Method for obtaining wind pressure distribution of near-earth wind-sensitive photovoltaic structure through separation vortex simulation

Abstract

The invention provides a method for obtaining wind pressure distribution of a near-earth wind-sensitive photovoltaic structure through separation vortex simulation. According to the separation vortex simulation method, an RANS method is adopted in a boundary layer area, an LES method is adopted in other areas, and a delay function is introduced to adjust the turbulence length scale to ensure that transition from the RANS to the LES is correctly achieved according to flow characteristics; and in order to compensate for attenuation of the turbulence characteristic of the wind field, an input spectrum is compensated, and a turbulence inlet is artificially synthesized by adopting an NSRFG method, so that the self-retentivity of the turbulence is improved. Compared with the prior art, the separation vortex simulation method provided by the invention is simple in implementation process, relatively high in precision while relatively small calculated amount is maintained, and therefore, the method has relatively good engineering application prospects.

Description

Technical Field

[0001] This application relates to the field of structural engineering, specifically a method for obtaining the wind pressure distribution law on the surface of near-ground wind-sensitive photovoltaic structures by using the detached eddy numerical simulation method. Technical Background

[0002] Solar photovoltaic power generation is an important part of China's renewable energy structure. Most of the areas with rich solar energy resources in China are located in desert areas with strong wind and sand. Therefore, the service environment of photovoltaic structures has high wind speed and strong turbulence. Although the height of photovoltaic structures is low, they generally have a large span and a small cross-section height, making them typical wind-sensitive structures. As a result, under the action of wind loads, photovoltaic modules will experience significant wind-induced vibrations and even be damaged or collapsed under strong winds. Therefore, with the wide application of photovoltaic power generation, the research on the wind resistance performance of photovoltaic structures has attracted more and more attention from scientific researchers and engineering designers.

[0003] To conduct wind resistance analysis and design of a photovoltaic structure, it is first necessary to understand the wind pressure distribution law on the surface of the photovoltaic structure. Currently, the methods for obtaining wind pressure distribution include wind tunnel tests and numerical simulations. The implementation process of wind tunnel tests is as follows: Combining the cross-sectional dimensions of the wind tunnel test section, determine an appropriate structural scale ratio, fabricate a scaled model, and through rigid pressure measurement or force measurement wind tunnel test methods, obtain the wind pressure time history at different inclination angles of different photovoltaic modules under different wind direction angles, and further obtain the wind load shape coefficient, wind vibration coefficient, etc.; or fabricate a scaled aeroelastic model, measure the wind-induced displacement response of the photovoltaic aeroelastic model at different wind speeds, different wind direction angles, and different inclination angles, and then obtain the wind vibration coefficient, etc. However, in addition to being time-consuming, laborious, and having limited simulation conditions, wind tunnel tests also have drawbacks such as scale effect and long cycle. With the rapid development of high-performance supercomputing technology, the CFD numerical simulation method has been widely used in the analysis of wind loads on the structure surface due to its higher flexibility, wider application range, and lower time and cost compared to experimental methods. Currently, the CFD numerical simulation methods that have been successfully applied to wind engineering research are RANS (Reynolds-averaged Navier-Stokes) and LES (Large Eddy Simulation). The RANS method uses the Reynolds-averaged flow equation and empirical turbulence models to solve the mean flow field and turbulence field, and is suitable for steady turbulent flow and flow problems with high Reynolds numbers. Its calculation cost is relatively low and it is widely used in the engineering field. The LES method divides turbulence into large-scale vortices and small-scale vortices. The large-scale vortices are solved by direct numerical simulation, and the small-scale vortices are closed by turbulence models to simulate the influence on the large-scale vortices. However, the RANS model has difficulties in capturing vortex structures and large-scale turbulence because it cannot effectively capture dynamic scale changes and irregular turbulence phenomena. In contrast, LES can better capture the vortex structures in the flow field and obtain richer wind load information. However, under high Reynolds number conditions, a very high-resolution grid is required in the flow region near the structure wall, otherwise the credibility of the results calculated by LES is very low; moreover, to divide a grid that meets the requirements of LES, the height of the first layer of the grid is very small, the number of grids is huge, the calculation efficiency is extremely low, and the numerical stability is poor. Therefore, the Detached-Eddy Simulation (DES) method has been developed. This method combines RANS and LES. It uses the LES method at larger scales to more accurately simulate large-scale structures; and uses the RANS method at smaller scales to save computing resources. However, in the numerical simulation of near-wall flow, especially when the wall grid configuration is inappropriate, the turbulent length scale in the DES method may be misjudged, causing the DES to transform into LES within the boundary layer, which will greatly reduce the turbulent viscosity, resulting in the generation of vortices and thus triggering grid separation phenomena.

