Roof structure wind vibration coefficient calculation method

Abstract

The invention relates to the field of roof structure design, and provides a roof structure wind vibration coefficient calculation method in order to improve the calculation precision of a wind vibration coefficient, and through a measuring point arrangement method and a load combination method, a numerical simulation method and a physical wind tunnel test method are combined together for wind load response research. The inherent defects of the two methods are avoided, and the advantages of the two methods are combined, so that the calculation precision of the wind vibration coefficient is higher.

Description

Technical Field

[0001] The invention relates to the field of roof structure design, in particular to a method for calculating the wind vibration coefficient of a roof structure. Background Art

[0002] Complex large-span roof structures are prone to large vibrations under wind loads due to their large roof flexibility. The biggest impact on structural safety is the instantaneous damage that may be caused by short-term excessive vibrations under non-steady fluctuating winds. However, it is difficult to obtain wind load data under fluctuating winds, and it is even more difficult to obtain structural responses under fluctuating winds. Therefore, the concept of wind vibration coefficient is proposed. The wind vibration coefficient is the ratio of the total structural response under fluctuating wind to the structural response caused by the average wind load. It is mainly used to quantify the relationship between the structural displacement or deformation of the structure under fluctuating wind and the structural displacement under steady-state wind. With the help of the wind vibration coefficient, the fluctuating wind response can be quickly converted based on the readily available average wind load.

[0003] When calculating the wind vibration coefficient, the existing technology mostly uses the wind tunnel test method for those with physical wind tunnel test conditions. However, wind tunnel tests basically use rigid model tests, which cannot truly reflect the elastic-plastic properties that the structure should have. Therefore, only the wind pressure distribution on the outer surface of the structural model can be obtained, but the deformation response data of the structure under wind load cannot be obtained. In order to save time and economic costs, numerical simulation methods are more often used. The data resolution of numerical simulation is higher, and the data obtained is more accurate, but it also has the same defect as the wind tunnel test, that is, it cannot consider the elastic-plastic deformation of the structural material.

[0004] Ansys APDL software can be used for finite element dynamic response calculation. However, the load data input when using this software for analysis is the wind tunnel test pressure measurement data. Due to the limitations of the wind tunnel test model, the number of pressure measuring holes that can be set is limited, and the distribution position is not completely reasonable. For example, the spacing between the upper and lower surfaces of the outer cantilevered eaves is often very small, and it is impossible to drill holes to install the pressure measuring tube equipment. Therefore, during the test, the measuring points that should be arranged at the edge in theory will be moved inward, which greatly affects the accuracy of the results. In addition, since the finite element analysis divides each structural unit into a finite number of small units, this leads to the model being discrete in the spatial dimension. It does not have continuity, which is specifically manifested in that the finite element model is composed of many nodes and rod units. As a result, the position of the pressure measuring point set on the surface of the structure during the wind tunnel pressure test may not have a node that can apply load at the same position in the finite element model. Even if there is, there is a small difference in its position. Therefore, in existing studies, the node closest to the measuring point position is usually selected as the load application point. The above defects will bring non-negligible deviations to the analysis results. Summary of the invention

[0005] In order to improve the calculation accuracy of the wind vibration coefficient, the present application provides a method for calculating the wind vibration coefficient of a roof structure.

[0006] The technical solution adopted by the present invention to solve the above problems is:

[0007] The calculation method of wind vibration coefficient of roof structure includes:

[0008] Step 1: Obtain geometric parameters and material property data of the building model;

[0009] Step 2: Establish a finite element elastic-plastic model based on geometric parameters and material property data;

[0010] Step 3, establishing a numerical model and a physical rigid model according to geometric parameters;

[0011] Step 4: Mesh the numerical model;

[0012] Step 5, extracting the node coordinates in the finite element elastic-plastic model;

[0013] Step 6: Arrange the monitoring points in the numerical model according to the node coordinates;

[0014] Step 7: Perform numerical simulation calculation based on the monitoring points;

[0015] Step 8, obtaining the wind pressure time history 1 of each monitoring point in the numerical model;

[0016] Step 9: The physical rigid model arranges measuring points according to the structural characteristics and records the measuring point coordinates;

[0017] Step 10, conducting a wind tunnel test on the physical rigid model;

[0018] Step 11, obtaining the surface wind load of the physical rigid model roof under the flow from different directions, and obtaining the wind pressure time history data 2;

[0019] Step 12: Overlay and merge the numerical model monitoring points with the physical rigid model measuring points. For two measuring points that are close in position, retain the numerical model monitoring points and remove the physical rigid model measuring points.

