A CFD numerical simulation method for wind-induced snow drift combined with actual measurements

Abstract

A CFD numerical simulation method for wind-induced snow drift combined with actual measurements, which relates to a CFD numerical simulation method for wind-induced snow drift. It aims to solve the technical problems that the existing numerical simulation methods do not consider the relative movement between wind and snow and have poor accuracy. This method: Based on the initial snow distribution morphology modeling measured on site and the measured snow particle data, the relative sliding speed between the wind phase and the snow phase is considered, and the UDF is written using the programming software Visual Studio and imported into the finite element numerical simulation analysis software Ansys Fluent to simulate the uneven distribution of snow on the structure, and the simulation data is post-processed using the software Tecplot, thereby establishing a CFD numerical simulation method for wind-induced snow drift that considers the relative movement of wind and snow combined with actual measurements. The present invention improves the accuracy of the numerical simulation method and can be used in the field of building structure design in cold areas.

Description

Technical Field

[0001] The invention belongs to the field of civil engineering and relates to a CFD numerical simulation method for wind-induced snow drift. Background Art

[0002] In recent years, heavy snowstorms in cold regions have brought great inconvenience to people's travel, communication, production and life, and have also greatly threatened the safety of houses and public buildings. Wind-induced snow drift can easily lead to excessive loads in local parts of buildings, and this uneven snow load on buildings often has additional effects on the structure, making it more likely to become unstable. Therefore, the study and prevention of wind and snow have become increasingly important.

[0003] In the early research on wind and snow disasters, due to the limitations of computers, scholars at home and abroad mainly conducted field measurements and experimental research. With the continuous progress of society and the rapid development of science and technology, numerical simulation research based on computational fluid dynamics has gradually been adopted by the public. The current numerical simulation research on wind-induced snow drift adopts the assumption of one-way coupling, and the parameters are set using empirical values. One-way coupling means that the influence of snow on the air phase is not considered, and the snow phase is regarded as a solid medium for passive transportation. The deposition and erosion of the snow phase will actually affect the boundary conditions of snow accumulation, thereby further affecting the flow field distribution. This method essentially calculates the diffusion of the snow phase under a stable wind field, which is very different from the actual situation and reduces the accuracy of the simulation. Summary of the invention

[0004] The present invention aims to solve the technical problem that the existing numerical simulation method does not consider the relative movement between wind and snow and has poor accuracy, and provides a CFD numerical simulation method for wind-induced snow drift combined with actual measurements. The method considers the relative movement of wind and snow, optimizes a series of processes from modeling of wind and snow two-phase flow, parameter setting, simulation calculation, etc., and obtains the CFD simulation value of wind-induced snow drift.

[0005] The CFD numerical simulation method of wind-induced snow drift combined with actual measurement of the present invention is carried out according to the following steps:

[0006] Step 1: Collect basic information needed for numerical simulation: collect the geometric characteristics of the building under study, the geographical location of the building, geomorphic parameters, average wind speed V0 at the reference height H0, wind profile specifications, turbulent kinetic energy specifications, turbulent intensity specifications, integrated turbulence scale specifications, turbulent energy dissipation rate specifications; at the same time, measure the snow particle density and snow particle diameter d of the snow at the location of the building s , angle of repose, wind speed, wind direction and initial snow depth of buildings, calculate threshold friction velocity, leaping snow particle velocity, ground roughness, building wall roughness, snow surface roughness and snow particle settling velocity w f ;

[0007] Step 2: Establish the grid model required for numerical simulation: Use the building geometry information and initial snow distribution pattern collected in step 1 to establish the model and divide the grid to obtain the grid file;

[0008] Step 3: Write the boundary conditions of the air term for numerical simulation: Use the geographical location information of the building location, the average wind speed at the reference height, the wind profile, the turbulent kinetic energy, the turbulent intensity, the integrated turbulence scale, and the turbulent dissipation rate specification information collected in step 1 to write the boundary conditions of the air phase. Then, in Visual Studio, based on the C language and the Fluent secondary development interface, write the exponential wind profile, turbulent kinetic energy, and turbulent dissipation rate program, save the C language file, and put it in the same folder as the grid model file in step 2;