[0004] The complexity of the near-surface airflow is mainly reflected in its turbulent characteristics at a certain height. The influence on near-ground photovoltaic structures is often more complex and intense than that on structures at other heights. This is mainly because the near-surface airflow is in a complex turbulent state with a relatively high turbulent intensity, and its variation with time and spatial position is very significant. Therefore, when using numerical simulation methods to analyze the wind loads on the surface of photovoltaic structures, accurately simulating the turbulent inlet conditions is crucial. In addition, the self-preservation of near-surface turbulence, that is, the ability of the statistical characteristics of the inlet turbulence to be continuously maintained in the flow field, also directly affects the accuracy of the simulation results. Since the detached eddy simulation (DES) method uses large eddy simulation (LES) to solve the flow field far from the structure, and the sub-grid turbulence model in LES has insufficient accuracy in simulating the motion of tiny vortices smaller than the filtering scale. These factors may cause the turbulent characteristics in the computational domain to change along the flow direction, making the turbulent characteristics at the target position inconsistent with those at the inlet. To ensure that the calculation results are consistent with the actual situation, it is necessary to strictly verify the self-preservation of the turbulent inlet and compensate it according to the actual situation. However, the existing compensation methods are relatively complex and difficult to implement.

[0005] Therefore, the limitations of existing research lead to the following problems:

[0006] (1) Since the DES method uses the Reynolds-averaged Navier-Stokes (RANS) method in the boundary layer region and the LES method in the remaining regions, for photovoltaic structures serving near the ground, it is necessary to correctly implement the transition from RANS to LES according to the characteristics of their separated flow.

[0007] (2) Although various turbulent synthesis methods have been proposed to improve the retention of the boundary layer wind field along the computational domain, in the actual engineering application process, the synthesized turbulent wind field still has the problem of obvious attenuation along the computational domain. Moreover, for photovoltaic structures serving near the ground, the turbulent intensity of the wind field is greater, and a simple and easy-to-implement compensation method is needed to ensure the turbulent characteristics of the wind field.

[0008] In view of this, the present invention proposes to use a delay function to adjust the turbulent length scale to ensure the use of the RANS method within the boundary layer of the photovoltaic structure. After compensating the input spectrum, the artificial synthesis of the turbulent inlet is carried out using the NSRFG method, so as to obtain the wind pressure distribution law on the surface of the near-ground wind-sensitive photovoltaic structure by using the detached eddy simulation method. Summary of the Invention

[0009] The present invention proposes a method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation. This method can flexibly adjust through the introduced delay function, select RANS or LES simulation according to the characteristics of the separated flow, so as to obtain more accurate separated flow simulation results; and by compensating the input spectrum and using artificial synthesis of the turbulent inlet, the self-preservation of turbulence is improved, so that the pulsating wind effect on the surface of the near-ground wind-sensitive photovoltaic structure can be accurately simulated.

[0010] To achieve the above technical objectives, the present invention provides a method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, which is characterized by specifically including the following steps:

[0011] S1. First, establish a numerical model of the photovoltaic structure according to its size, then set the computational domain of the wind field and the BOI area that needs to be locally refined, and use a hybrid grid division strategy to divide the grid of the computational domain. Unstructured grids are used in the area where turbulence is significant near the building, and structured grids are used far from the building, and the BOI area is locally refined;

[0012] S2. Set the boundary conditions of the computational domain. The inlet selects the velocity inlet boundary condition, the outlet selects the pressure outlet or the fully developed outflow boundary condition, the ground and the building surface select the no-slip wall boundary condition, and the top and both sides of the computational domain select the free-slip boundary condition;

[0013] S3. Use the delayed detached eddy simulation method after introducing a delay function in the detached eddy simulation method, select an appropriate turbulence model, and correctly select RANS or LES simulation according to the characteristics of the separated flow based on the control of the new turbulence length scale d DDES ;

[0014] The introduced delay function satisfies the following expression:

[0015] f d = 1 - tanh([8r d 3 )

[0016]