[0020] Step 13, respectively correspond the wind pressure time history data of the corresponding measuring points one by one, and select the wind pressure time history data corresponding to the physical rigid model for the points that are close in position and retained, thereby obtaining wind pressure time history data 3;

[0021] Step 14, respectively importing wind pressure time history 1, wind pressure time history 2 and wind pressure time history 3 into finite element analysis software and applying them to corresponding points, and using modal iteration method to perform wind-induced vibration elastoplastic analysis to obtain three sets of structural displacement time history curves;

[0022] Step 15, obtaining the displacement pulsation value and average value at each node based on the three sets of structural displacement time history curves;

[0023] Step 16, calculating wind vibration coefficient 1, wind vibration coefficient 2 and wind vibration coefficient 3 respectively based on the data obtained in step 15;

[0024] Step 17, superimposing the wind vibration coefficient 1 and the wind vibration coefficient 2 to obtain the wind vibration coefficient 4, wherein the weighted average value of the overlapping points is taken;

[0025] Step 18: Combine wind vibration coefficient 3 and wind vibration coefficient 4 to obtain an envelope value to obtain wind vibration coefficient 5, which is the final wind vibration coefficient.

[0026] Furthermore, the step 2 also includes: simplifying the finite element elastic-plastic model, omitting the lower main body part, and using the connection part between the roof and the lower main body part as the constraint boundary condition of the finite element elastic-plastic model.

[0027] Furthermore, the step 4 also includes: encrypting the grid at the location where the fluid separation is severe.

[0028] Furthermore, in step 6, when selecting monitoring points, the wind tunnel test measurement point arrangement principle is followed and encrypted processing is performed.

[0029] Furthermore, the encryption processing means that the spacing of measuring points in the flat area should be kept within 1m corresponding to the original building size; in the part where the local morphology changes suddenly, the measuring points are arranged at a spacing of no more than 0.1m on both sides of the top of the protrusion; the curved area is arranged at a spacing of no more than 0.5m along the two-dimensional curve, and the measuring points are arranged at the vertices of the peaks and troughs of the curve; the measuring points are arranged at a spacing of no more than 1m in the direction of linear change.

[0030] Furthermore, the step 7 performs numerical simulation calculation based on transient calculation and LES large eddy simulation turbulence model.

[0031] Furthermore, in step 9, the principle of arranging the measurement points is to encrypt the edges and local mutations and evenly distribute the other positions.

[0032] Furthermore, the method further includes step 19, converting the point distribution wind vibration coefficient into a partition surface distribution wind vibration coefficient by weighted average according to the coverage area.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The numerical simulation and physical wind tunnel test methods are combined through the measurement point arrangement method and the load combination method to study the wind load response, avoiding the inherent defects of the two methods and combining the advantages of the two methods, so that the calculation accuracy of the wind vibration coefficient is higher. The measurement point arrangement method refers to: extracting the unit node coordinates of the building finite element model, and selecting points from the unit nodes as measurement points according to the measurement point arrangement principle. The load response combination method refers to: first, superimposing the pressure measurement points to superimpose the loads, and then loading the superimposed load data into the finite element model for dynamic analysis; second, loading the load data of numerical simulation and wind tunnel test into the finite element model for dynamic analysis, and then superimposing the calculated wind load response; finally, the wind load responses obtained by the two methods are combined to obtain the envelope value to obtain the final response result.

[0035] In terms of measurement point arrangement, the unit coordinates of the finite element model and the geometric shape characteristics of the wind tunnel physical model are combined to make the measurement point arrangement more reasonable and the coverage density higher.

[0036] When conducting finite element elastic-plastic dynamic analysis, the loading position of the load is more reasonable and accurate. The loads loaded at most loading points are the wind load data at the same position, and the number of approximately equivalent loading points is greatly reduced.

[0037] More dense measuring points and more accurate loading positions make the dynamic response results obtained by finite element calculation more accurate, and the wind load response parameters obtained thereby have higher validity and availability.

