A method for extracting geometric information of a bend forming FEM simulation result

Abstract

The application discloses a kind of geometric information extraction methods of elbow forming FEM simulation result, first read the node coordinates of elbow forming FEM simulation result, obtain the point cloud data of pipe section node;Acquire pipe end section node;According to section spacing, obtain the next section plane equation, utilize clustering method to obtain the same section node;Using OBB bounding box method obtains section center point;After the above two steps, obtain the center point of all sections, obtain the central point set of elbow FEM simulation result;Obtain the geometric information of elbow axis, section contour.This method can process circular, rectangular and triangular arbitrarily shaped cross-section pipe, and the plastic large deformation condition that exists grid redivision, high in calculation efficiency, accurate and reliable.

Description

A method for extracting geometric information from FEM simulation results of pipe bending. Technical Field

[0001] This invention relates to the field of metal plastic forming, and more specifically to a method for extracting geometric information from FEM simulation results of pipe bending. Background Technology

[0002] Pipe bending is a common metal tube forming process. To meet the requirements of different materials, tube shapes, and part precision, various bending methods have been developed to date, including pressure bending, wrap bending, tension bending, roll bending, and free bending (3D roll bending). When applying these methods, it is crucial to accurately grasp the correspondence between the die movement and the stress and deformation of the tube during the bending process to obtain qualified bent parts. However, due to the complexity of bending deformation, especially the susceptibility to abnormal deformation phenomena such as cross-sectional distortion, wrinkling, cracking, and springback during bending, extensive and time-consuming adjustments are often required to obtain suitable forming process parameters. Numerical simulation based on the FEM (Finite Element Method) is currently the most effective theoretical analysis tool for metal plastic forming processes. Numerical simulation can obtain detailed information about the entire forming process from blank to final part, helping engineers to identify potential problems in process and die design in advance. However, the actual formed pipe fittings can be inspected for shape and size using physical fixtures, while numerical simulation is based on a digital virtual model. FEM simulation results mainly include geometric information such as node coordinates and wall thickness variations of part elements, as well as mechanical information such as stress and strain. They cannot directly provide the process quality information that engineers are concerned with, such as springback, dimensional deviations, and cross-sectional distortion. Since the geometric information of the pipe fitting is represented by tens of thousands of discrete node coordinates, it is difficult to evaluate the forming effect using analytical data of the three-dimensional curved surface of an ideal pipe shape. If manual analysis is used, the massive amount of simulation data makes the operation cumbersome, time-consuming, and prone to errors. Therefore, how to correctly process the data information from FEM simulations of pipe bending to provide theoretical and technical support for forming processes and mold design has become an important issue in numerical simulation.

[0003] To evaluate the effect of FEM simulation for pipe forming, the literature [Yue Yongbao, Yang He, Zhan Mei, Xu Xudong, Li Guangjun. Post-processing of numerical simulation of CNC pipe bending based on Python. Journal of Plasticity Engineering, 2009, 16(3): 48-54] used Python to carry out secondary development of the ABAQUS post-processing module; based on the common quality indicators of CNC pipe bending, functions such as calculation of cross-sectional flattening rate, wall thickness change rate, springback angle, and pipe wrinkling judgment were developed. However, this study only considered the pipe bending simulation of conventional round pipes and the case where the topological information of the mesh nodes remained unchanged before and after deformation. The literature [Zengkun Zhang, Jianjun Wu, Qi Shang, et al. A new strategy for describing the characteristics of bending line in flexible push bending. Engineering Computations, 2019] manually extracted the coordinate points of the specified cross-section, calculated the center point, imported it into the three-dimensional drawing software to fit the axis, and calculated its curvature distribution. However, this method is cumbersome and inefficient for extracting geometric information from pipe bends with multiple curved segments, and it is also prone to errors. Currently, there is a lack of effective methods for extracting geometric information from FEM simulation results of pipe bends, particularly for non-circular cross-section irregular pipes and their complex spatial bends. Summary of the Invention

[0004] Based on the above background, this invention proposes a method for extracting geometric information from FEM simulation results of pipe bending, which can accurately and efficiently obtain geometric information such as axis and cross-sectional shape from FEM simulation data of spatial bending of pipes with various cross-sections.

[0005] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0006] On the one hand, the present invention provides a method for extracting geometric information from FEM simulation results of pipe bending, the method specifically including the following steps:

[0007] S1, read the node coordinates of the FEM simulation results of the bent pipe;

[0008] S2, end section node acquisition;

[0009] S3, the center points of the remaining sections are obtained;

[0010] S4, Obtain the axis of the bend;

[0011] S5, Calculation of pipe bend dimensions.