[0017] where r d is the ratio of the turbulence length scale to the distance to the wall; v t is the kinematic viscosity coefficient of the fluid molecules;

[0018] v is the fluid molecular viscosity coefficient; U i,j is the velocity gradient; k is the Karman constant; d is the turbulence length scale;

[0019] The new turbulence length scale d DDES is defined as:

[0020] d DDES = d - f d max(0, d - C DES Δ max );

[0021] where Δ max is the filtering scale in the large eddy simulation method​

[0022] C DES is an empirical constant, taking the value of 0.75 in Realizable k-ε;

[0023] S4. Generate a turbulent inlet based on the principle of the NSRFG method and verify the self-preservation of the turbulent inlet before calculation, that is, compare the pulsating wind speed power spectrum at the obtained structure with the Karman spectrum. If they are basically in agreement, it indicates that the self-preservation of the mean wind speed and turbulence intensity profiles generated by this method is very good; if not, use a compensation method to compensate and correct the input spectrum so that the pulsating wind speed power spectrum at the high-frequency part coincides or approximately coincides with the Karman spectrum when it reaches the structure;

[0024] S5. Adopt the delayed detached eddy simulation method and connect the corrected turbulent inlet to the fluent inlet, so as to calculate the time history of the pulsating wind pressure of the near-ground wind-sensitive photovoltaic structure at different wind direction angles and different angles of attack, and further obtain the time history of the total wind pressure; when not considering the coupling effect between the wind and the structure, the total wind pressure on the surface of the photovoltaic structure satisfies the following formula:

[0025]

[0026] In the formula: W i (t) is the total wind pressure at point i at time t; ρ is the air density, taking 1.29 Kg / m3;

[0027] is the mean wind speed at point i; w i (t) is the pulsating wind pressure at point i at time t.

[0028] In the preferred technical solution of the present invention: in the S1 step, establish its rigid model in the modeling software Rhino according to the size of the photovoltaic structure; set the calculation domain of the wind field and the BOI area that needs to be locally refined in SpaceClaim.

[0029] In the further technical solution of the present invention: in the S1 step, the selection of the calculation domain must first meet the requirement of the blockage ratio, and the blockage ratio should be less than 3%. The blockage ratio ρ satisfies the following formula:

[0030]

[0031] In the formula: A b is the maximum windward area; A d is the cross-sectional area of the calculation domain.

[0032] Secondly, it is necessary to meet the requirement that the top of the building is not less than 5H from the upper boundary, the two side surfaces are not less than 5H from the boundary, the building is not less than 5H from the inlet, and not less than 15H from the outlet, where H is the height of the building.

[0033] A preferred technical solution of the present invention: In the S1 step, unstructured tetrahedral meshes with stronger adaptability to fine features are adopted on the surface and nearby areas of the photovoltaic structure, and structured hexahedral meshes are adopted on the outside, that is, a computational domain mesh is established in a nested manner of inner and outer domains; local refinement is performed on the flow field near the structure, that is, the BOI region, to ensure that the mesh near the photovoltaic structure model in the inner computational domain is refined, and the mesh in the outer computational domain is set slightly sparser in the direction away from the model, and the inner and outer domain meshes transition smoothly.

[0034] A preferred technical solution of the present invention: In the S2 step, the boundary conditions of the computational domain are set. The boundary conditions of the computational domain are set such that the left end boundary is a velocity inlet, and the right end boundary is a pressure outlet or a fully developed outflow boundary condition; the two sides and the top of the computational domain are free-slip wall surfaces; the bottom of the computational domain and the surface of the photovoltaic structure are no-slip wall boundary surfaces, which conforms to the actual wind field situation.

[0035] A further technical solution of the present invention: In the S3 step, through the adjustment of the introduced delay function, and then based on the control of the new turbulent length scale d DDES , the detached eddy simulation method can correctly select RANS or LES simulations according to the characteristics of the separated flow; when d DDES is less than C DES Δ max , then the RANS method is used to solve the motion of the fluid in this region; when the turbulent length scale d DDES is greater than C DES Δ max , then the LES method is used to solve the motion of the fluid.

[0036] A preferred technical solution of the present invention: In the S3 step, the turbulence model adopts the Realizable k-ε model; the non-linear convection term is discretized by adopting the second-order upwind scheme; the momentum equation is solved by adopting the bounded central difference scheme; in terms of time discretization, the second-order fully implicit scheme is selected; the pressure-velocity coupling equation adopts the SIMPLEC algorithm, so as to promote the early development of the turbulent flow field in the detached eddy simulation and improve the convergence of the simulation; then the time step, the number of time steps, the maximum number of iterations and the simulation duration are set according to the requirements.