[0038] It combines the high-density and high-precision data of numerical simulation with the high credibility of wind tunnel tests, and corrects the result errors by exchanging the order of load superposition and response superposition. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of the calculation method of wind vibration coefficient of roof structure;

[0040] Figure 2 This is a top view of the finite element model of a large-span roof structure;

[0041] Figure 3 It is a simplified diagram of the finite element model and the constraint conditions;

[0042] Figure 4 This is a schematic diagram of the preliminary estimated arrangement of measuring points on a roof line according to the measuring point arrangement principle;

[0043] Figure 5 are several unit nodes on the ridge line in the finite element model;

[0044] Figure 6 A schematic diagram of the combination of measuring points and unit nodes for preliminary estimation;

[0045] Figure 7 Schematic diagram of measuring point arrangement in the curve change area;

[0046] Figure 8 Schematic diagram of measuring point layout in different areas;

[0047] Fig. 9 It is a schematic diagram of the arrangement of measuring points on a roof with linear changes in the vertical curve change direction;

[0048] Fig.10 Schematic diagram of the measurement point layout for wind tunnel tests. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0050] like Figure 1 As shown in the figure, the calculation method of wind vibration coefficient of roof structure includes:

[0051] Step 1: Use structural design software (such as PKPM, YJK, etc.) to establish a complete building model of the target complex large-span roof structure. The main load-bearing components such as beams, slabs, and columns are fully restored in the model. The structural design scheme and material properties are defined. The geometric parameters and material properties of the building model are exported for subsequent numerical simulation and dynamic calculation.

[0052] Step 2: Use ANSYS APDL to establish a finite element elastic-plastic model based on the exported building geometry parameters and material properties. The roof load-bearing members use rod elements and are positioned according to the spatial coordinates in the building model, such as Figure 2 As shown in the figure. Since the upper roof part is the main part affected by wind load in the long-span roof structure, while the lower main structure is less affected. Therefore, the finite element model can be simplified, the lower main part is omitted, and the connection between the roof and the lower structure is used as the constraint boundary condition of the finite element model, as shown in the figure. Figure 3 As shown in the figure. Since the elastoplastic model takes into account the dynamic response such as material deformation, it requires not only geometric information but also material properties. Computational fluid dynamics and wind tunnel tests focus more on the load conditions on the surface of the structure, so they mostly use rigid models, so only geometric information modeling is required.

[0053] Step 3: Establish a numerical model and a physical rigid model based on the geometric parameters. Unlike the finite element dynamic analysis which only requires a model of the upper roof part of the component, numerical simulation and physical wind tunnel testing need to measure the load caused by the flow field movement on the building surface in the wind field environment. At this time, the building form will have a great impact on the flow field, so complete building geometric parameters are required to establish the model. Use structural design software to export its geometry file or use CAD to export the iges file of the three-dimensional model, and import the geometry file into common numerical modeling software (such as Gambit, ICEM, Meshing, etc.) to generate a numerical model. At the same time, use 3D printing technology to manufacture a scaled physical model.

[0054] Step 4: Use modeling software to divide the numerical model grid. In order to measure the accurate wind pressure on the building surface, the grid needs to be encrypted where the fluid separation is severe.

[0055] Step 5: Extract the node coordinates of the end points and intersections of the rod unit in the finite element model. Since the concentrated load needs to be loaded on the nodes in the finite element analysis, it is necessary to arrange the monitoring points accordingly according to the unit node coordinates.

[0056] Step 6: Arrange the monitoring points in the numerical model according to the node coordinates to monitor the wind pressure data. Taking the ridge line of a typical part as an example, according to the principle of measuring point arrangement, the locations of the measuring points are preliminarily estimated as follows: Figure 4 As shown in the figure, several unit nodes on the ridge line in the finite element model are as follows Figure 5 As shown in the figure, the locations of the preliminary estimated measurement points are combined with the unit nodes of the finite element model, as shown in the figure. Figure 6 As shown, the unit node that is closest to the two is selected as the location for the final arrangement of the measuring points.

[0057] Since all loads need to be loaded on unit nodes in the final finite element dynamic calculation, when calculating the wind load on the roof surface in numerical simulation, points are directly selected from the unit nodes of the finite element model as measuring points for data collection according to the measuring point arrangement principle to ensure the complete accuracy of the load loading position in the subsequent finite element analysis.

[0058] Furthermore, since the number of upper roof rod unit nodes is much greater than the number of measuring points arranged in conventional wind tunnel tests, the monitoring points should be selected in accordance with the wind tunnel test measuring point arrangement principle and appropriately encrypted. In this embodiment, for large flat areas where the structural form does not change significantly, the measuring point arrangement spacing should be kept within 1m corresponding to the original building size, and the measuring points should be arranged at intervals of no more than 0.1m on both sides of the top of the protrusion at the local morphological mutation site, and the curved surface area should be arranged at intervals of no more than 0.5m along the two-dimensional curve, and measuring points should be arranged at the peaks and troughs of the curve. Figure 7 As shown in the figure, the arrangement principles of measuring points in different areas are as follows: Figure 8As shown in the figure. Since the long-span roof usually changes in a unidirectional curve, it changes linearly along its vertical direction. In the direction of linear change, the measuring points can be arranged with a spacing of no more than 1m, such as Fig. 9 shown.

