume and fast processing speed.

[0076] Secondly, the full matrix data is subjected to imaging processing to obtain a two-dimensional image, including:

[0077] First, determine the motion parameters of the ultrasonic phased array probe: axial direction: Z coordinate (mm), radial direction: R coordinate (radius from the axis, mm), circumferential direction: θ coordinate (angle, °).

[0078] Then, the imaging region is acquired, which is a cylindrical space; specifically, the ultrasonic phased array probe moves along the threaded axis (Z-axis) while rotating itself (θ-axis). Therefore, the final imaging region is a cylindrical space. Mesh generation is typically more convenient and accurate in a cylindrical coordinate system.

[0079] For each pixel P within the imaging region p , θ , z), p Indicates radial length. θ Let z represent the angle and z represent the axial length. The echo signals contributed by all acoustic paths to this pixel are calculated and superimposed to obtain a TFM (Total Focusing Method) image, i.e., a two-dimensional image. This includes the following steps:

[0080] For each pixel P and each transmit-receive event (i, j) in the full matrix data (i.e., element i transmits, element j receives):

[0081] 1) Calculate the sound path time:

[0082] Calculate the time T_ix of the sound wave propagating from the transmitting element i to the pixel point P.

[0083] Calculate the time T_xj of the sound wave propagating from the pixel point P to the receiving element j after being scattered.

[0084] Total sound path time T_ij = T_ix + T_xj. Where the sound speed c is known, the time can be obtained by the geometric distance / sound speed.

[0085] 2) Signal extraction and superposition:

[0086] In the recorded A-scan signal of the transmitting element i and the receiving element j, find the signal amplitude A_ij corresponding to the time T_ij.

[0087] Superimpose (sum) the signal amplitudes A_ij calculated by all N×N transmission-reception events.

[0088] (sum for i from 1 to N, and for j from 1 to N).

[0089] Assignment:

[0090] The absolute value (or energy) of the superimposed amplitude I(P) represents the possibility of the presence of defects or interfaces at point P. Assign the absolute value of the amplitude I(P) to the pixel point P.

[0091] Repeat the above process for each pixel point in the imaging grid to obtain a complete TFM image, i.e. a two-dimensional image, which is accurately focused at each point. The two-dimensional image is a high-resolution, high-signal-to-noise ratio two-dimensional cross-sectional image of the entire thread area, which can clearly show the thread profile, any cracks and corrosion pits at the root, etc. The above method greatly improves the signal-to-noise ratio and resolution of the image, and can clearly image small cracks and corrosion, laying the foundation for high-precision three-dimensional modeling.

[0092] Then, based on the two-dimensional image, extract the thread three-dimensional feature point cloud, including,

[0093] Set an intensity threshold to obtain pixel points in the TFM image with intensity higher than the intensity threshold; wherein points with intensity higher than the threshold are considered to be valid defect signals or surface signals

[0094] Use an image segmentation algorithm to separate these highlighted areas (regions composed of pixel points with intensity higher than the intensity threshold) from the background; wherein the image segmentation algorithm includes but is not limited to region growing, level set, and machine learning segmentation.

[0095] Obtain the three-dimensional coordinates of all valid pixel points (i.e. pixel points with intensity higher than the intensity threshold) p ,θ z) are extracted, forming a three-dimensional point cloud. The three-dimensional point cloud describes the geometry of the thread surface and internal defects.

[0096] Finally, a three-dimensional digital model of the thread is reconstructed based on the obtained three-dimensional feature point cloud. Poisson surface reconstruction or rolling ball method algorithm is used for model reconstruction. Specifically, the three-dimensional model reconstruction process preferably adopts Poisson surface reconstruction (Poisson Surface Reconstruction) algorithm. This method can robustly reconstruct a smooth and watertight triangular mesh model from a point cloud with normal vectors, and is suitable for digitization of complex industrial parts such as threads. The core steps are as follows:

[0097] First, an indicator function is constructed, where the indicator function χ is a scalar function defined in the entire three-dimensional space. Its key feature is that the function value is greater than a given threshold (usually 0) inside the thread entity to be reconstructed; outside the entity, the function value is less than the threshold, i.e. χ(p)>0 when point p is inside the thread; χ(p)<0 when point p is outside the thread; and the points that constitute χ(p)=0 form a surface, which is also called the isosurface.