[0037] A further technical solution of the present invention: In the S4 step, the compensation of the input spectrum satisfies the following formula:

[0038]

[0039] In the formula: S is the input spectrum; f is the frequency; f min is the minimum frequency value of the pulsating wind speed power spectrum;

[0040] f maxis the maximum frequency value of the pulsating wind speed power spectrum;

[0041] A further technical solution of the present invention: in step S4, the NSRFG method is used to generate the pulsating turbulent inlet of the three-dimensional along-wind, cross-wind and vertical pulsating wind fields. The three-dimensional pulsating wind speed based on the Karman spectrum in the atmospheric boundary layer satisfies the following expression:

[0042]

[0043]

[0044] In the formula: f is the frequency; S u (f), S v (f) and S w (f) are the power spectral density functions of the along-wind, cross-wind and vertical pulsating wind speeds respectively; U ɑv is the mean wind speed; L u , L v , L w are the turbulence integral scales in the three directions respectively.

[0045] A further technical solution of the present invention: in step S4, based on the principle of the NSRFG method, basic wind field parameters including the mean wind speed, turbulence intensity and turbulence integral scale need to be input in advance to generate the turbulent inlet;

[0046] The mean wind speed profile satisfies the following formula:

[0047]

[0048] In the formula: v z is the mean wind speed at the height z above the ground; v 10 is the mean wind speed at the height of 10 m above the ground; the roughness index α of class B terrain is 0.15; the basic wind pressure with a return period of 50 years in the local area is w. According to w = v 2 / 1600, v 10 is calculated;

[0049] According to the building load code and specifications, the turbulence intensity and turbulence integral scale in the along-wind, cross-wind and vertical directions satisfy the following expressions:

[0050]

[0051] In the formula: I 10 is the nominal turbulence intensity at the height of 10 m, which is related to the surface roughness. For class B terrain, the value is 0.14; ɑ takes the value of 0.15. σ u , σ v and σ w are the root mean squares of the wind speeds in the along-wind, cross-wind and vertical directions respectively; when there is no measured data, Iv =0.88I u , I w =0.55I u ;

[0052]

[0053] Where: z0 is the ground roughness parameter, which is taken as 0.05 for Class B landform; σ v / σ u and σ w / σ u The value is determined according to ESDU85020.

[0054] Beneficial effects of the present invention:

[0055] The present invention provides a method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation. The method first uses the detached eddy simulation method to adjust the turbulence length scale through a delay function, thereby flexibly selecting RANS or LES simulation according to the characteristics of the detached flow, and more accurately simulating the wind pressure distribution on the surface of the photovoltaic structure and the flow field around it, thereby effectively improving the accuracy of the numerical simulation; by compensating and correcting the turbulent wind speed power spectrum, the NSRFG method is used to generate a turbulent inlet, thereby ensuring that the wind pressure simulation result on the surface of the near-ground wind-sensitive photovoltaic structure is consistent with the actual turbulence characteristics, thereby reducing the turbulence attenuation problem and ensuring the accurate development of the wind field in the calculation domain; finally, the above method is used to numerically simulate the wind field around the photovoltaic structure, thereby obtaining the wind pressure distribution on the surface of the near-ground wind-sensitive photovoltaic structure under different wind direction angles and different attack angles.

[0056] Since the present invention refers to the delay function, it can reduce the amount of calculation while still providing more accurate separation flow simulation results, thereby ensuring that the separation eddy simulation method can be implemented on the near-ground wind-sensitive photovoltaic structure; at the same time, by performing a simple and easy-to-implement correction method on the input spectrum, it is ensured that the generated turbulent inlet has good self-sustaining properties, thereby more accurately reflecting the wind pressure distribution of the near-ground photovoltaic structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the implementation methods of the present application, the drawings required for use in the implementation methods will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 A flowchart for realizing the generation of a turbulent inlet based on the principle of the NSRFG method according to an embodiment of the present invention;

[0059] Figure 2a Front view of the numerical model of the 0° angle of attack near - ground wind - sensitive photovoltaic structure according to an embodiment of the present invention;

[0060] Figure 2b Side view of the numerical model of the 0° angle of attack near - ground wind - sensitive photovoltaic structure according to an embodiment of the present invention;

