An accurate finite element modeling method for the cross-section of a multi-layer and multi-material complex pipe and cable

Abstract

An accurate finite element modeling method for the cross-section of a multi-layer and multi-material complex pipe and cable, belonging to the field of offshore oil and gas resource development. First, the cross-section of the pipe and cable to be modeled and analyzed is photographed in a vertical shooting manner to obtain the picture information of the pipe and cable cross-section; and color assignment processing is performed on each component of the pipe and cable cross-section to obtain a color picture. Secondly, after filtering the color picture, the relative position coordinate values of the pixel points at the structural boundary in the Cartesian coordinate system are obtained through a boundary extraction algorithm. Thirdly, the pixel coordinate values stored based on the Cartesian coordinate system are converted into pixel coordinate values in the polar coordinate system, so that the pixel coordinates can be extracted in a clockwise or counterclockwise form; the coordinate values of the pixel points of each color correspond to an Excel file, and four Excel files are obtained. Finally, using the data in the Excel file as the original data, the ABAQUS finite element analysis method is used for modeling. The present invention can accurately extract the geometric information of the structure and has good applicability and accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of offshore oil and gas resource development, and relates to a precise finite element modeling method for the design and analysis of flexible pipe and cable equipment for offshore oil and gas development, which is used to accurately identify the cross-section of submarine cables. Background Art

[0002] With the continuous innovation of deep-sea mining technology, resource development in the marine field has gradually become the main development direction of resource extraction. However, offshore mining faces the problem of long-distance resource transportation, and the requirements for long-distance power transmission technology in the deep and far sea are also constantly increasing. Due to the complex environmental conditions in the deep and far sea, it is necessary to analyze the performance of submarine cables with different configurations to explore the safety performance of the cables.

[0003] In actual engineering, the performance calculation of cables mainly adopts the theoretical formula method and the finite element analysis method. Due to the highly nonlinear problems in the pipe and cable structure, the analysis based on theoretical formulas is relatively difficult. However, the current finite element analysis research on submarine cables and umbilical cables is all carried out based on the ideal geometric configuration. For example, for the functional structures such as copper conductors, armor wires, inner and outer sheaths of dynamic submarine cables, ideal circular configurations are used for modeling. In actual processing, manufacturing, transportation and installation, defects are likely to occur in the structure, and it is very difficult for the structural geometric configuration to meet the ideal design state. Therefore, there is a certain error between the finite element modeling analysis based on the ideal structure and the actual structure, which leads to a deviation between the finite element analysis results under the ideal configuration and the actual experimental values, and it is difficult to guide the analysis and calculation of actual service pipe and cables.

[0004] Based on the above analysis, it is very important to accurately model and analyze the actual marine pipe and cable structure. Pipe and cable structures similar to submarine cables have the following geometric characteristics: First, the pipe and cable structure similar to submarine cables has obvious hierarchical distribution characteristics geometrically: the inner core of the submarine cable structure usually consists of one layer of copper wires wrapped by another layer of copper wires, and the outermost layer is composed of a sheath layer wrapping all the copper wires. Therefore, in the process of image recognition, it is necessary to consider distinguishing the boundary relationship between structural layers; Second, for the same submarine cable, its inner core is usually composed of copper wires of the same material and color arranged closely, and there is mutual extrusion between the copper wires produced in actual production and manufacturing. This means that for the cross-section of the submarine cable obtained by photographing, the boundaries between the copper wires are very unclear. However, in the process of finite element analysis and calculation, it is necessary to model and analyze each copper wire, which poses a high requirement for the recognition algorithm and requires the ability to capture the boundary information of each copper wire; Finally, due to the complex service conditions of structures such as submarine cables, the cross-section of submarine cables usually does not have a regular and clear cross-sectional drawing. Uneven cutting or local shrinkage during the copper wire cutting process will result in poor quality of the cross-sectional picture, which also poses a high requirement for the image recognition algorithm.