[0098] The specific construction method of the indicator function is as follows:

[0099] First, convert the point cloud into a vector field: consider each three-dimensional feature point cloud data point q and its estimated unit normal vector n(q) as a sample of a vector field V, i.e. V(q)=n(q). This vector field can be regarded as an approximation of the gradient field of the ideal indicator function χ at the point cloud position, i.e. V≈▽χ;

[0100] Then, solve the Poisson equation: the goal is to find a smooth indicator function χ that is closest to the sample vector field, which is converted into solving a classic Poisson equation:

[0101] ▽ 2 χ=▽V

[0102] Where ▽ 2 is the Laplace operator and ▽ is the divergence operator. This process essentially reverses the discrete, possibly noisy point cloud normal vector field to a continuous, smooth indicator function χ by solving a partial differential equation.

[0103] Function value solving: In practical numerical computation, the space containing the point cloud is usually adaptively divided using an Octree structure. Then at each tree node, a cubic B-spline is used as the basis function to represent the indicator function χ. By discretizing the Poisson equation, a linear system of equations Aχ = b (where A is the system matrix and b consists of the divergence of the vector field for each data point) is formed and solved using iterative algorithms such as the conjugate gradient method, ultimately obtaining the indicator function values at all grid nodes in the space.

[0104] Isosurface extraction: After obtaining the indicator function values at the discrete grid nodes, the isosurface is defined as the surface consisting of all points p that satisfy χ(p) = 0. This surface is the reconstructed digitized thread surface.

[0105] Extraction algorithm: This step is usually implemented using the Marching Cubes Algorithm, which includes the following steps:

[0106] First, traverse the voxel: The algorithm traverses each cubic unit (voxel) composed of the leaf nodes of the Octree.

[0107] Second, vertex classification: Check the indicator function values of the 8 vertices of the cube and compare them with the isosurface threshold (0). If the function value is greater than or equal to 0, mark the vertex as "inside"; otherwise, mark it as "outside".

[0108] Then, configuration lookup: According to the inside / outside marking of the 8 vertices, the algorithm determines the topology of the isosurface within the current cubic unit (i.e., how the isosurface will pass through this cube) through a predefined lookup table. Specifically, there are 2 8 = 256 possible configurations, and the topology of the isosurface within the current cubic unit is one of the 256 configurations.

[0109] Then, triangle generation: According to the configuration determined by the lookup table, the exact intersection points of the isosurface and the edges of the cube are calculated by linear interpolation on the edges. Then, these intersection points are connected into one or more triangular patches as the local surface within the cubic unit.

[0110] Finally, traversal completion: After performing the above operations on all cubic units, the entire isosurface is extracted as a continuous mesh composed of a large number of triangular patches, i.e., the isosurfaces of all cubic units are combined into a continuous mesh composed of triangular patches.

[0111] Based on the extracted isosurface, generating a three-dimensional digitized model includes preprocessing the continuous mesh composed of triangular patches to form a digitized surface of the thread, i.e., a three-dimensional digitized model. The preprocessing includes,

[0112] Post-processing of the original triangular mesh extracted by the marching cubes algorithm (i.e., the continuous mesh described above) to optimize model quality:

[0113] Mesh simplification: Reduce the number of triangular facets while maintaining shape accuracy, reducing model complexity.

[0114] Mesh smoothing: Remove minor noise that may be introduced by the discretization calculation, making the surface smoother.

[0115] Output: A watertight, triangular facet-based 3D digital model is generated, which can be exported as STL, OBJ, and other standard 3D file formats for subsequent comparison and analysis.

[0116] In the embodiment of the present application, the method further comprises comparing the three-dimensional digital model with a standard thread CAD model (usually a parameterized feature model, which is constructed based on the standard parameters of the thread (such as nominal diameter, pitch, thread angle, thread count, etc.) through feature modeling), and automatically calculating and labeling the quantitative indicators, including wear, corrosion depth, and crack length. The specific steps of comparing the three-dimensional digital model with the standard thread CAD model include:

[0117] Step 1: Model alignment (registration), align the actual scanned three-dimensional digital model with the standard CAD thread model in space, ensuring that both are compared in the same coordinate system. Wherein, using the Iterative Closest Point algorithm (ICP, Iterative Closest Point) or other point cloud / mesh registration algorithm (i.e., loop pixel points), first perform coarse registration (such as initial alignment based on thread axis, end face, etc. features), then perform fine registration (ICP optimization), so as to ensure that the two models are aligned in the axial (Z), radial (R), and circumferential (θ) directions.

