A method for predicting axial compression performance of steel columns based on morphological defect scanning

Abstract

A method for predicting the axial compression performance of a steel column based on morphological defect scanning relates to the field of axial compression performance evaluation of steel columns, and the method comprises the following steps: S1: three-dimensional morphological scanning; S2: three-dimensional design model matching and alignment; S3: acquisition of elevation distribution of steel column morphological deviation; S4: establishment of a finite element analysis model; S5: eigenvalue buckling analysis, initial morphological defect fitting; S6: axial compression bearing analysis, extraction of bearing capacity and axial compression failure characteristics; S7: neural network prediction model construction; S8: prediction of axial compression bearing capacity and failure mode of steel columns. The present invention directly and quickly predicts the bearing capacity and failure mode of steel columns by scanning the apparent morphology of steel columns, which helps to timely discover the problem of insufficient bearing capacity of steel columns caused by excessive morphological deviation of steel columns at the construction site, so as to achieve timely deformation correction and reinforcement control and ensure structural safety.

Description

Technical Field

[0001] The invention relates to the field of steel column axial compression performance evaluation, and in particular to a steel column axial compression performance prediction method based on morphology defect scanning. Background Art

[0002] Steel structures have the advantages of light weight, high structural reliability, high degree of mechanization in manufacturing and installation, good sealing performance, low carbon and energy saving, and green environmental protection, so they are widely used in the field of civil engineering. Due to its high material strength, steel can provide a large load-bearing capacity with a small cross-sectional size. Therefore, the slenderness of steel columns is relatively large when used, and they will be unstable and damaged under axial compression. In engineering applications, steel columns often adopt open-section structures. Factors such as manufacturing deviations of components and plates make it inevitable that initial morphological defects such as initial eccentricity and initial bending will occur in the steel columns. These initial morphological defects directly affect the axial compression bearing capacity and failure mode of the steel columns.

[0003] At present, the axial compression stability bearing capacity of steel columns is mainly calculated by modifying the Euler formula based on empirical data to obtain empirical column curves. However, in practical applications, there are often irregular structural steel columns or cold-bent thin-walled steel columns, and their axial bearing capacity often needs to be tested to obtain their true bearing capacity. However, the initial defects of steel columns manufactured and sampled in the laboratory often cannot represent the morphological deviation of steel columns in actual projects; the bearing capacity of steel columns sampled in the laboratory cannot directly represent the bearing performance of steel columns under different batches and structures in actual projects. In view of this, it is urgent to provide a prediction method that can directly predict the axial bearing capacity and failure mode of steel columns based on the surface morphology scanning of steel columns on site, so as to provide scientific guidance for the safety control of steel columns and the control of construction deviations. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method for predicting the axial compression performance of steel columns based on morphological defect scanning, in response to the shortcomings described in the above-mentioned background technology, so as to solve the technical problems such as the manpower, material resource investment and time lag that may be caused by on-site random inspections, and the inability of indoor random inspections to reflect the deviation characteristics of morphological defects on the construction site in real time.

[0005] The technical solution adopted by the present invention to solve the technical problem is: a method for predicting the axial compression performance of a steel column based on morphological defect scanning, comprising the following steps:

[0006] S1: 3D topography scanning;

[0007] For the steel column to be evaluated, a three-dimensional scanner is used to perform a three-dimensional morphology scan on the surface of the steel column, so that the scanning area covers the entire steel column and the top and bottom end surfaces, and a three-dimensional point cloud model of the overall surface morphology of the steel column is obtained;

[0008] S2: 3D design model matching and alignment;

[0009] Import the 3D design model of the steel column, and perform feature matching and model alignment on the bottom centroid, the centroid axis and the bottom end face of the 3D design model with the bottom center point, the fitting axis and the bottom end face of the 3D point cloud model;

[0010] S3: Acquisition of steel column shape deviation elevation distribution;

[0011] Using the surface of the 3D design model as a reference surface, project the surface of the 3D point cloud model with the corresponding surface of the 3D design model, calculate the 3D point cloud projection and the elevation deviation of the scanning point compared to the position of the 3D design model surface, and obtain the 3D point cloud deviation elevation distribution;

[0012] S4: Establish finite element analysis model;

[0013] Establish a finite element analysis model of the steel column based on the three-dimensional design model of the steel column;

[0014] S5: eigenvalue buckling analysis, initial morphology defect fitting;

[0015] The eigenvalue buckling analysis of the steel column is performed using finite element method to obtain the overall buckling mode and local buckling mode of the steel column. The different overall buckling modes and local buckling modes obtained are selected, and the buckling mode amplitudes are scaled and superimposed to obtain the fitted initial morphological defect distribution, that is, the elevation deviation distribution from the surface of the three-dimensional design model.