[0061] Figure 2c Three - dimensional view of the numerical model of the 0° angle of attack near - ground wind - sensitive photovoltaic structure according to an embodiment of the present invention;

[0062] Figure 3 Schematic diagram of the computational domain according to an embodiment of the present invention;

[0063] Figure 4 Comparison diagram of the power spectrum of the pulsating wind speed at the inlet and the Karman spectrum for verifying the self - preservation of synthetic turbulence according to an embodiment of the present invention;

[0064] Figure 5 Comparison diagram of the power spectrum of the pulsating wind speed at the structure before and after compensation and the Karman spectrum for verifying the self - preservation of synthetic turbulence according to an embodiment of the present invention;

[0065] Figure 6 Time - history diagram of the total wind pressure at a certain monitoring point on the surface of the near - ground wind - sensitive photovoltaic structure at 0° wind direction angle and 0° angle of attack according to an embodiment of the present invention. Detailed implementation manners

[0066] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0067] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application claimed, but merely represents the selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0068] An embodiment provides a method for obtaining the wind pressure distribution of a near - ground wind - sensitive photovoltaic structure by using detached eddy simulation. Taking the classical near - ground wind - sensitive flat single - axis tracking photovoltaic structure as an example, the wind pressure distribution at 0° wind direction angle and 0° angle of attack is calculated. The photovoltaic structure includes nine columns and two photovoltaic panels connected by beams, and its numerical model is shown in Figure 2; the detailed parameters of the photovoltaic structure are shown in Table 1.

[0069] Table 1 Detailed parameter table of the photovoltaic structure

[0070]

[0071] Specifically, it includes the following steps:

[0072] S1. Use the modeling software Rhino to establish a full-scale rigid model of the photovoltaic structure according to the detailed dimensions of the photovoltaic structure in Table 1 above; set the calculation domain size of the wind field in SpaceClaim to 140×200×80 m, and calculate the blockage ratio ρ according to the following formula:

[0073]

[0074] In the formula: A b is the maximum windward area; A d is the cross-sectional area of the calculation domain.

[0075] The obtained blockage ratio is 2.4%. The selection of the calculation domain primarily meets the requirement of the blockage ratio, and the blockage ratio should be less than 3%. The obtained blockage ratio meets the requirement; at the same time, the top of the photovoltaic structure is not less than 5H from the upper boundary, the two side surfaces are not less than 5H from the boundary, the building is not less than 5H from the entrance, and not less than 15H from the exit, where H is the height of the highest part of the photovoltaic structure. The calculation domain is as Figure 3 shown. Then, set the size of the BOI area that needs local encryption to 80×25×15 m.

[0076] Perform mesh division on the calculation domain. The mesh type uses poly-hexcore polyhedron meshes. On the surface and near the photovoltaic structure, use unstructured tetrahedron meshes that are more adaptable to fine features, and use structured hexahedron meshes on the outside, that is, establish the calculation domain mesh in the way of nesting the inner and outer domains. Locally encrypt the flow field near the structure, that is, the BOI area. The mesh size in this area and the mesh size in the near-wall boundary layer are both crucial for ensuring the simulation accuracy. For the first layer height of the near-wall mesh on the surface of the photovoltaic structure, it is 0.05 m, the growth rate is 1.15, the number of boundary layer meshes is 10 layers, and the dimensionless distance y + is between 30 and 100, meeting the requirements of the scalable wall function; the skewness coefficient of the surface mesh is less than 0.70, and the minimum orthogonal quality of the mesh is greater than 0.15, meeting the calculation requirements.

[0077] S2. Set the boundary conditions of the calculation domain. Select the velocity inlet boundary condition for the inlet, the pressure outlet for the outlet, the no-slip wall boundary condition for the ground and the building surface, and the free-slip boundary condition for the top and two side surfaces of the calculation domain, which conforms to the actual wind field situation.

[0078] S3. To ensure that the DES method can correctly achieve the proper transition from RANS to LES, the delayed detached eddy simulation method after introducing a delay function is adopted using the detached eddy simulation method, and the turbulence model uses the Realizable k-ε model; the nonlinear convection term is discretized by adopting the second-order upwind scheme; the bounded central difference scheme is used for solving the momentum equation; in terms of time discretization, the second-order fully implicit scheme is selected; the SIMPLEC algorithm is used for the pressure-velocity coupling equation; the number of time steps is set to 1200 steps, the time step size is 0.1 s, and the total simulation duration is 2 min.