[0118] Step 2: Difference detection (deviation analysis) to identify geometric differences between the three-dimensional digital model and the standard thread CAD model. Wherein, for each triangular facet or point cloud point of the three-dimensional digital model, calculate the shortest distance (signed distance field) to the surface of the standard thread CAD model.

[0119] Set a tolerance range (such as ±0.1mm (millimeter)), the area beyond the range is the abnormal area, such as wear, corrosion, crack, etc. Further, a deviation chromatogram can be generated to visually display the difference distribution, such as blue for concave / wear and red for convex / deposition.

[0120] Step 3: Region segmentation and feature recognition to classify the difference regions by type and defect type, including but not limited to crest, root, and flank. Based on the geometry of the standard thread CAD model, the thread region is divided into: crest region, root region, thread flank, root arc transition zone, and each region is automatically identified using region growing or semantic segmentation algorithms (such as curvature-based, normal direction). Then, difference statistics are performed for each region. Different regions are set for detection, and differences represent thread abnormalities, improving the efficiency of abnormal detection.

[0121] In the embodiments of the present application, the quantitative indicators are used to give the specific situation of thread damage, and the quantitative indicators include wear depth, corrosion depth, and crack length. The calculation method of each quantitative indicator is as follows:

[0122] Wear depth: the depth of material loss on the thread crest or flank due to friction. The calculation method includes:

[0123] First, in the crest or flank region, take multiple sampling points;

[0124] Then, calculate the radial distance difference between each point and the standard thread CAD model;

[0125] Then, take the maximum negative deviation (inward indentation) of all points as the maximum wear depth;

[0126] Finally, calculate the average deviation of the crest or flank region as the average wear depth.

[0127] 2. Corrosion depth: the depth of local pits caused by chemical corrosion; the calculation method includes,

[0128] First, identify the local indentation area obtained by comparing the three-dimensional digital model with the standard thread CAD model, denoted as corrosion pit;

[0129] Then, for each corrosion pit, calculate the height difference between its lowest point and the surrounding uncorroded area;

[0130] Finally, record the maximum corrosion depth and average corrosion depth.

[0131] 3. Crack length: the extension length of the crack on the thread surface or inside; the calculation method includes:

[0132] First, identify the crack using the continuity and fracture features in the point cloud or three-dimensional digital model;

[0133] Second, extract the skeleton of the crack path;

[0134] Then, the total length of the skeleton curve is calculated as the crack length.

[0135] Finally, if it is a surface crack, the crack projection length in the two-dimensional image can be used for verification.

[0136] The embodiments of the present application are not limited to the above-mentioned quantitative indicators, and also include other quantitative indicators, such as thread profile deviation: profile angle, pitch error, etc.; roundness / cylindricity error: reflecting the overall deformation degree of the thread; defect volume: three-dimensional volume calculation of corrosion pit or crack area.

[0137] In the embodiments of the present application, the method further includes final output results, and the forms of the results include but are not limited to:

[0138] Visual report: generate a three-dimensional comparison graph with color mapping;

[0139] Data table: list the quantitative indicators of each area, such as the maximum, minimum, average and standard deviation of the quantitative indicators;

[0140] Automatic labeling: directly label the over-limit area on the model, such as red highlighting the wear-out over-limit part;

[0141] Pass / fail judgment: automatically judge whether the thread is usable according to the preset threshold.

[0142] In the embodiments of the present application, after generating the point cloud, the periodic geometric features of the thread (such as known pitch, profile angle, etc. Parameters) can be used as constraint conditions for model reconstruction, which can further improve the accuracy and robustness of the reconstructed model, especially when part of the data is missing or of poor quality.