[0012] Furthermore, the steps for obtaining the node coordinates of the bent pipe FEM model in S1 specifically include: establishing a finite element model and completing the simulation in numerical simulation software such as Abaqus and Ansys, and extracting the node coordinates of the bent pipe FEM model from the result file.

[0013] Furthermore, the step of obtaining the S2 end section node specifically includes:

[0014] S21, converts the axial direction of the end straight segment to be parallel to the X-axis;

[0015] S22, use bubble sort to obtain the coordinates of the end section.

[0016] Furthermore, the step of obtaining the center points of the remaining cross-sections in S3 specifically includes:

[0017] S31, obtain the coordinates of the next cross section;

[0018] S32, the center point of the cross section is obtained using the OBB bounding box method;

[0019] S33, repeat the above two steps to obtain the center points of all cross sections.

[0020] Furthermore, the step of obtaining the next cross-sectional coordinate point in S31 specifically includes:

[0021] S311, obtain the three coordinate points with the most dispersed distribution in the previous cross section, denoted as F. n-1 ;

[0022] S312, obtain the coordinates of points other than the previous section;

[0023] S313, obtain and F n-1 The nearest set of points is used to obtain the plane equation denoted as P using the RANSAC algorithm. n ;

[0024] S314, using clustering to obtain the relationship with P n The set of points at the nearest cross-section.

[0025] Furthermore, step S311 obtains the three coordinate points with the most dispersed distribution in the previous cross-section, denoted as F. n-1 The specific steps include:

[0026] S3111, combine every three section nodes and calculate the area of ​​the enclosed triangle; the formula for calculating the area of ​​a triangle is:

[0027]

[0028] Where l1, l2, and l3 are the distances between each pair of points.

[0029] S3112, using bubble sort, we obtain the triangle with the largest area and its three corresponding coordinates, denoted as F. n-1 .

[0030] Furthermore, S313 acquires and F n-1 The plane equation is obtained using the RANSAC algorithm from the nearest set of points, denoted as P. n The specific steps include:

[0031] S3131, calculate the remaining points and F. n-1 The distance;

[0032] S3132, using bubble sort to obtain the result with F n-1 The closest point is denoted as S. n ;

[0033] S3133, using the RANSAC algorithm, obtain S n The equation of the plane in which it lies is denoted as P. n The equation of a plane is expressed in the form of:

[0034]

[0035] Where a, b, and c are unknown parameters, obtained through fitting.

[0036] Furthermore, step S314 uses clustering to obtain the relationship with P. n The steps for finding the nearest set of cross-section points specifically include:

[0037] S3141, calculate the remainder point and P. n The distance; the distance formula is

[0038]

[0039] Where, x n y n z n Here are the coordinate values, and a, b, and c are the plane equation parameters obtained from S3133.

[0040] S3142, grouping points within a certain distance into one category;

[0041] S3143, Delete nodes that are not from the same cross section in the class;

[0042] S3144, the remaining points are the cross-sectional coordinate points, denoted as SS. n .

[0043] Furthermore, the step of obtaining the S4 bend axis specifically includes:

[0044] S41, Number the center points of the cross section;

[0045] S42, connect sequentially to obtain the axis.

[0046] Furthermore, the steps for calculating the S5 bend size information specifically include:

[0047] S51, Enter the start and end point numbers of the arc segment from which you want to extract information;

[0048] S52, obtain arc segment size information.

[0049] This invention can handle pipes with arbitrary irregular cross-sections such as circles, rectangles, and triangles; it does not require the use of the FEM mesh coordinates of the pipe before deformation, therefore it can be used in cases of large plastic deformation where mesh re-division is required. Data extraction and automatic calculation are achieved through programming, resulting in high efficiency, accuracy, and reliability. Attached Figure Description

[0050] Figure 1 is a flowchart of a method for extracting geometric information from FEM simulation results of pipe bending according to the present invention;

[0051] Figure 2 is a detailed flowchart of S3 in Figure 1 of the present invention;

[0052] Figure 3 is a detailed flowchart of S31 in Figure 2 of the present invention;

[0053] Figure 4 illustrates the process of obtaining the next cross-sectional coordinate point according to an embodiment of the present invention;

[0054] Figure 5 is a schematic diagram of the axis of the triangular cross-section bent pipe FEM model according to an embodiment of the present invention;