[0009] Step 4: Write snow phase boundary conditions for numerical simulation: Using the snow particle density, gravitational acceleration at the building location, and threshold friction velocity collected in step 1, using the snow concentration empirical formula and the snow phase motion velocity, snow particle friction velocity, and critical height information of saltation and suspension layer used in the formula, write the snow phase volume fraction formula in Visual Studio based on C language and Fluent secondary development interface as the inlet boundary condition of the snow term, save the C language file, and put it in the same folder as the model grid file in step 2 and the C language file in step 3;

[0010] Step 5: Access the program required for numerical simulation. The specific operations are as follows: open the Ansys-Fluent interface and select 2D / 3D, double precision, and parallel. Under General Options-Working Directory, select the folder directory containing the grid file of step 2, the C language file of step 3, and the C language file of step 4. Select the Set up Compilation Environment for UDF interface in the environment to open the software; select User Defined-Functions-Compiled under the Ansys-Fluent main menu, select the C language files of steps 3 and 4 to load and compile, and obtain the inlet velocity and volume fraction of the air phase and snow phase;

[0011] Step 6: Using Ansys-Fluent software, first set the model, material, and solver parameters, then import the grid model obtained in step 2, the inlet velocities and volume fractions of the air phase and snow phase obtained in step 5, perform wind and snow two-phase flow simulation, and obtain the convergence files of wall shear force, air phase and snow phase velocity and volume fraction;

[0012] Step 7: Use the convergence file calculated in step 6 to post-process the wall shear force and velocity: Save the convergence file calculated by Ansys-Fluent in a format supported by Tecplot, use Tecplot for post-processing calculations, extract the components of velocity, pressure, and wall shear force in the x, y, and z directions in the three-dimensional space of the calculated domain, and use the Tecplot custom function Specify Equations to make a cloud map and contour map of pressure, velocity, and wall shear force within the calculation domain;

[0013] Step 8: Post-processing calculation of numerical simulation of snow deposition erosion: Based on the wall shear force calculated in step 7, the wall shear force on the bottom surface of the model or on the horizontal and vertical midlines of the bottom surface is extracted through Tecplot, and the formula The wall shear force is converted into the wall friction velocity, where τ is the wall shear force and ρ is the air density; the wall friction velocity is compared with the threshold friction velocity. When the wall friction velocity is greater than the threshold friction velocity, the snow is eroded; when the wall friction velocity is less than the threshold friction velocity, the snow is deposited; the erosion amount and deposition amount are calculated respectively, and the changed thickness of the snow is calculated. Finally, the changed thickness of the snow is superimposed with the initial snow depth of the building to obtain the final distribution of the snow, and complete the CFD numerical simulation of wind-induced snow drift combined with actual measurements.

[0014] Furthermore, the solving process of the mesh file described in step 2 is: in Ansys-Icem, the model calculation domain is determined according to the geometric characteristics of the building and the calculation domain is set to meet the blocking rate of less than 3%, and then a mesh model is established in Icem according to the model. In order to ensure smooth convergence of the calculation and accuracy of the simulation, a reasonable mesh size is set according to the model size, the mesh skewness (Skewness) of triangles and tetrahedrons is adjusted to be no more than 0.95, the mesh aspect ratio (Aspect Ratio) is adjusted, and the mesh squashing degree (Squish) is adjusted between 0-1, and finally the solved mesh file is output in Ansys-Fluent format.

[0015] Furthermore, a reasonable grid size means that the grid model controls the structured grid quality between 0.6 and 1, and the unstructured grid quality between 0.3 and 1; the grid aspect ratio is less than 5:1 in the flow core area and less than 10:1 in the boundary layer.

[0016] Furthermore, the empirical formula for snow concentration described in step 4 is the empirical formula for snow concentration proposed by the 1990 version of Pomeroy & Gray and the 1992 version of Pomeroy & Male.

[0017] Furthermore, the wind and snow two-phase flow simulation described in step 6 has the following specific process:

[0018] 1) Use Ansys-Fluent software and select Transient in the Time column under General;

[0019] 2) Select Mixture model in the Multiphase column under Models, select Realizable k-epsilon model under the k-epsilon two-equation model in the Viscous column under Models, select Phase Interaction-slip in the Multiphase column under Models, and set the sliding velocity algebraic equation;

[0020] 3) Select Fluid under Materials, add air phase, set Viscosity and Density, and create a new material snow phase. Set the snow phase Viscosity to be consistent with the air phase and set Density and particle size according to step 1.