[0059] Step 7. Import the numerical model grid file in step 4 into Fluent for numerical calculation. Add monitoring points in Fluent according to the measuring point coordinates arranged in step 6 and select pressure as the monitoring sampling data. Use transient calculation to obtain the change of monitoring data over time. Use the LES large eddy simulation turbulence model for simulation calculation to obtain more accurate data results.

[0060] Step 8: Obtain the wind pressure time series 1 of each monitoring point through numerical simulation calculation.

[0061] Step 9. Arrange the measuring points of the physical rigid model made in step 3 according to the structural characteristics. The basic arrangement principle is to encrypt the edges and local mutations and evenly distribute the other positions. Record the coordinates of the measuring points to determine the load application position in the subsequent finite element analysis. Since the test model in the wind tunnel test is usually scaled within the range of 1:50-1:200, the internal space of the scaled model is small, and the measuring points need to drill holes on the surface of the building to insert pressure measuring tubes and connect pressure measuring valves. The narrow space and the large number of pressure measuring tubes and pressure measuring valves placed inside jointly limit the number of measuring points. Therefore, when arranging the measuring points of the test model, only a very small number of measuring points can be selected at each typical position. For example, in the middle raised ridge area commonly seen in large-span curved roofs, usually only one measuring point is arranged at the highest point of the top of the ridge, and then one is arranged in the middle of the slope on both sides. This is considered to be a local encryption process, and for the flat area that occupies most of the roof area, 3-5 points are often arranged, such as Fig.10 As shown. However, the sizes of various cavitation vortices generated by the flow separation on the roof surface are extremely small, and the density of the test measurement points is too low, so it is very likely that the roof wind load extreme value cannot be identified. Therefore, the accuracy of the results of the existing technology using wind tunnel test data for finite element analysis needs to be improved.

[0062] Step 10: Place the physical rigid model in an atmospheric boundary layer wind tunnel for a wind tunnel scale test. Select different terrains according to the topographic conditions of the target building's proposed location. Class B terrain is usually used in towns. Arrange the spoiler wedges and ground friction roughness elements at the front end of the wind tunnel duct according to the selected terrain to generate incoming flow that meets the terrain wind field characteristics.

[0063] Step 11: Since most large-span roof structures have complex geometric shapes and are usually not symmetrical, the flow conditions of the wind on the roof surface under wind fields with different directions of flow vary greatly. Therefore, it is necessary to obtain the surface wind load of the roof under wind flows from different directions. In combination with the structural morphology characteristics, the physical model is rotated. In this embodiment, the model is rotated every 10°, and a total of 36 angles from 10° to 360° are measured to obtain wind pressure time history data 2.

[0064] Step 12: Overlay and merge the numerical model measurement points with the wind tunnel test model measurement points. For two measurement points close to each other, the numerical model measurement points are retained and the test model measurement points are removed, thereby obtaining a measurement point distribution with a higher measurement point coverage density.

[0065] Step 13, the wind pressure time history data of the corresponding measuring points are matched one by one, and the wind pressure time history data measured by the corresponding wind tunnel test are selected for the points that are close in position and retained, thereby obtaining wind pressure time history data 3 with higher data density. The accuracy of the numerical simulation measuring point position and the high credibility of the wind tunnel test data are retained by the superposition of steps 12 and 13, thereby improving the calculation accuracy of the subsequent wind vibration coefficient.

[0066] Step 14, import the three groups of wind pressure time history data into ANSYS respectively and apply them to the corresponding points, and use the modal iteration method to perform wind-induced vibration elastoplastic analysis. Among them, since the measuring point coordinates of wind pressure time history data 1 are arranged according to the coordinates of the unit nodes of the finite element model, the corresponding nodes can be directly found for load loading. Wind pressure time history data 2 is the measuring point coordinates determined according to the morphology of the large-span roof structure. The coordinates are almost impossible to be consistent with the coordinates of the unit nodes of the finite element model. Therefore, only the unit node closest to the measuring point position can be selected as an approximate substitute, which is also the error defect of the traditional method. Wind pressure time history data 3 combines the characteristics of wind pressure time history data 1 and wind pressure time history data 2. The numerical model measuring points in wind pressure time history data 3 are directly applied, and the measuring points of the wind tunnel test also use the approximate substitution method. The distribution point density of this group of data is the largest, and the accuracy of the response calculation results obtained is also the highest.