[0005] Some scholars have conducted research on image recognition-based modeling. For example, in a finite element modeling method based on the true microstructure of materials in the invention patent CN104063902A, processes such as grayscale processing, binarization processing, threshold segmentation, and finite element modeling are performed on the electron microscope scanning pictures to be processed using Photoshop software to achieve the process of defect scanning modeling in a planar structure. However, this method has certain limitations. The present invention proposes an algorithm process with stronger generality and lower hardware requirements for photographic equipment. The differences from the method of the invention patent CN104063902A are as follows:

[0006] First, for the method proposed in the invention patent CN104063902A, after grayscale processing, binarization processing is performed, and finally, threshold filtering is directly carried out to extract the required structural information in the picture. This operation is not sensitive to the boundary features of defects. That is, when the distance between two small defects is relatively close, using the algorithm of this invention patent will identify a larger defect area instead of two small defect areas with the same true distribution. When dealing with pipe and cable structures such as submarine cables, due to the above characteristics of submarine cables, that is, the copper wires contained in the submarine cable have the same color and material, and it is difficult to have a clear image boundary between the copper wires ( Figure 2 ), in the process of finite element analysis and calculation of the pipe and cable structure, each copper wire needs to be modeled separately as a smallest unit. When using the method proposed in the invention patent CN104063902A, it is impossible to accurately identify the contour of the copper wire. Even for the existing mainstream boundary extraction algorithms, it is very difficult to directly extract accurate boundary parameters for each copper wire in the image. Based on the above problems, the present invention considers introducing simple manual operations. By assigning colors, the information of the image is enhanced, so as to be able to extract accurate geometric parameters for structures with extremely blurred boundaries. This operation can reduce the requirements of the algorithm for image quality, and the structural boundary can be subjectively defined by people, making the algorithm highly flexible and ensuring the accuracy of structural boundary extraction.

[0007] Secondly, for the method proposed in the invention patent CN104063902A, binarized images are used for threshold filtering processing to achieve the process of structural feature extraction. This method essentially divides the defective and non-defective regions in the structure based on two colors (black and white) to achieve the function of boundary extraction. However, when dealing with pipe and cable structures such as submarine cables, due to the hierarchical structure relationship of the structure, the image extraction algorithm not only needs to distinguish the boundaries between structures of the same layer, but also needs to distinguish the structures between different layers. The method proposed in the invention patent CN104063902A cannot distinguish the structures between two layers. Based on the four-color theorem, the present invention uses four colors to divide the structure according to specific rules, which can meet the recognition of all hierarchical structure boundaries and also has excellent applicability to complex non-hierarchical structures.

[0008] Finally, compared with the method proposed in the invention patent CN104063902A, since no image information enhancement processing is performed, it is necessary to identify based on the SEM micrograph images. To ensure the recognition accuracy, there are high requirements for the photos to be recognized. The present invention proposes an image enhancement processing method based on assigning four colors, which reduces the requirements of the algorithm for image quality and has better applicability. Summary of the Invention

[0009] In order to fully consider the influence of structural geometric defects and relative position uncertainties caused by the processing, manufacturing, transportation, installation, service damage, etc. of submarine pipe and cable structures on the overall pipe and cable structure, the present invention is based on the "four-color theorem", uses image recognition technology, extracts geometric features based on submarine pipe and cable cross-section images, identifies cable conductor cross-sections with geometric defects and irregular positions, and then accurately establishes a finite element model through the recognized structural geometric positions and shapes, realizing the accurate establishment of the finite element model. This method can model according to the actual cross-section shape and relative relationship of the pipe and cable, enhance the picture information through color threshold technology, and can reduce the requirements of the algorithm for the graphic quality.

[0010] In order to achieve the above object, the technical solution adopted by the present invention is:

[0011] An accurate finite element modeling method for a multi-layer and multi-material complex pipe and cable cross-section, which is a finite element modeling method based on color threshold filtering image boundary recognition, and includes the following steps:

[0012] (1) Obtain the original pipe and cable cross-section picture information

[0013] Using a camera device, the cross-section of the cable to be modeled and analyzed is photographed in a vertical shooting manner to obtain the image information of the cable cross-section. There are multiple closely arranged copper wires inside the cable. Each copper wire serves as a sub-structure, and all the copper wires form a multi-layer structure. For example, a copper wire at the center of the cable is a single-layer structure, a circle of copper wires surrounding the central copper wire is the second layer structure, and a circle of copper wires surrounding the second layer structure is the third layer structure, and so on.