[0143] As shown in Figure 4 The embodiments of the present application also introduce a downhole thread digital modeling system capable of performing the above method, the system includes an ultrasonic phased array probe, an imaging module, a point cloud generation module and a model establishment module, wherein the ultrasonic phased array probe is used to obtain full matrix data when the complex motion is made relative to the thread; the imaging module is used to image process the full matrix data to obtain a two-dimensional image; the point cloud generation module is used to extract a three-dimensional feature point cloud of the thread based on the two-dimensional image; and the model establishment module is used to reconstruct a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud.

[0144] As shown in Figure 5As shown, the embodiment of the present application also introduces a downhole thread digital modeling system capable of executing the above method, comprising an ultrasonic phased array probe, an adaptive coupling module, a mechanical rotary drive device, an ultrasonic emission / reception acquisition card, and an upper computer processing module, wherein the adaptive coupling module is used for wrapping the thread area; the ultrasonic phased array probe is used for being fixed on the adaptive coupling module, doing a compound motion under the driving of the mechanical rotary drive device, and acquiring full matrix data by the ultrasonic emission / reception acquisition card; the upper computer processing module is used for imaging processing on the full matrix data, obtaining a two-dimensional image, extracting a thread three-dimensional feature point cloud based on the two-dimensional image, and reconstructing a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud.

[0145] Although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of downhole threaded digitization modeling, the method comprising: The application relates to a method for acquiring a full matrix data of a thread, comprising the following steps: The thread region is wrapped by an adaptive coupling module; An ultrasonic phased array probe is fixed on the adaptive coupling module, and a composite motion is performed, and full matrix data is acquired, The application relates to a method for acquiring a full matrix data of a thread, comprising the following steps: A mechanical driving device drives the ultrasonic phased array probe to move at a constant speed along a thread axis, and the ultrasonic phased array probe rotates at a preset angle; At each acquisition position, all the array elements of the ultrasonic phased array probe emit ultrasonic waves in turn, and all the array elements receive echo signals at the same time, so that full matrix data of N*N*T is obtained, wherein N is the number of array elements, and T is the number of time sampling points; The full matrix data is subjected to imaging processing, so that a two-dimensional image is obtained, comprising the following steps: The motion parameters of the ultrasonic phased array probe are determined, including an axial Z coordinate, a radial R coordinate and a circumferential theta coordinate; Based on the determined motion parameters of the ultrasonic phased array probe, an imaging region is acquired, and the imaging region is a cylindrical space; For each pixel P within the imaging region Rho , Theta The process involves calculating the echo signals contributed by all acoustic paths to the pixel (p, z), and superimposing them to obtain a two-dimensional image. This includes calculating the total acoustic path time T_ij based on each pixel P and each transmit-receive event (i, j) in the full matrix data, where i represents the transmission of array element i and j represents the reception of array element j. Specifically, this includes... The time T_ix that the sound wave propagates from the transmitting array element i to the pixel point P is calculated; The time T_xj that the sound wave propagates from the pixel point P to the receiving array element j after being scattered is calculated; The total sound path time satisfies T_ij = T_ix + T_xj; In the A-scan signal recorded by the array element i emission and the array element j reception, the signal amplitude A_ij corresponding to the time T_ij is found, and the signal amplitudes A_ij obtained by all N*N times of emission-reception events are superimposed: The absolute value of the superimposed amplitude I(P) represents the possibility strength of defects or interfaces existing at the pixel point P, and the absolute value of the amplitude I(P) is assigned to the pixel point P; The above process is repeated for each pixel point in the imaging region, so that a two-dimensional image is obtained; Based on the two-dimensional image, a three-dimensional feature point cloud of the thread is extracted, comprising the following steps: An intensity threshold is set, and the pixel points with intensity higher than the intensity threshold in the two-dimensional image are acquired; An image segmentation algorithm is used to separate the highlight region composed of the pixel points with intensity higher than the intensity threshold from the background; The three-dimensional coordinates of all pixels with an intensity higher than the intensity threshold. Rho , Theta Extracting z) forms a three-dimensional feature point cloud; A three-dimensional digital model of the thread is reconstructed based on the obtained three-dimensional feature point cloud.