[0067] Step 15, obtain three sets of structural displacement time history curves. Since the structural response is abrupt when the initial load is applied, it has no statistical significance. It is necessary to wait for the response to gradually stabilize and take the stable period time history data to calculate the displacement pulsation value and average value at each node.

[0068] Step 16, the load effect generated by the average wind is Rs, and the load effect generated by the equivalent wind vibration force caused by the pulsating wind is Rd, then the total response of the structure under the wind load is Ra=Rs+Rd. Three sets of wind vibration coefficients are calculated. The wind vibration coefficient is the ratio of the total wind-induced vibration response to the response generated by the average wind, βz=(Rs+Rd) / Rs=1+Rd / Rs. The response concerned by this method is the structural displacement, and the displacement wind vibration coefficient can be obtained by replacing the load effect with the position. For large-span roof structures, since there are many modes involved in the vibration, the displacement wind vibration coefficient can more accurately reflect the dynamic wind load on the structure. The wind vibration coefficient of each node is calculated based on the displacement pulsation value and average value of each node obtained in step 15. The wind vibration coefficients 1, 2, and 3 are calculated from the three sets of displacement time-history data 1, 2, and 3, respectively.

[0069] Step 17. Since the measurement point arrangement logic of Method 1 and Method 2 is completely different, the measurement points and load application points in Method 1 are based on the coordinates of the intersection nodes of the finite element model units, and each unit is basically subjected to end force, while in Method 2, approximate substitution is used to load the load on the closest unit node. Since each large component unit in the finite element model is divided into a finite number of small units, the nodes of approximate substitution are mostly the middle position of the component rod unit. At this time, each unit is subjected to the concentrated load in the middle. The two force characteristics of the two methods make the load response of the structure different. However, both methods have their advantages and disadvantages. Method 1 is accurate in load loading, and the response results obtained should be more accurate, but because the load it loads is calculated using CFD numerical simulation, the credibility is lower than the test results. The load used in Method 2 is the measurement result of the wind tunnel test, which is more reliable, but there are errors in load loading. The wind vibration coefficient 1 and wind vibration coefficient 2 calculated by the two methods are superimposed to obtain the wind vibration coefficient 4, and the overlapping points are weighted averaged according to 0.5 and 0.5, thereby combining the advantages of the two methods and improving the accuracy and feasibility of the results.

[0070] Step 18, wind vibration coefficient 3 and wind vibration coefficient 4 are obtained by combining the two advantages of accurate numerical simulation loading and reliable wind tunnel test data. However, the core ideas of the two methods are different. Wind vibration coefficient 3 is to superimpose the load first and then calculate the response, while wind vibration coefficient 4 is to calculate the loads separately first and then superimpose the response. Regardless of the order of superposition, the data points of wind vibration coefficient 3 and wind vibration coefficient 4 are the same. For the sake of engineering application safety, wind vibration coefficient 3 and wind vibration coefficient 4 are combined to obtain the envelope value to obtain wind vibration coefficient 5, which is the final wind vibration coefficient.

[0071] Furthermore, due to the large number of points, it is not convenient for actual engineering application. At the same time, due to the large plane size of the large-span roof structure, the wind load and wind-induced response of the structure in a large area with similar geometric shapes are often similar. At this time, listing the data of all the measurement points will no longer have characteristic significance. The roof surface is divided according to the geometric shape characteristics. The basic principle is to divide the same curved surface area into the same area, divide the curved surface at the change, and divide the corner. The area shape is the best square. For irregular curved surfaces, it should be divided into four equal-length areas as much as possible. The typical loads in the same area are similar (for example, for the obvious protruding part in the middle of the roof, when one side is impacted by the incoming flow and shows positive pressure, the other side will show negative pressure wind suction due to the separation of the incoming flow at the top. Therefore, when dividing the area, it should be divided by the top ridge line). In order to provide simple and easy-to-use data, the point distribution wind vibration coefficient βzi is converted into the partition surface distribution wind vibration coefficient βA according to the weighted average of the coverage area. βA=ΣβziAi / ΣAi, where: Ai is the area covered by each measuring point, ΣAi is the sum of the areas of each measuring point in the partition, and βzi is the wind vibration coefficient calculated at each measuring point.