[0016] S6: Axial compression analysis, extracting bearing capacity and axial compression failure characteristics;

[0017] The initial morphological defect distribution is imported into the finite element analysis model and the node coordinates of the finite element analysis model are updated. The axial compressive bearing capacity of the steel column considering the initial morphological defects is calculated to obtain the axial ultimate bearing capacity of the steel column. The failure state is defined as when the bearing capacity of the steel column decreases by 85%. The failure morphological characteristics of the steel column in the failure state are extracted, that is, the deviation elevation distribution of the deformation surface of the steel column in the failure state compared with the three-dimensional design surface position;

[0018] S7: Neural network prediction model construction;

[0019] The initial morphological defect distribution is used as the training input, and the failure morphological characteristics corresponding to the steel column bearing capacity and failure state are used as the training output. A prediction model for the axial compressive bearing capacity and failure morphological characteristics of steel columns is established and trained based on the graph neural network.

[0020] S8: Axial compression bearing capacity and failure mode prediction of steel columns;

[0021] The three-dimensional point cloud deviation elevation distribution of the steel column to be evaluated obtained through steps S1-S3 is imported into the input layer of the established graph neural network steel column axial compressive bearing capacity and failure morphology characteristics prediction model for analysis, and the output layer obtains the axial compressive bearing capacity and failure morphology characteristics of the steel column under actual three-dimensional morphological defects.

[0022] Further, in step S2:

[0023] Based on the point cloud fitting plane of the bottom end face of the 3D point cloud model, the fitting axis of the 3D point cloud model of the steel column is obtained based on the principal component analysis method. The intersection of the 3D point cloud model fitting axis and the bottom end face of the 3D point cloud model is the bottom center point of the 3D point cloud model.

[0024] Furthermore, in the calculation of the 3D point cloud deviation elevation distribution in step S3, when the surface of the 3D point cloud model is away from the centroid axis of the 3D design model compared to the corresponding surface of the 3D design model, the scanning point deviation elevation takes a positive value; when the surface of the 3D point cloud model is biased towards the centroid axis of the 3D design model compared to the corresponding surface of the 3D design model, the scanning point deviation elevation takes a negative value; when the surface of the 3D point cloud model is coplanar with the centroid axis of the 3D design model, the counterclockwise direction of rotation around the centroid axis of the 3D design model is positive, and the clockwise direction is negative.

[0025] Furthermore, in the calculation of the scanning point deviation elevation of the three-dimensional point cloud model in step S3, the surface of the steel column can be numbered and gridded, and the grid division size should be less than B / 5 to ensure the accuracy of the morphology deviation description, and B is the minimum width of the cross section; then the compared point can be uniquely marked with a set of surface numbers and relative position numbers, and the projection height of the point obtained by projecting the three-dimensional point cloud model surface onto the corresponding three-dimensional design model surface is used as the scanning point deviation elevation value at the point.

[0026] Furthermore, the deviation elevation amplitude of the initial morphological defect distribution fitted in step S5 is in the range of [-L / 50, L / 50] to cover the actual possible initial defect range, where L is the height of the steel column.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention performs three-dimensional morphology scanning on the installed steel columns at the construction site and compares it with the three-dimensional design model of the steel columns in real time, so as to realize non-destructive acquisition of the elevation distribution of the three-dimensional morphology defect deviation and the deviation amplitude on each section of the steel columns; in the early stage, a correlation prediction model of the initial morphology defect deviation distribution and the bearing capacity and failure mode of the steel columns is established based on a large number of artificial initial defect distribution components and the numerical analysis results of the bearing capacity level and failure mode under the corresponding initial morphology defects, which can quickly predict the bearing performance of the steel columns based on the apparent morphology deviation characteristics of the steel columns, and help the construction sites to promptly discover the problem of insufficient bearing capacity of the steel columns caused by excessive morphology deviation of the steel columns, so as to achieve timely deformation correction and reinforcement control, and ensure structural safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the method for predicting the axial compression performance of a steel column based on morphological defect scanning of the present invention;