[0079] The key to the detached eddy simulation method lies in the mutual conversion between RANS and LES, which is mainly controlled by the turbulence length scale d. When d is less than C DES Δ max , the RANS method is used to solve the fluid motion in this region; when the turbulence length scale d is greater than C DES Δ max , the LES method is used to solve the fluid motion. Among them, Δ max is the filtering scale in the large eddy simulation method, and C DES is an empirical constant (taking a value of 0.75 in Realizable k-ε).

[0080] The introduced delay function satisfies the following expression:

[0081] f d = 1 - tanh([8r d 3 )

[0082]

[0083] In the formula, r d is the ratio of the turbulence length scale to the distance to the wall; v t is the kinematic viscosity coefficient of fluid molecules; v is the viscosity coefficient of fluid molecules; U i,j is the velocity gradient; k is the Karman constant; d is the turbulence length scale.

[0084] The new turbulence length scale d DDES is defined as:

[0085] d DDES = d - f d max(0, d - C DES Δ max )

[0086] Through the flexible adjustment of the introduced delay function, and then based on the new turbulence length scale d DDES ​For control, the DES method can correctly select RANS or LES simulations according to the characteristics of separated flows. When d DDES is less than C DES Δ max , the RANS method is used to solve the fluid motion in this region; when the turbulent length scale d DDES is greater than C DES Δ max , the LES method is used to solve the fluid motion, where Δ max is the filtering scale in the large eddy simulation method, and C DES is an empirical constant (taking the value of 0.75 in Realizable k-ε).

[0087] S4. Write the turbulent inlet based on the principle of the NSRFG method. The basic wind field parameters of the location where the photovoltaic structure is located need to be input in advance, as shown in Table 2.

[0088] Table 2 Basic wind field parameters of the location where the photovoltaic structure is located

[0089]

[0090] Compile and connect the UDF of the turbulent inlet according to the flow chart with the fluent inlet, and the time history of the pulsating wind speed in the x, y, and z directions can be obtained.

[0091] However, there is a problem that the turbulent wind field synthesized by this method decays significantly along the computational domain, and the turbulent intensity at the near-ground end is large. The photovoltaic structure is a thin plate with a large angle of attack and is extremely sensitive to wind. It is necessary to verify the self-preservation of the turbulent inlet before calculation. Establish an empty wind field with a size of 140×200×80m, select a point at the structure as the monitoring point, calculate according to the above method, and obtain the power spectrum of the pulsating wind speed at this point. Compare the power spectrum of the pulsating wind speed at the obtained structure with the Karman spectrum. The comparison diagrams are as shown in Figure 4 and Figure 5 . It is found that there is a large error between the power spectrum of the pulsating wind speed at this point and the Karman spectrum in the high-frequency part. To ensure the accuracy of the calculation results, the input spectrum is compensated and corrected by the following formula, and the verification is carried out again with the corrected input spectrum. The verification results are as shown in Figure 5 . It is found that the high-frequency part of the power spectrum of the pulsating wind speed at the structure is approximately consistent with the Karman spectrum, which is more in line with the actual wind field situation, and thus more accurate calculation results are obtained.

[0092]

[0093] In the formula: S is the input spectrum; f is the frequency; f min is the minimum frequency value of the power spectrum of the pulsating wind speed, which is taken as 0 here; f maxis the maximum frequency value of the pulsating wind speed power spectrum, and the value here is 10.

[0094] S5. The wind pressure distribution of the near-ground wind-sensitive photovoltaic structure at a wind direction angle of 0° and an angle of attack of 0° was analyzed. The delayed detached eddy simulation method with a delay function was adopted, and the corrected turbulent inlet was connected to the fluent inlet. Other settings were made according to the steps in S3, so that the time history of the pulsating wind pressure of the photovoltaic structure at a wind direction angle of 0° and an angle of attack of 0° could be calculated more accurately, and further the time history of the total wind pressure could be obtained, thus providing parameter data for the subsequent analysis of the wind-induced response of the photovoltaic structure. The time history of the total wind pressure at one monitoring point of the near-ground wind-sensitive photovoltaic structure at a wind direction angle of 0° and an angle of attack of 0° is as Figure 6 shown. When the coupling effect between the wind and the structure is not considered, the total wind pressure on the surface of the photovoltaic structure satisfies the following formula:

[0095]

[0096] In the formula: W i (t) is the total wind pressure at point i at time t; ρ is the air density, taken as 1.29 Kg / m3; is the average wind speed at point i; w i (t) is the pulsating wind pressure at point i at time t.