[0014] Since the cross-sectional dimensions of submarine cables and the like are relatively small, vertical shooting is preferably used to ensure the accuracy of geometric dimensions. The method determines the boundaries of the structure based on pixel points, so the requirements for the shooting device are not high. The current mainstream mobile phone shooting pixels can already meet the requirements. However, the boundary accuracy of the structure will increase with the increase in the distribution of pixel points on the structure boundary. Therefore, obtaining clearer photos helps to improve the recognition accuracy of the method.

[0015] (2) Color assignment processing is performed on each component of the cable cross-section to obtain a color picture. Considering the applicability of the "Four Color Theorem" in submarine cables with an obvious hierarchical structure and in order to obtain the maximum RGB segmentation efficiency, four colors, namely red (255, 0, 0), green (0, 255, 0), blue (0, 0, 255), and gray (125, 125, 125), are used to perform the color assignment operation on the cable structure. Among them, during the color assignment process, it is required that adjacent copper wires have different colors and the same colors do not touch. The assignment operation follows the following rules:

[0016] For cable structures with hierarchical structure characteristics such as submarine cables, only two colors are used for the sub-structures in the same layer, and any sub-structure in this layer has a different color from the adjacent sub-structures. For example, in a certain layer structure, the colors are painted alternately in the order of red, green, red, green; at the same time, it is required that the two colors used in this layer structure are different from the colors used in the upper and lower layer structures. For example, if the first layer is painted alternately in red and green, the second layer should be painted alternately in blue and gray, and the third layer still needs to use red and green colors, and so on.

[0017] The structure color assignment can be carried out in the following two ways:

[0018] The first way is: before shooting in step (1), the real experimental components are painted with four colors, namely red, green, blue, and gray (the differences between the colors are large), and then shooting is carried out;

[0019] The second way is: based on manual judgment, using image processing software such as Photoshop, the cross-sections of the copper wires in the structure are subjected to color assignment processing according to the above painting rules.

[0020] (3) Filter the color picture obtained in step (2) with the three color values of R, G, and B, and then use the boundary extraction algorithm to obtain the pixel points of the boundary of each copper wire in the Cartesian coordinate system, that is, obtain the relative position coordinate values of the pixel points at the structural boundary in the Cartesian coordinate system.

[0021] (4) Convert the pixel coordinate values stored in the Cartesian coordinate system into pixel coordinate values in the polar coordinate system, so that the pixel coordinates can be extracted in a clockwise or counterclockwise form, thereby meeting the subsequent modeling requirements. Specifically as follows:

[0022] 4.1) After obtaining the coordinate information of each sub-structure boundary pixel point in the Cartesian coordinate system, use the method of taking the average value to obtain the approximate cross-sectional center point coordinates of each sub-structure;

[0023] 4.2) According to this center point, convert the sub-structure boundary pixel point coordinates into the polar coordinate system, and store the angle values corresponding to each pixel point on each sub-structure boundary in the polar coordinate system;

[0024] 4.3) For the same sub-structure, through the angle values of the corresponding pixel points obtained in step 4.2) in the polar coordinate system, in the order of the pixel point polar coordinate angle values from small to large (counterclockwise) or from large to small (clockwise), sequentially extract the Cartesian coordinate values of the pixel points on the sub-structure boundary.

[0025] Because based on the traditional Cartesian coordinate system, the extraction of pixel points can only be carried out horizontally or vertically, and it is difficult to sort the geometric relationships between pixel points in the order of arc connection (that is, clockwise or counterclockwise) by other methods. Therefore, the polar coordinate is used to extract the relative relationship of pixel points, and according to this relative relationship (clockwise or counterclockwise), the Cartesian coordinate values of the pixel points at the structural boundary are stored.

[0026] (5) The pixel coordinate values of each color correspond to an Excel file, and four Excel files are obtained.