2. The method of thread digitizing modeling downhole according to claim 1, characterized in that, Reconstructing a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud comprises the following steps: An indicator function is constructed; An isosurface of the indicator function is extracted; Based on the extracted isosurface, a three-dimensional digital model is generated.

3. The method of thread digitizing modeling downhole according to claim 2, characterized in that, Constructing the indicator function comprises the following steps: The point cloud is converted into a vector field: each three-dimensional feature point cloud data point q and its estimated unit normal vector n(q) are regarded as a sample of a vector field V, that is, V(q)=n(q), the vector field V is an approximation of the gradient field of an ideal indicator function chi at the position of the point cloud, and the following conditions are met: V≈ χ The Poisson equation is solved: 2 χ= V wherein 2 is the Laplacian operator, is the divergence operator; Numerical solution: the space is discretized by an octree, a cubic B-spline basis function is used to represent chi, the Poisson equation is converted into a linear equation group A chi=b, b is each data point, and a conjugate gradient method is used for iterative solution, so that the indicator function value chi of the entire space grid node is obtained, and the indicator function value chi is the indicator function defined on the entire space grid.

4. The method of thread digitizing modeling downhole according to claim 3, characterized in that, Extracting the isosurface of the indicator function comprises the following steps: Traverse the voxel: traverse each cubic unit composed of the octree leaf nodes; Vertex classification: check the indicator function value of each of the 8 vertices of a cube cell, compare it with the isosurface threshold value, if the indicator function value is greater than or equal to 0, mark the vertex as "inside"; otherwise, mark it as "outside"; Configuration lookup: according to the marking of each of the 8 vertices of a cube cell, determine the configuration of the isosurface in the current cube cell through a predefined lookup table; Triangle patch generation: according to the determined configuration of the isosurface in the current cube cell, calculate the intersection points of the isosurface and the edges by linear interpolation on the edges, and connect the calculated intersection points of the isosurface and the edges into one or more triangle patches as the local surface in the cube cell; Traversal completion: after executing the above steps on all cube cells, extract the isosurfaces of all cube cells as a continuous mesh composed of triangle patches.

5. The method of thread digitizing modeling downhole according to claim 4, characterized in that, Based on the extracted isosurface, generating a three-dimensional digitized model includes, forming a threaded digitized surface by preprocessing the continuous mesh composed of triangle patches, which is the three-dimensional digitized model.

6. The method of thread digitization modeling downhole as claimed in any one of claims 1 to 5, wherein, It also includes comparing the three-dimensional digitized model with the standard thread CAD model, specifically including, aligning the three-dimensional digitized model with the standard thread CAD model; identifying the difference area between the three-dimensional digitized model and the standard thread CAD model; classifying the identified difference area by type and defect type.

7. The method of thread digitization modeling downhole as defined in claim 6, wherein, It also includes automatically calculating and labeling quantitative indicators, including wear amount, corrosion depth, and crack length.

8. A downhole threaded digitized modeling system, characterized by, It includes, an ultrasonic phased array probe for acquiring full matrix data when making a compound motion relative to the threaded part, including, a mechanical driving device drives the ultrasonic phased array probe to move uniformly along the thread axis, while the ultrasonic phased array probe itself rotates at a preset angle; at each acquisition position, all elements of the ultrasonic phased array probe emit ultrasonic waves in turn, and all elements receive echo signals at the same time to obtain N × N × T full matrix data, where N is the number of elements, and T is the number of time sampling points; an imaging module for imaging processing the full matrix data to obtain a two-dimensional image, including, determining the motion parameters of the ultrasonic phased array probe: axial: Z coordinate, radial: R coordinate, and circumferential: θ coordinate; based on the determined motion parameters of the ultrasonic phased array probe, an imaging region is obtained, which is a cylindrical space; For each pixel P within the imaging region ρ , θ The process involves calculating the echo signals contributed by all acoustic paths to the pixel (p, z), and superimposing them to obtain a two-dimensional image. This includes calculating the total acoustic path time T_ij based on each pixel P and each transmit-receive event (i, j) in the full matrix data, where i represents the transmission of array element i and j represents the reception of array element j. Specifically, this includes... calculate the time T_ix of sound wave propagation from transmitting element i to pixel point P; calculate the time T_xj of sound wave propagation from pixel point P to receiving element j after scattering; the total sound path time satisfies: T_ij = T_ix + T_xj; in the A-scan signal recorded by element i emission and element j reception, find the signal amplitude A_ij corresponding to time T_ij, and superimpose the signal amplitudes A_ij calculated by all N × N times of transmission-reception events: the absolute value of the superimposed amplitude I(P) represents the possibility strength of the existence of defects or interfaces at pixel point P, and the absolute value of the amplitude I(P) is assigned to pixel point P; repeat the above process for each pixel point in the imaging region to obtain a two-dimensional image; A point cloud generation module is configured to extract a three-dimensional feature point cloud of the thread based on the two-dimensional image, including, An intensity threshold is set, and pixel points with intensity higher than the intensity threshold in the two-dimensional image are obtained; An image segmentation algorithm is used to separate a highlight region composed of the pixel points with intensity higher than the intensity threshold from the background; The three-dimensional coordinates of all pixels with an intensity higher than the intensity threshold. rho , theta Extracting z) forms a three-dimensional feature point cloud; A model establishment module is configured to reconstruct a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud.