[0097] The present invention adjusts the turbulent length scale through a delay function, flexibly selects RANS or LES simulation according to the characteristics of separated flow, and more accurately simulates the wind pressure distribution on the surface of the photovoltaic structure and the flow field around it, effectively improving the accuracy of numerical simulation; by compensating and correcting the turbulent wind speed power spectrum and generating a turbulent inlet using the NSRFG method, it ensures that the simulated results of the wind pressure on the surface of the near-ground wind-sensitive photovoltaic structure are more consistent with the actual turbulent characteristics, thereby reducing the problem of turbulent decay and ensuring the accurate development of the wind field within the computational domain; finally, the above method can be used to numerically simulate the wind field around the photovoltaic structure, so as to obtain the wind pressure distribution on the surface of the near-ground wind-sensitive photovoltaic structure at different wind direction angles and different angles of attack.

[0098] The above is only an embodiment of the present invention, and its description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the appended claims.

Claims

1. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that, Specifically, it includes the following steps: S1. First, establish a numerical model according to the size of the photovoltaic structure, then set the calculation domain of the wind field and the BOI area that needs local refinement, and use a hybrid grid division strategy to divide the grid of the calculation domain. Unstructured grids are used in the areas with significant turbulence near the building, and structured grids are used far from the building, and the BOI area is locally refined; S2. Set the boundary conditions of the calculation domain. The velocity inlet boundary condition is selected for the inlet, the pressure outlet or the fully developed outflow boundary condition is selected for the outlet, the no-slip wall boundary condition is selected for the ground and the building surface, and the free-slip boundary condition is selected for the top and both sides of the calculation domain; S3. The delayed detached eddy simulation method after introducing a delay function using the detached eddy simulation method, select an appropriate turbulence model, and based on the new turbulence length scale d DDES for control, correctly select RANS or LES simulation according to the characteristics of separated flow; The introduced delay function satisfies the following expression: f d = 1 - tanh([[8r d 3 )​ where r d is the ratio of the turbulent length scale to the distance to the wall; v t is the kinematic viscosity coefficient of the fluid molecules; v is the fluid molecular viscosity coefficient; U i,j is the velocity gradient; k is the Karman constant; d is the turbulent length scale; New turbulent length scale d DDES is defined as: d DDES = d - f d max(0, d - C DES Δ max ); where, Δ max is the filtering scale in the large eddy simulation method C DES is an empirical constant and takes a value of 0.75 in Realizable k-ε; S4. Generate a turbulent inlet based on the principle of the NSRFG method, and verify the self-preservation of the turbulent inlet before calculation, that is, compare the pulsating wind speed power spectrum at the obtained structure with the Karman spectrum. If they are basically consistent, it means that the self-preservation of the mean wind speed and turbulent intensity profiles generated by this method is very good; if they are not consistent, use a compensation method to compensate and correct the input spectrum so that the pulsating wind speed power spectrum at the high frequency when it reaches the structure coincides or approximately coincides with the Karman spectrum; S5. Adopt the delayed detached eddy simulation method, and connect the corrected turbulent inlet to the fluent inlet, so as to calculate the time history of the pulsating wind pressure of the near-ground wind-sensitive photovoltaic structure at different wind direction angles and different attack angles, and further obtain the time history of the total wind pressure; when the coupling effect between the wind and the structure is not considered, the total wind pressure on the surface of the photovoltaic structure satisfies the following formula: Where: W i (t) is the total wind pressure at point i at time t; ρ is the air density, taken as 1.29 Kg / m3; is the average wind speed at point i; w i (t) is the fluctuating wind pressure at point i at time t.

2. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In the step S1, establish its rigid model in the modeling software Rhino according to the size of the photovoltaic structure; set the calculation domain of the wind field and the BOI area that needs local refinement in SpaceClaim.

3. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation according to claim 1, characterized in that: In the step S1, the selection of the calculation domain first meets the requirement of the blockage ratio, and the blockage ratio should be less than 3%. The blockage ratio ρ satisfies the following formula: Where: A b is the maximum windward area; A d is the cross-sectional area of the computational domain. Secondly, it is necessary to meet the requirement that the top of the building is not less than 5H from the upper boundary, the two sides are not less than 5H from the boundary, the building is not less than 5H from the inlet, and not less than 15H from the outlet, where H is the height of the building.