[0072] Through the above research method, the wind vibration coefficients of different parts of the complex large-span roof structure surface under different wind direction angles can be finally obtained. When the wind vibration coefficients obtained by this method are used to calculate the relevant equivalent static wind load and other key control indicators of the wind-resistant design of large-span roof structures, more accurate wind load response results can be obtained, which can significantly improve the wind-resistant design performance of the structure.

Claims

1. The calculation method of wind vibration coefficient of roof structure is characterized by: include: Step 1: Obtain geometric parameters and material property data of the building model; Step 2: Establish a finite element elastic-plastic model based on geometric parameters and material property data; Step 3, establishing a numerical model and a physical rigid model according to geometric parameters; Step 4: Mesh the numerical model; Step 5, extracting the node coordinates in the finite element elastic-plastic model; Step 6: Arrange the monitoring points in the numerical model according to the node coordinates; Step 7: Perform numerical simulation calculation based on the monitoring points; Step 8, obtaining the wind pressure time history 1 of each monitoring point in the numerical model; Step 9: The physical rigid model arranges measuring points according to the structural characteristics and records the measuring point coordinates; Step 10, conducting a wind tunnel test on the physical rigid model; Step 11, obtaining the surface wind load of the physical rigid model roof under the flow from different directions, and obtaining the wind pressure time history data 2; Step 12: Overlay and merge the numerical model monitoring points with the physical rigid model measuring points. For two measuring points that are close in position, retain the numerical model monitoring points and remove the physical rigid model measuring points. Step 13, respectively correspond the wind pressure time history data of the corresponding measuring points one by one, and select the wind pressure time history data corresponding to the physical rigid model for the points that are close in position and retained, thereby obtaining wind pressure time history data 3; Step 14, respectively importing wind pressure time history 1, wind pressure time history 2 and wind pressure time history 3 into finite element analysis software and applying them to corresponding points, and using modal iteration method to perform wind-induced vibration elastoplastic analysis to obtain three sets of structural displacement time history curves; Step 15, obtaining the displacement pulsation value and average value at each node based on the three sets of structural displacement time history curves; Step 16, calculating wind vibration coefficient 1, wind vibration coefficient 2 and wind vibration coefficient 3 respectively based on the data obtained in step 15; Step 17, superimposing the wind vibration coefficient 1 and the wind vibration coefficient 2 to obtain the wind vibration coefficient 4, wherein the weighted average value of the overlapping points is taken; Step 18: Combine wind vibration coefficient 3 and wind vibration coefficient 4 to obtain an envelope value to obtain wind vibration coefficient 5, which is the final wind vibration coefficient.

2. The method for calculating the wind vibration coefficient of a roof structure according to claim 1, characterized in that: The step 2 also includes: simplifying the finite element elastic-plastic model, omitting the lower main body part, and using the connection between the roof and the lower main body part as the constraint boundary condition of the finite element elastic-plastic model.

3. The method for calculating the wind vibration coefficient of a roof structure according to claim 1, characterized in that: The step 4 also includes: encrypting the grid at the location where the fluid separation is severe.

4. The method for calculating the wind vibration coefficient of a roof structure according to claim 1, characterized in that: In step 6, when selecting monitoring points, the wind tunnel test measurement point arrangement principle is followed and encrypted.

5. The method for calculating the wind vibration coefficient of a roof structure according to claim 4, characterized in that: The encryption processing means that the spacing of measuring points in flat areas should be kept within 1m corresponding to the original building size; in parts where local morphology changes suddenly, measuring points are arranged at a spacing of no more than 0.1m on both sides of the top of the protrusion; in curved areas, measuring points are arranged at a spacing of no more than 0.5m along the two-dimensional curve, and at the vertices of the curve's peaks and troughs; in the direction of linear change, measuring points are arranged at a spacing of no more than 1m.

6. The method for calculating the wind vibration coefficient of a roof structure according to claim 1, characterized in that: The step 7 performs numerical simulation calculation based on transient calculation and LES large eddy simulation turbulence model.

7. The method for calculating the wind vibration coefficient of a roof structure according to claim 1, characterized in that: In step 9, the principle of arranging the measuring points is to encrypt the edges and local mutations and evenly distribute the other positions.

8. The method for calculating the wind vibration coefficient of a roof structure according to any one of claims 1 to 7, characterized in that: The step 19 is also included, converting the point distribution wind vibration coefficient into the partition surface distribution wind vibration coefficient by weighted average according to the coverage area.

Images