[0030] Figure 2 It is a schematic diagram of the three-dimensional morphological defect scanning process of a steel column;

[0031] Figure 3 It is a schematic diagram of the three-dimensional point cloud model of the overall surface morphology of the steel column obtained by scanning;

[0032] Figure 4 It is a schematic diagram of point, line and surface feature matching and model alignment between the 3D point cloud model of the steel column and the 3D design model;

[0033] Figure 5 The elevation diagram of the point cloud deviation is obtained by projecting the surface of the three-dimensional point cloud model onto the surface of the three-dimensional design model;

[0034] Figure 6 is the obtained three-dimensional morphology defect deviation elevation grayscale map;

[0035] Figure 7 It is a schematic diagram of the global buckling mode and local buckling mode obtained by the eigenvalue buckling analysis of the steel column;

[0036] Figure 8 It is a schematic diagram of the three-dimensional morphological characteristics of steel column failure obtained by analyzing a set of initial defect distribution modes;

[0037] Description of reference numerals in the figures:

[0038] 1. Steel column, 2. 3D scanner, 3. Scanning area, 4. 3D point cloud model, 41. Bottom center point of 3D point cloud model, 42. Fitting axis of 3D point cloud model, 43. Bottom end surface of 3D point cloud model, 5. 3D design model, 51. Bottom centroid of 3D design model, 52. Centroid axis of 3D design model, 53. Bottom end surface of 3D design model, 6. Surface of 3D point cloud model, 7. Surface of 3D design model, 8. 3D point cloud projection, 9. Scanning point deviation elevation, 10. 3D point cloud deviation elevation distribution, 11. Overall buckling mode, 12. Local buckling mode, 13. Failure morphology characteristics. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments, not all of the embodiments. Based on this embodiment, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] Reference Figure 1 A method for predicting the axial compression performance of a steel column based on morphological defect scanning specifically comprises the following steps:

[0041] S1: 3D topography scanning;

[0042] For the steel column 1 to be evaluated, a three-dimensional scanner 2 is used to perform a three-dimensional morphology scan on the surface of the steel column 1, such as Figure 2 As shown, multiple scans are required to make the scanning area 3 cover the entire steel column 1 and the top and bottom end surfaces, and obtain a three-dimensional point cloud model 4 of the overall appearance of the steel column 1, such as Figure 3 As shown;

[0043] S2: 3D design model matching and alignment;

[0044] Based on the point cloud fitting plane of the bottom end face 43 of the three-dimensional point cloud model, and based on the principal component analysis method, the three-dimensional point cloud model fitting axis 42 of the steel column 1 is obtained, and the intersection of the three-dimensional point cloud model fitting axis 42 and the bottom end face 43 of the three-dimensional point cloud model is the bottom center point 41 of the three-dimensional point cloud model;

[0045] Import the 3D design model 5 of the steel column 1, and perform feature matching and model alignment on the bottom centroid 51, the centroid axis 52, and the bottom end face 53 of the 3D design model with the bottom center point 41, the fitting axis 42, and the bottom end face 43 of the 3D point cloud model in sequence, as shown in FIG. Figure 4 As shown;

[0046] S3: Acquisition of steel column shape deviation elevation distribution;

[0047] Taking the 3D design model surface 7 as the reference surface, the 3D point cloud model surface 6 is projected with the corresponding 3D design model surface 7, and the 3D point cloud projection 8 and the scanning point deviation elevation 9 at the position compared with the 3D design model surface 7 are calculated, such as Figure 5 As shown, the three-dimensional point cloud deviation elevation distribution 10 is obtained at the same time. For details, refer to the three-dimensional morphology defect deviation elevation grayscale map, as shown in Figure 6 As shown;

[0048] S4: Establish finite element analysis model;

[0049] A finite element analysis model of the steel column 1 is established based on the three-dimensional design model 5 of the steel column 1;

[0050] S5: eigenvalue buckling analysis, initial morphology defect fitting;

[0051] The eigenvalue buckling analysis of the steel column 1 is performed using finite element method, and the overall buckling mode 11 and the local buckling mode 12 of the steel column 1 are obtained, as shown in Figure 7 As shown, different global buckling modes 11 and local buckling modes 12 are selected, and the buckling mode amplitudes are scaled and superimposed to obtain the fitted initial morphological defect distribution, that is, the elevation deviation distribution from the surface of the three-dimensional design model;

[0052] S6: Axial compression analysis, extracting bearing capacity and axial compression failure characteristics;