[0021] 4) Under Cell Zone Conditions, set the mesh calculation domain to the required fluid and solid, and set the boundary conditions under Boundary Conditions: the inlet boundary is the velocity inlet, the two sides and the top of the calculation domain are symmetric boundaries, the ground and the model surface are no-slip walls, and the outlet boundary is a free outlet;

[0022] 5) Connect the air phase boundary conditions of step 3 to Boundary Conditions-inlet (inlet): set the wind profile, turbulent kinetic energy and dissipation rate, and connect the volume fraction of the snow phase in step 4 to Volume Fraction under Boundary Conditions-inlet (inlet)-snow phase-Multiphase; the Fluent-Mixture model is calculated using the finite volume method. Since the wind-snow flow has only two phases and the sum of the integral fractions of the two phases is 1, only the volume fraction of the snow phase needs to be given, and the air phase is calculated by the software without setting;

[0023] 6) In the Methods section of the Solver Settings, select the SIMPLE algorithm.

[0024] 7) In the Monitors-Residual selection under Solution, set the continuity equation, momentum equation, and k-epsilon convergence accuracy, where the k-epsilon convergence accuracy is set to 1e -6 Accuracy;

[0025] 8) In the Initialization under Solution setting, select Standard Initialization and initialize at the entrance;

[0026] 9) Finally, in Run Calculation, set the calculation time step and number of calculation steps, start the calculation, and obtain the convergence file of wall shear force, air phase and snow phase velocity and volume fraction.

[0027] Furthermore, in step eight, the calculation formula for the erosion amount is: Among them A ero is a constant coefficient, u * is the inlet friction velocity, u *t is the threshold speed.

[0028] Furthermore, in step eight, the calculation formula for the deposition amount is: where w f where is the snowfall velocity, φ s is the mass concentration of snow, φ s =ρ s f, f is the volume fraction of snow, ρ s is the density of snow particles.

[0029] Furthermore, in step eight, the calculation formula for the change in snow thickness is: where q s Q er o or q dep ,γ is the maximum volume fraction of snow, ρ s is the density of snow particles.

[0030] The method of the present invention uses Icem CFD software to establish a model, applies the measured cube and surrounding snow distribution to the establishment of the model, considers the initial snow distribution of wind-induced snow drift, and uses Ansys Fluent for numerical simulation analysis. The method is based on the two-phase flow theory of the Euler-Euler model, combines the numerical calculation development and analysis program of the turbulence model improved by Visual Studio, applies the measured wind speed, wind direction, physical properties of snow particles (shape, density, size), threshold speed of wind and snow drift, falling speed of snow particles, movement mode of wind and snow drift, angle of repose and other data to the simulation, simulates the snow distribution of the building, and completes the CFD numerical simulation of wind-induced snow drift combined with actual measurement. The simulation results are compared with the classical experiment of wind-induced snow drift, and compared with the method not combined with actual measurement, and the accuracy of the simulation method is verified. The present invention considers the relative movement between wind and snow and combines it with actual measurement, thereby improving the accuracy of numerical simulation of wind-induced snow drift.

[0031] The overall process of the CFD numerical simulation method of wind-induced snow drift combined with actual measurement is shown as follows: Figure 1 As shown, the numerical simulation method of wind-induced snow drift combined with actual measurements of the present invention has the following advantages:

[0032] 1. The relative slip velocity between the wind phase and the snow phase is considered, and the interaction between the two phases is considered by solving the relative velocity between the two phases.

[0033] 2. The measured snow distribution pattern of windblown snow is used in the establishment of the model, and the influence of snow deposition and erosion on the flow field distribution in the calculation domain is considered.

[0034] 3. Experimental tools were used to measure various data based on local conditions and applied them to numerical simulations, making the simulation results more consistent with reality and more accurate.