9. A downhole threaded digitized modeling system, characterized by, The ultrasonic phased array probe, the adaptive coupling module, the mechanical rotation driving device, the ultrasonic emission / reception acquisition card, and the host computer processing module are included, wherein, The adaptive coupling module is configured to wrap the thread region; The ultrasonic phased array probe is configured to be fixed on the adaptive coupling module and perform a compound motion under the driving of the mechanical rotation driving device, and the ultrasonic emission / reception acquisition card is configured to obtain full matrix data, including, The mechanical driving device drives the ultrasonic phased array probe to move at a uniform speed along the thread axis, while the ultrasonic phased array probe itself rotates by a preset angle; At each acquisition position, all array elements of the ultrasonic phased array probe sequentially emit ultrasonic waves, and all array elements simultaneously receive echo signals to obtain N x N x T full matrix data, wherein N is the number of array elements, and T is the number of time sampling points; The host computer processing module is configured to perform imaging processing on the full matrix data to obtain a two-dimensional image, extract a three-dimensional feature point cloud of the thread based on the two-dimensional image, and reconstruct a three-dimensional digital model of the thread based on the obtained three-dimensional feature point cloud, wherein The imaging processing on the full matrix data to obtain a two-dimensional image includes, Determine the motion parameters of the ultrasonic phased array probe: axial: Z coordinate, radial: R coordinate, and circumferential: theta coordinate; Based on the determined motion parameters of the ultrasonic phased array probe, an imaging region is obtained, which is a cylindrical space; For each pixel P within the imaging region rho , theta The process involves calculating the echo signals contributed by all acoustic paths to the pixel (p, z), and superimposing them to obtain a two-dimensional image. This includes calculating the total acoustic path time T_ij based on each pixel P and each transmit-receive event (i, j) in the full matrix data, where i represents the transmission of array element i and j represents the reception of array element j. Specifically, this includes... Calculate the time T_ix of sound wave propagation from the transmitting array element i to the pixel point P; Calculate the time T_xj of sound wave propagation from the pixel point P to the receiving array element j after scattering; The total sound path time satisfies: T_ij = T_ix + T_xj; In the A-scan signal recorded when the array element i emits and the array element j receives, find the signal amplitude A_ij corresponding to the time T_ij, and superimpose the signal amplitudes A_ij obtained by all N x N times of transmitting-receiving events: The absolute value of the superimposed amplitude I(P) represents the possibility strength of the existence of defects or interfaces at the pixel point P, and the absolute value of the amplitude I(P) is assigned to the pixel point P; Repeat the above process for each pixel point in the imaging region to obtain a two-dimensional image; Based on the two-dimensional image, a three-dimensional feature point cloud of the thread is extracted, including, An intensity threshold is set, and pixel points with intensity higher than the intensity threshold in the two-dimensional image are obtained; An image segmentation algorithm is used to separate a highlight region composed of the pixel points with intensity higher than the intensity threshold from the background; The three-dimensional coordinates of all pixels with an intensity higher than the intensity threshold. rho , theta The z) are extracted to form a three-dimensional feature point cloud.

Images