[0053] The initial morphological defect distribution is imported into the finite element analysis model and the node coordinates of the finite element analysis model are updated. The axial compressive bearing capacity of the steel column considering the initial morphological defects is calculated to obtain the axial ultimate bearing capacity of the steel column. The failure state is defined when the bearing capacity of the steel column decreases by 85%, and the failure morphological characteristics of the steel column in the failure state are extracted, such as Figure 8 As shown, it is the deviation elevation distribution of the deformation surface of the steel column under the failure state compared with the three-dimensional design surface position;

[0054] S7: Neural network prediction model construction;

[0055] The initial morphological defect distribution is used as the training input, and the failure morphological characteristics corresponding to the steel column bearing capacity and failure state are used as the training output. A prediction model for the axial compressive bearing capacity and failure morphological characteristics of steel columns is established and trained based on the graph neural network.

[0056] S8: Axial compression bearing capacity and failure mode prediction of steel columns;

[0057] The three-dimensional point cloud deviation elevation distribution of the steel column 1 to be evaluated obtained through steps S1-S3 is imported into the input layer of the established graph neural network steel column axial compressive bearing capacity and failure morphology characteristics prediction model for analysis, and the output layer obtains the axial compressive bearing capacity and failure morphology characteristics 13 of the steel column under actual three-dimensional morphological defects.

[0058] Among them, in the calculation of the three-dimensional point cloud deviation elevation distribution 10 in step S3, when the surface of the three-dimensional point cloud model 4 is away from the direction of the three-dimensional design model centroid axis 52 compared to the corresponding surface of the three-dimensional design model 5, the scanning point deviation elevation 9 takes a positive value; when the surface of the three-dimensional point cloud model 4 is biased towards the direction of the three-dimensional design model centroid axis 52 compared to the corresponding surface of the three-dimensional design model 5, the scanning point deviation elevation 9 takes a negative value; when the surface of the three-dimensional point cloud model 4 is coplanar with the three-dimensional design model centroid axis 52, the counterclockwise direction of rotation around the three-dimensional design model centroid axis 52 is positive, and the clockwise direction is negative.

[0059] Among them, in the calculation of the scanning point deviation elevation 9 of the three-dimensional point cloud model 4 in step S3, the surface of the steel column 1 can be numbered and gridded, and the grid division size should be less than B / 5 to ensure the accuracy of the morphology deviation description, and B is the minimum width of the cross section; then the compared point can be uniquely marked with a set of surface numbers and relative position numbers, and the projection height of the point obtained by projecting the surface of the three-dimensional point cloud model 4 onto the surface of the corresponding three-dimensional design model 5 is used as the value of the scanning point deviation elevation 9 at the point.

[0060] The deviation elevation amplitude of the initial morphological defect distribution fitted in step S5 is in the range of [-L / 50, L / 50] to cover the actual possible initial defect range, where L is the height of the steel column.

[0061] The present invention performs three-dimensional morphology scanning of the installed steel columns at the construction site and compares it with the three-dimensional design model of the steel columns in real time, thereby realizing non-destructive acquisition of the elevation distribution of three-dimensional morphology defect deviations and the deviation levels on each section of the steel columns; in the early stage, a correlation prediction model of the initial morphology defect deviation distribution and the bearing capacity and failure mode of the steel columns is established based on a large number of artificial initial defect distribution components and the numerical analysis results of the bearing capacity level and failure mode under the corresponding initial morphology defects. The bearing performance of the steel columns can be quickly predicted based on the apparent morphology deviation characteristics of the steel columns, which helps to promptly discover the problem of insufficient bearing capacity of the steel columns caused by excessive morphology deviations at the construction site, so as to achieve timely deformation correction and reinforcement control and ensure structural safety.

[0062] Although the preferred embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments, which are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which fall within the scope of protection of the present invention.