[0035] The method of the invention is helpful to understand the law of wind and snow movement, to prevent building collapse, and to reduce casualties and economic losses. The method can be used in the field of building structure design in cold regions. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the overall process of the CFD numerical simulation method of wind-induced snow drift combined with actual measurement of the present invention;

[0037] Figure 2 is a diagram of the grid division and computational domain of Example 1;

[0038] Figure 3 The snow distribution diagram obtained in step 7 of Example 1;

[0039] Figure 4 It is a comparison chart of the snow flux per unit time obtained in Example 1, the snow flux per unit time obtained in the classical experiment, and the snow flux per unit time obtained in Example 2, i.e., the method not combined with actual measurement. DETAILED DESCRIPTION

[0040] Use the following examples to verify the beneficial effects of the present invention:

[0041] Embodiment 1: This embodiment is a numerical simulation method for wind-induced snow drift combined with actual measurements, and is carried out in the following steps:

[0042] Step 1: Collect basic information needed for numerical simulation: Collect the geometric characteristics of the specific building under study: a cube model with a side length of 1m, the geographical location of the building with a geomorphic parameter of 0.16, an average wind speed of 5m / s at a reference height of 10m, and a wind profile using the formula z0 and v0 represent the height of the reference point (1m) and the corresponding reference average wind speed (5m / s), respectively. z represents any height. α represents the landform parameter. For Class B landform, α is 0.16. The turbulent kinetic energy is calculated using the formula The turbulent energy dissipation rate is given by the formula Among them, C μ is an empirical constant, which is taken as 0.09; I and l are the turbulence intensity and turbulence length scale respectively, where:

[0043] The turbulent kinetic energy intensity is given by the formula The integrated turbulence scale is given by the formula Among them, z G 、z b , α', are all geomorphic parameters. In the Japanese AIJ standard, Class B topography corresponds to Class II flat land. The values ​​of the above three geomorphic parameters are 5m, 350m and 0.15 respectively.

[0044] The snow particle diameter, repose angle and snow density were measured by using a sample sieve, a conical funnel, a Snow Fork and a HOBO small weather station. The results showed that the snow particle diameter was 0.00015m, the repose angle was 45° and the snow density was 145kg / m 3 , the calculated threshold friction velocity is 0.2 m / s, the leap snow particle velocity is 0.2 m / s, the ground roughness coefficient is 0.25, and the snow particle settling velocity is 0.2 m / s.

[0045] Step 2: Establish the mesh model required for numerical simulation: In Ansys-Icem CFD, establish the model and divide the mesh according to the building geometry information and initial snow distribution pattern collected in step 1, determine the model calculation domain, the length, width and height are 20m, 10m and 8m respectively, and the calculation domain is set to have a blocking rate of 1.25%, then establish the mesh model in Icem CFD based on the model, and the number of meshes is 3.6 million based on the model size. The mesh model controls the structured mesh quality at 0.8 and the unstructured mesh quality at 0.8, and finally outputs the solved mesh file in Ansys-Fluent format. The mesh division and calculation domain are shown in the figure below. Figure 2 As shown; Figure 2 In the figure, 1 is the symmetric boundary, 2 is the free development boundary, 3 is the velocity inlet, and 4 is the no-slip wall.

[0046] Step 3: Write the boundary conditions of the air term for numerical simulation: Use the geographical location information of the building, the average wind speed at the reference height, the wind profile, the turbulent kinetic energy, the turbulent intensity, the integrated turbulence scale, and the turbulent dissipation rate specification information collected in step 1 to write the boundary conditions of the air phase. In Visual Studio, based on the C language and the Fluent secondary development interface, write the exponential wind profile, turbulent kinetic energy, and turbulent dissipation rate, save the C language file, and put it in the same folder as the grid model file in step 2;

[0047] Step 4: Write the snow phase boundary conditions for numerical simulation: Use the snow particle density of 145 kg / m collected in step 1 3 、Local gravity acceleration 9.81m / s 2 , the threshold friction velocity is 0.2 m / s, and the snow concentration empirical formula proposed by Pomeroy & Gray (1990) and Pomeroy & Male (1992) is adopted, that is, the formula where ρ s represents the density of snow particles, g is the acceleration due to gravity, which is 9.81 m / s 2 ,u p is the average velocity of snow particles in the saltation layer, u p =2.8u *t ,u * is the inlet friction velocity; the friction velocity obtained by fitting based on the actual measurement by Beyers et al. and the wind speed u at a height of 10 m 10 The relationship is determined, The snow phase motion speed is 5m / s, the snow particle friction speed is 0.15m / s, and u *t is the threshold speed; take 0.15m / s; h s is the critical height of the transition layer and the suspension layer, The snow phase movement speed is 5m / s, the snow particle friction speed is 0.15m / s, and the snow phase volume fraction formula is written as the inlet boundary condition of the snow term based on C language and Fluent secondary development interface in Visual Studio. Save the C language file and put it in the same folder as the model grid file in step 2 and the C language file in step 3;