4. A method for obtaining the wind pressure distribution of a near - ground wind - sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In the step S1, unstructured tetrahedral grids that are more adaptable to fine features are used on the surface and near areas of the photovoltaic structure, and structured hexahedral grids are used outside, that is, the calculation domain grid is established in a nested manner of inner and outer domains; local refinement is carried out on the flow field near the structure, that is, the BOI area, to ensure that the grids near the photovoltaic structure model in the inner calculation domain are refined, and the grids in the direction far from the model in the outer calculation domain are set slightly sparser, and the inner and outer domain grids transition smoothly.

5. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In the step S2, set the boundary conditions of the calculation domain. The boundary conditions of the calculation domain are set as the left end boundary is the velocity inlet, and the right end boundary is the pressure outlet or the fully developed outflow boundary condition; the two sides and the top of the calculation domain are free-slip walls; the bottom of the calculation domain and the surface of the photovoltaic structure are no-slip wall boundaries, which conforms to the actual wind field situation.

6. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In the step S3, through the adjustment of the introduced delay function and based on the new turbulent length scale d DDES control, the detached eddy simulation method can correctly select the RANS or LES simulation according to the characteristics of the separated flow; when d DDES is less than C DES Δ max , the RANS method is used to solve the fluid motion in this region; when the turbulent length scale d DDES is greater than C DES Δ max , the LES method is used to solve the fluid motion.

7. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In step S3, the realizable k-ε turbulence model is adopted; the non-linear convection term is discretized by using the second-order upwind scheme; the bounded central difference scheme is used for solving the momentum equation; in terms of time discretization, the second-order fully implicit scheme is selected; the SIMPLEC algorithm is adopted for the pressure-velocity coupling equation, so as to promote the early development of the turbulent flow field in the separated vortex simulation and improve the convergence of the simulation; then, the time step, the number of time steps, the maximum number of iteration steps and the simulation duration are set according to requirements.

8. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation, characterized in that: In step S4, the compensation for the input spectrum satisfies the following formula: Where: S is the input spectrum; f is the frequency; f min is the minimum frequency value of the fluctuating wind speed power spectrum; f max is the maximum frequency value of the pulsating wind speed power spectrum.

9. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by detached eddy simulation according to claim 1 or 8, characterized in that: In step S4, the NSRFG method is used to generate the pulsating turbulent inlet of the three-dimensional along-wind, across-wind and vertical pulsating wind fields. The three-dimensional pulsating wind speed based on the Karman spectrum in the atmospheric boundary layer satisfies the following expression: where: f is the frequency; S u (f), S v (f) and S w (f) are the power spectral density functions of the fluctuating wind speeds in the along-wind, cross-wind, and vertical directions, respectively; U ɑv is the mean wind speed; L u , L v , L w are the turbulence integral scales in the three directions, respectively.

10. A method for obtaining the wind pressure distribution of a near-ground wind-sensitive photovoltaic structure by using detached eddy simulation according to claim 1 or 8, characterized in that: In step S4, based on the principle of the NSRFG method, basic wind field parameters including the mean wind speed, the turbulence intensity and the turbulence integral scale need to be input in advance to generate the turbulent inlet. The mean wind speed profile satisfies the following formula: where: v z is the average wind speed at a height z above the ground; v 10 is the average wind speed at a height of 10 m above the ground; the roughness index α of B-type terrain is 0.15; the basic wind pressure with a return period of 50 years at the local area is w, and v 2 is calculated according to w = v 10 / 1600; According to the building load code and the code, the turbulence intensity and the turbulence integral scale in the along-wind, across-wind and vertical directions satisfy the following expressions: Where: I 10 is the nominal turbulence intensity at a height of 10 m, which is related to the surface roughness and takes a value of 0.14 for B-type terrain; ɑ takes the value of 0.

15. σ u 、σ v and σ w are respectively the root-mean-square of the wind speeds in the along-wind, cross-wind, and vertical directions; in the absence of measured data, I v =0.88I u ,I w =0.55I u ; where: z0 is the ground roughness parameter, which is taken as 0.05 for B-type terrain; σ v / σ u and σ w / σ u are taken according to the specification ESDU85020.

Images