[0027] First, create four Excel files, corresponding to the four colors respectively;

[0028] Secondly, store the pixel point coordinate values corresponding to the sub-structure boundaries of the same color into an Excel file, where each sub-structure boundary includes a number of pixel points, and the number of pixel points of different sub-structure boundaries is different; specifically:

[0029] According to the extraction order of the structural boundary pixel points in step (4), store the horizontal and vertical coordinate values of each pixel point coordinate into two adjacent columns of the Excel file of this color (one column for the horizontal coordinate value and one column for the vertical coordinate value), and each sub-structure is stored separately in two adjacent columns;

[0030] Finally, the maximum number of pixel points on the boundary of the sub-structures with the same color is used as the maximum number of rows in the Excel file of that color. For the columns where the number of pixel points on the boundary does not reach the maximum number of rows, the sub-structures are filled with the number 0 up to the maximum number of rows to facilitate subsequent data reading;

[0031] (6) The data of the Excel file obtained in step (5) is used as the original data, and the ABAQUS finite element analysis method is used for modeling. Specifically as follows:

[0032] ① Model construction steps:

[0033] The first step: Based on the secondary development of ABAQUS, open any one of the four Excel files generated in step (5);

[0034] The second step: Through the ABAQUS sketch modeling function, construct a separate sub-structure Part by reading two columns of data each time until the Excel file is read completely;

[0035] The third step: Repeat the previous two steps to complete the reading of the four Excel files and construct the Part files of all sub-structures;

[0036] ② Assign material parameters such as the elastic modulus or Poisson's ratio corresponding to the material to all sub-structures;

[0037] ③ Assemble all Parts to form a whole;

[0038] ④ Measure the maximum diameter of the original pipe cable and scale the ABAQUS model proportionally so that the size of the model structure is the same as that of the real structure;

[0039] ⑤ Set the analysis step category and adjust the analysis step size and interval;

[0040] ⑥ According to the required number of mesh elements and mesh quality requirements of the problem, set the number of structural mesh seeds and select the mesh type, and perform mesh division on the overall structure obtained in step ③;

[0041] ⑦ According to the working conditions borne by the pipe cable, apply boundary conditions and load conditions, and at the same time, according to the analysis needs, set the contact relationship between structures and set relevant parameters such as the friction coefficient;

[0042] ⑧ Submit the model in ABAQUS and perform calculations, and view the calculation results after the calculations are completed.

[0043] Compared with the existing classical ocean pipe cable modeling methods, the beneficial effects of the present invention are as follows:

[0044] (1) Traditional finite element modeling methods usually assemble the marine cable structure using ideal circles and periodic and regular geometric relationships. However, the method proposed in the present invention accurately extracts the geometric information of the structure based on digital images, and can perform accurate finite element modeling analysis of the structure considering the actual structural defects and irregular distribution relationships of the cable.

[0045] (2) The method proposed in the present invention enhances the picture based on the image color threshold technology, enabling accurate extraction of the geometric information of the structure even in the case of poor image quality or shooting environment.

[0046] (3) The color threshold image enhancement technology proposed in the present invention, based on the "Four Color Theorem", can be applied to almost all problems using four fixed colors, with good applicability. Especially for such cable problems with hierarchical structures, it has good universality, eliminating the process of operator debugging, enhancing the user experience, and enabling the method to be widely adopted in various problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is the precise finite element modeling process based on color-enhanced image recognition;

[0048] Figure 2 is the cross-section of the physical cable of the image recognition of the present invention;

[0049] Figure 3 is the result diagram after color assignment by the image processing software;

[0050] Figure 4 is the schematic diagram of the color distribution in the RGB space;

[0051] Figure 5 is the pixel extraction method based on clockwise or counterclockwise; Figure (a) is the traditional pixel extraction method, and Figure (b) is the pixel extraction method based on polar coordinates.

[0052] Figure 6 is the schematic diagram of the definition of the overall structure and sub-structures

[0053] Figure 7 is the schematic diagram of the sub-structures at the same layer in the overall structure

[0054] Figure 8 is the geometric model imported into the ABAQUS finite element software based on the Excel coordinate values;

[0055] Figure 9 is the analysis result of the precise finite element model based on the image recognition method; Figure (a) is the stress nephogram, and Figure (b) is the displacement nephogram. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] The present invention will be further described below in conjunction with specific embodiments.