Claims

1. A method for predicting the axial compression performance of a steel column based on morphological defect scanning, characterized in that: The following steps are involved: S1: 3D topography scanning; For the steel column to be evaluated, a three-dimensional scanner is used to perform a three-dimensional morphology scan on the surface of the steel column, so that the scanning area covers the entire steel column and the top and bottom end surfaces, and a three-dimensional point cloud model of the overall surface morphology of the steel column is obtained; S2: 3D design model matching and alignment; Import the 3D design model of the steel column, and perform feature matching and model alignment on the bottom centroid, the centroid axis and the bottom end face of the 3D design model with the bottom center point, the fitting axis and the bottom end face of the 3D point cloud model; S3: Acquisition of elevation distribution of steel column morphology deviation; Using the surface of the 3D design model as a reference surface, project the surface of the 3D point cloud model with the corresponding surface of the 3D design model, calculate the 3D point cloud projection and the elevation deviation of the scanning point compared to the position of the 3D design model surface, and obtain the 3D point cloud deviation elevation distribution; S4: Establish finite element analysis model; Establish a finite element analysis model of the steel column based on the three-dimensional design model of the steel column; S5: eigenvalue buckling analysis, initial morphology defect fitting; The eigenvalue buckling analysis of the steel column is performed using finite element method to obtain the overall buckling mode and local buckling mode of the steel column. The different overall buckling modes and local buckling modes obtained are selected, and the buckling mode amplitudes are scaled and superimposed to obtain the fitted initial morphological defect distribution, that is, the elevation deviation distribution from the surface of the three-dimensional design model. S6: Axial compression analysis, extracting bearing capacity and axial compression failure characteristics; The initial morphological defect distribution is imported into the finite element analysis model and the node coordinates of the finite element analysis model are updated. The axial compressive bearing capacity of the steel column considering the initial morphological defects is calculated to obtain the axial ultimate bearing capacity of the steel column. The failure state is defined as when the bearing capacity of the steel column decreases by 85%. The failure morphological characteristics of the steel column in the failure state are extracted, that is, the deviation elevation distribution of the deformation surface of the steel column in the failure state compared with the three-dimensional design surface position; S7: Neural network prediction model construction; The initial morphological defect distribution is used as the training input, and the failure morphological characteristics corresponding to the steel column bearing capacity and failure state are used as the training output. A prediction model for the axial compressive bearing capacity and failure morphological characteristics of steel columns is established and trained based on the graph neural network. S8: Axial compression bearing capacity and failure mode prediction of steel columns; The three-dimensional point cloud deviation elevation distribution of the steel column to be evaluated obtained through steps S1-S3 is imported into the input layer of the established graph neural network steel column axial compressive bearing capacity and failure morphology characteristics prediction model for analysis, and the output layer obtains the axial compressive bearing capacity and failure morphology characteristics of the steel column under actual three-dimensional morphological defects.

2. The method for predicting axial compression performance of a steel column based on morphological defect scanning according to claim 1 is characterized in that: In step S2: Based on the point cloud fitting plane of the bottom end face of the 3D point cloud model, the fitting axis of the 3D point cloud model of the steel column is obtained based on the principal component analysis method. The intersection of the 3D point cloud model fitting axis and the bottom end face of the 3D point cloud model is the bottom center point of the 3D point cloud model.

3. The method for predicting the axial compression performance of a steel column based on morphological defect scanning according to claim 1 or 2, characterized in that: In the calculation of the 3D point cloud deviation elevation distribution in step S3, when the surface of the 3D point cloud model is away from the centroid axis direction of the 3D design model compared to the corresponding surface of the 3D design model, the scanning point deviation elevation takes a positive value; when the surface of the 3D point cloud model is biased towards the centroid axis direction of the 3D design model compared to the corresponding surface of the 3D design model, the scanning point deviation elevation takes a negative value; When the surface of the 3D point cloud model is coplanar with the centroid axis of the 3D design model, the counterclockwise direction of rotation around the centroid axis of the 3D design model is positive, and the clockwise direction is negative.

4. The method for predicting axial compression performance of a steel column based on morphological defect scanning according to claim 1 or 2, characterized in that: In the calculation of the scanning point deviation elevation of the three-dimensional point cloud model in step S3, the three-dimensional point cloud model surface of the steel column is numbered and gridded, and the grid division size should be less than B / 5 to ensure the accuracy of the morphology deviation description, and B is the minimum width of the cross section; the scanning point on the three-dimensional point cloud model is uniquely marked with a set of surface numbers and relative position numbers, and the projection height of the scanning point obtained by projecting the three-dimensional point cloud model surface onto the corresponding three-dimensional design model surface is used as the deviation elevation value of the scanning point.

5. The method for predicting axial compression performance of a steel column based on morphological defect scanning according to claim 1 or 2, characterized in that: The deviation elevation amplitude of the initial morphological defect distribution fitted in step S5 is in the range of [-L / 50, L / 50], where L is the height of the steel column.

Images