[0055] Figure 6 is a schematic diagram of the axis of the circular cross-section and the rectangular cross-section bent pipe FEM model of an embodiment of the present invention; Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Referring to Figures 1-6, this invention provides a method for extracting geometric information from FEM simulation results of bent pipe forming. This method enables efficient acquisition of geometric information from FEM simulation results of spatially complex bent pipes with irregular cross-sections, thereby evaluating the simulation forming effect. Taking a spatially bent pipe with a triangular cross-section as an example, the FEM simulation software used is Abaqus, and the implementation steps are as follows:

[0058] Step S1: Read the node coordinates of the bent pipe FEM model. Specifically, this includes: creating a finite element model of the bent pipe forming process in Abaqus and adding COORD field output. Based on the shape of the target 3D bent pipe model, setting the motion of the bending die, submitting the calculation, and obtaining the Abaqus result odb file. Using a Python script, read the specified odb file, the analysis step and the frame number within the analysis step, and the pipe node range of the COORD variable to obtain the node coordinates of the deformed pipe fitting FEM model.

[0059] Step S2, obtaining the end section node. Specifically, this includes:

[0060] Step S21: Convert the axial direction of the end straight segment to be parallel to the X-axis. Perform coordinate transformation on all coordinates of the bend, transform the coordinate system so that the X-coordinate values ​​of nodes on each cross-section of the end straight segment are the same, and store them in the list CoordinatesAll.

[0061] Step S22: Obtain the coordinates of the end section using bubble sort. Use a for loop to iterate through the elements of the list CoordinatesAll, sort them by their X-coordinates (the elements at index 0), obtain the coordinates of the end section, and write them into the list Coordinates1.

[0062] Step S3, obtaining the center points of the remaining cross-sections, the process is shown in Figure 2, specifically including:

[0063] Step S31: Obtain the coordinates of the next cross-section point. The process is shown in Figure 3, and specifically includes:

[0064] Step S311: Obtain the three coordinate points with the most dispersed distribution in the previous cross section, denoted as F. n-1 Use a for loop to iterate through the list Coordinates1, combining every three points. Calculate the area of ​​the triangle using the formula: [Formula omitted for brevity].

[0065]

[0066] Where l1, l2, and l3 are the distances between each pair of points.

[0067] Find the three points corresponding to the triangle with the largest area, denoted as F. n-1 These three relatively dispersed points represent this cross section, as shown in Figure 4.

[0068] Step S312: Obtain the coordinates of points other than the previous cross-section. Use a for loop to iterate through the elements of the list CoordinatesAll. If an element is not in the list Coordinates1, store it in a new list CoordinatesRemain.

[0069] Step S313, obtain the information related to F n-1 The set of nearest points is shown in Figure 4. The equation of the plane is obtained using the RANSAC algorithm, denoted as P. n Use a for loop to iterate through the elements of the list CoordinatesRemain, and use the point spacing formula to calculate the distance between the points and F. n-1 If the distance is within ±2mm of the cross-sectional spacing, write it into the list coor n 0. List coor n Convert 0 to an array, define the plane equation x + a * y + b * z + c = 0, use the RANSAC algorithm to obtain the parameters a, b, and c, and denote the plane as P. n As shown in Figure 4.

[0070] Step S314: Use clustering to obtain the data related to P. n Find the set of closest cross-section points. Use a for loop to iterate through the elements of the list `CoordinatesRemain` and apply the distance formula. Calculate the remaining points and P n The distance to P is obtained. n Element points within 2mm of each other are stored in list S. n In order to remove S n For nodes that are not on the same cross section, use a for loop to traverse list S. n An element is removed from S if the distance between a node and the center point of the previous section is greater than 1.5 times the maximum distance between points within the section (denoted as D), or if the distance between a node and any other node is greater than 1.1 times D. n Remove from the middle to obtain the cross-sectional point set, denoted as SS. n As shown in Figure 4.

[0071] Step S32: Obtain the center point of the cross-section using the OBB bounding box method. The specific process is as follows: Calculate P... n The angle between plane P and plane YZ, will be determined by P. n Transform the surface into a surface P perpendicular to the X-coordinate axis. n '; By multiplying with the coordinate transformation matrix, the SS is transformed. n A coordinate system is established such that its X-coordinate values ​​are equal; SS is obtained. n Find the extreme values ​​of the Y and Z coordinates in the middle, and establish a rectangular bounding box; obtain the coordinates of the center point of the rectangular bounding box, transform back to the initial coordinate system, and obtain the coordinates of the center point of this cross section.