[0057] An accurate finite element modeling method for the cross-section of a multi-layer and multi-material complex pipe and cable, which is a finite element modeling method based on color threshold filtering for image boundary recognition. Taking an ocean cable with a five-layer hierarchical structure (four layers of copper conductors and one outer sheath) as an example, the specific implementation steps of the present invention are as follows:

[0058] The first step: Take a photo of the solid structure cable. Since a monocular system is used, it is best to take the photo perpendicular to the structure to be measured, as Figure 2 shown; Divide the structure into sub-structures and the overall structure. For an ocean cable, the entire cable cross-section is the overall structure, and each single copper conductor and outer sheath inside it is a sub-structure, as Figure 6 shown;

[0059] The second step: Paint the overall structure with four colors, namely red (255, 0, 0), green (0, 255, 0), blue (0, 0, 255), and gray (125, 125, 125) in such a way that adjacent sub-structures have different colors and only two colors are used for sub-structures of the same layer. For an ocean cable, sub-structures of the same layer are as Figure 7 shown, and the overall coloring result is as Figure 3 shown;

[0060] The third step: Filter the overall structure according to the four colors. First, convert the boundary node coordinate information of the sub-structures into the polar coordinate system, and extract the Cartesian coordinate values of the pixel points of each sub-structure in sequence according to the clockwise or counterclockwise boundary pixel point extraction method;

[0061] The fourth step: The storage rules of the Excel file are as follows:

[0062] ① For sub-structures of the same color, store them in the same Excel file,

[0063] ② For different pixel points on the boundary of the same sub-structure, store their horizontal and vertical coordinate values of the pixel points in two adjacent columns in Excel respectively, and store the boundary coordinate data between different sub-structures adjacent to each other;

[0064] ③ In the same Excel file, use the maximum number of pixel points of the boundary of the sub-structure of this color as the maximum number of rows in this color Excel file. For sub-structures whose number of boundary pixel points of this color does not reach the maximum value, the remaining parts of the two columns storing their data are filled with 0s to the maximum number of rows;

[0065] The fifth step: Based on the secondary development of ABAQUS, implement steps such as data import, model construction, and model analysis. The specific operations are as follows:

[0066] ①Read each Excel file in sequence. In each Excel file, construct a substructure model by reading two columns at a time, and store the substructure model as a separate Part in the ABAQUS software;

[0067] ②Assign material parameters such as elastic modulus and Poisson's ratio to each Part of the structure. For the marine cable structure, the elastic modulus and Poisson's ratio are taken as 120 GPa and 0.33 respectively;

[0068] ③Assemble each substructure, i.e., Part, to form the overall structure, as Figure 7 shown;

[0069] ④Set the analysis step category and adjust the analysis step size;

[0070] ⑤Set the contact form between substructures. For the marine cable example, the normal contact between structural layers is hard, and there is friction in the tangential direction, with a friction coefficient value of 0.2;

[0071] ⑥Set the load case, and set a uniform pressure on the boundary of the outermost layer of the overall structure, with a value of 50 MPa;

[0072] ⑦Perform mesh division on the structure. In the example, tetrahedral elements are used to divide the overall structure;

[0073] ⑧Submit the analysis and calculation, and view the operation results as Figure 8 shown.

[0074] The above-described embodiments only represent the implementation modes of the present invention, but should not be construed as limiting the scope of the present invention patent. It should be noted that for those skilled in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. An accurate finite element modeling method for the cross-section of a multi-layer and multi-material complex pipe and cable, characterized in that, It includes the following steps: (1) Obtain the original cross-sectional picture information of the pipe cable Take a vertical photo of the cross-section of the pipe cable to be modeled and analyzed to obtain the cross-sectional picture information of the pipe cable; there are multiple closely arranged copper wires inside the pipe cable, and each copper wire is used as a sub-structure, and all copper wires form a multi-layer structure; (2) Perform color assignment processing on each component of the pipe cable cross-section to obtain a color picture, and perform color assignment operations on the pipe cable structure using four colors: red, green, blue, and gray; the color assignment operation follows the following rules: only two colors are used for the sub-structures in the same layer, and any sub-structure in this layer has a different color from the adjacent sub-structure; at the same time, it is required that the two colors used for this layer structure are different from the colors used for the upper and lower layer structures; (3) After filtering the color picture obtained in step (2) with the three color values of R, G, and B, through the boundary extraction algorithm, obtain the pixel points of the boundary of each copper wire in the Cartesian coordinate system, that is, obtain the relative position coordinate values of the pixel points at the structure boundary in the Cartesian coordinate system; (4) Convert the pixel coordinate values stored in the Cartesian coordinate system into pixel coordinate values in the polar coordinate system, and then the pixel coordinates can be extracted in a clockwise or counterclockwise form for subsequent modeling; specifically as follows: 4.1) After obtaining the coordinate information of each sub-structure boundary pixel point in the Cartesian coordinate system, obtain the cross-sectional center point coordinate of each sub-structure by taking the average value; 4.2) Convert the sub-structure boundary pixel point coordinates into the polar coordinate system according to this center point, and store the angle value corresponding to each pixel point on the boundary of each sub-structure in the polar coordinate system; 4.3) For the same sub-structure, through the angle values of the corresponding pixel points obtained in step 4.2) in the polar coordinate system, in the order of the pixel point polar coordinate angle values counterclockwise or clockwise, sequentially extract the Cartesian coordinate values of the pixel points on the sub-structure boundary; and store the Cartesian coordinate values of the pixel points at the structure boundary according to this relative relationship; the relative relationship refers to clockwise or counterclockwise; (5) The pixel point coordinate values of each color correspond to an Excel file, and four Excel files are obtained First, create four Excel files, corresponding to the four colors respectively; Secondly, store the pixel point coordinate values corresponding to the boundaries of the sub-structures of the same color into an Excel file, where each sub-structure boundary includes a number of pixel points, and the number of pixel points of different sub-structure boundaries is different; Specifically: according to the order of extracting the pixel points at the structure boundary in step (4), store the horizontal and vertical coordinate values of each pixel point into two adjacent columns of the Excel file of this color respectively, and each sub-structure is stored separately in two adjacent columns; Finally, use the largest number of pixel points on the boundary of the sub-structures of the same color as the maximum number of rows in the Excel file of this color, and fill the columns where the sub-structures with the number of pixel points on the boundary not reaching the maximum number of rows with the number 0 to facilitate subsequent data reading; (6) Use the data of the Excel files obtained in step 5 as the original data and perform modeling using the ABAQUS finite element analysis method; specifically as follows: ① Model construction steps: Step 1: Based on the secondary development of ABAQUS, open any one of the four Excel files generated in step (5). Step 2: Through the ABAQUS sketch modeling function, construct a separate sub-structure Part by reading two columns of data each time until the Excel file is read completely. Step 3: Repeat the above two steps to complete the reading of the four Excel files and construct the Part files of all sub-structures. ② Assign the elastic modulus or Poisson's ratio material parameters corresponding to the materials to all sub-structures respectively. ③ Assemble all Parts to form a whole by Assemble. ④ Measure the maximum diameter of the original pipe cable and scale the ABAQUS model proportionally so that the size of the model structure is the same as that of the real structure. ⑤ Set the analysis step category and adjust the analysis step size and interval. ⑥ According to the required number of mesh and the mesh quality requirements of the problem, set the number of structure mesh seeds and select the mesh type, and perform mesh division on the overall structure obtained in step ③. ⑦ Apply boundary conditions and load conditions according to the working conditions borne by the pipe cable. At the same time, according to the analysis needs, set the contact relationship between structures and set the relevant parameters of the friction coefficient. ⑧ Submit the model in ABAQUS and perform the calculation. After the calculation is completed, view the calculation results.

2. The accurate finite element modeling method for the cross-section of a multi-layer and multi-material complex pipe and cable according to claim 1, characterized in that, In step (2) above, the structure color assignment can be carried out in the following two ways: The first way is: before shooting in step (1), paint the real experimental parts with four colors of red, green, blue, and gray, and then shoot. The second way is: based on manual judgment, use the Photoshop image processing software to perform color assignment processing on the cross-section of the copper wire in the structure according to the above painting rules.

Images