[0072] Step S33: Repeat steps S31 and S32 to obtain the center points of all cross sections, as shown in Figure 5.

[0073] The above steps can also be used to obtain the axis points of the FEM simulation results of the spiral pipe with a circular cross section and the axis points of the FEM simulation results of the curved pipe with a rectangular cross section, as shown in Figure 6.

[0074] Step 4, Obtaining the pipe bend axis. This specifically includes:

[0075] Step S41: Number the center points of the cross-section. Number the center points according to the extraction order, and output the coordinates of the center points as a txt file.

[0076] Step S42: Connect the axes sequentially to obtain the axis.

[0077] Step 5, calculate the bend dimensions. This includes:

[0078] Step S51: Input the start and end point numbers of the arc segment from which the information to be extracted is required. First, taking the end section of the pipe fitting as the starting point, obtain the distances from the start point to the entry and exit points of the arc segment to be extracted, and divide them by the cross-sectional spacing to obtain the start and end point numbers.

[0079] Step S52: Obtain the arc segment dimension information. The axis points between the start and end points of the arc segment are obtained by numbering the start and end points. The plane equation and sphere equation of the arc segment are fitted using the least squares method, and the bending radius of the arc segment is obtained by solving them simultaneously. The bending angle of the arc segment is calculated using the center points of the start and end points of the arc segment.

[0080] To verify the accuracy of the extracted geometric information from the FEM model of the bent pipe, a 3D model of the bent pipe was built in UG based on the acquired geometric information and compared with the simulated FEM model. The 3D model of the bent pipe largely overlapped with the FEM model, proving that the extracted geometric information was correct.

Claims

1. A method for extracting geometric information from FEM simulation results of pipe bending, characterized in that, Specifically, the following steps are included: S1, Read the node coordinates of the FEM model of the bend; S2, Obtain the end section node; S3, Obtain the center point of the remaining sections; S4, Obtain the axis of the bend; S5, Calculate the bend dimension information. Step S1 specifically includes: establishing a bent pipe FEM model in numerical simulation software for simulation, and extracting node coordinates from the result file; Step S2 specifically includes: S21, converting the axial direction of the end straight segment to a direction parallel to the X-axis; S22, obtaining the coordinate points of the end section using the bubble sort method; Step S3 specifically includes: S31, obtaining the coordinate points of the next section; S32, obtaining the center point of the section using the OBB bounding box method; S33, repeating the above two steps to obtain all section center points; Step S31 specifically includes: S311, obtaining the three coordinate points with the most dispersed distribution in the previous section, denoted as F n-1 S312, obtain the coordinates of points other than the previous section; S313, obtain the coordinates of points relative to F. n-1 The plane equation is obtained using the RANSAC algorithm from the nearest set of points, denoted as P. n S314, using clustering to obtain the relationship with P n The nearest set of cross-sectional points; step S311 specifically includes: S3111, combining every three cross-sectional nodes and calculating the area of ​​the enclosed triangle; the formula for calculating the area of ​​the triangle is... Where l1, l2, and l3 are the distances between any two points, S3112, the triangle with the largest area is obtained using bubble sort, and the corresponding three coordinate points are obtained, denoted as F. n-1 The steps in S313 specifically include: S3131, calculating the remaining points and F. n-1 The distance between F and S3132 is obtained using bubble sort. n-1 The closest point is denoted as S. n S3133, using the RANSAC algorithm, obtain S n The equation of the plane in question is denoted as P. n The plane equation is expressed in the form of Where a, b, and c are unknown parameters obtained through fitting; step S314 specifically includes: S3141, calculating the remaining points and P n The distance; the distance formula is Where, x n y n z n Here are the coordinate values, a, b, and c are the plane equation parameters obtained in S3133, S3142 groups points within a certain distance into one class, S3143 deletes nodes that are not on the same cross section in the class, and S3144 denotes the remaining points as cross section coordinate points, denoted as SS. n .

2. The method for extracting geometric information from FEM simulation results of pipe bending as described in claim 1, characterized in that, The steps for obtaining the S4 bend axis specifically include: S41, numbering the center points of the cross section; S42, sequentially connecting them to obtain the axis.

3. The method for extracting geometric information from FEM simulation results of pipe bending as described in claim 1, characterized in that, The steps in S5 specifically include: S51, inputting the start and end point numbers of the arc segment from which the information to be extracted; S52, obtaining the arc segment size information.

Images