[0048] Step 5: Access the programs required for numerical simulation. The specific operations are as follows: Open the Ansys-Fluent interface and select 3D, double precision, and parallel. Under General Options-Working Directory, select the folder directory containing the mesh file in step 2 and the custom programs in steps 3 and 4, and select the environment to Set up Compilation Environmentfor UDF interface to open the software. Select User-Defined-Functions-Compiled under the Ansys-Fluent main menu and select the C language files in steps 3 and 4 to load and compile, and obtain the inlet velocity and volume fraction of the air phase and snow phase;

[0049] Step 6: Use Ansys-Fluent software to set the model, material, and solver parameters first, then import the grid model obtained in step 2, the inlet velocity and volume fraction of the air phase and snow phase obtained in step 5, and simulate the wind and snow two-phase flow to obtain the convergence files of wall shear force, air phase and snow phase velocity and volume fraction, as follows:

[0050] 1) Under General, in the Time column, select Transient;

[0051] 2) Select Mixture model in the Multiphase column under Models, select Realizable k-epsilon model under the k-epsilon two-equation model in the Viscous column under Models, select Phase Interaction-slip in the Multiphase column under Models, and set the relative sliding speed to 0.25 m / s;

[0052] 3) Select Fluid under Materials, add an air phase, set the Viscosity and Density, and create a new material snow phase. Set the snow phase Viscosity to be the same as air. Set the Density to 145kg / m according to step 1. 3 (measured by Snow Fork snow characteristics analyzer), particle size 0.00015m (measured by sample sieve);

[0053] 4) Set the required fluid and solid in the grid calculation domain under the volume grid Cell Zone Conditions, and set the boundary conditions in the Boundary Conditions: the inlet boundary is the velocity inlet, the two sides and the top of the calculation domain are symmetric boundaries, the ground and the model surface are no-slip walls, and the outlet boundary is a free outlet;

[0054] 5) Connect the air phase boundary conditions of step 3 to Boundary Conditions-inlet: set the wind profile, turbulent kinetic energy and dissipation rate. Connect the volume fraction of the snow phase to Volume Fraction of step 4 under Boundary Conditions-inlet-snow-multiphase. The Fluent-Mixture model is calculated using the finite volume method. Since the wind-snow flow has only two phases and the sum of the integral fractions of the two phases is 1, only the volume fraction of the snow phase needs to be given. The air phase is calculated by the software and does not need to be set.

[0055] 6) Select the SIMPLE algorithm in Methods under Solution.

[0056] 7) Under Solution, select Monitors-Residual and set the convergence accuracy of the continuity equation, momentum equation, and k-epsilon. In this example, the convergence accuracy is set to 1e -6 Accuracy.

[0057] 8) In the solver settings Solution, select Standard Initialization in Initialization to initialize at the entrance.

[0058] 9) In the final calculation, set the number of steps to 1000 and the precision to 10. -6 , start the calculation and obtain the convergence file of wall shear force, velocity and volume fraction of air phase and snow phase.

[0059] Step 7: Use the convergence file calculated in step 6 to post-process the wall shear force and velocity: Save the convergence file calculated by Ansys-Fluent in a format supported by Tecplot, use Tecplot for post-processing calculations, extract the components of velocity, pressure, and wall shear force in the x, y, and z directions in the three-dimensional space of the calculated domain, and use the Tecplot custom function Specify Equations to make a cloud map and contour map of pressure, velocity, and wall shear force within the calculation domain; the snow distribution diagram is shown in the figure below: Figure 3 shown.

[0060] Step 8: Post-processing calculation of numerical simulation of snow deposition erosion: Based on the wall shear force calculated in step 7: Based on the wall shear force calculated in step 7, the wall shear force of the bottom surface of the model is extracted through Tecplot, and the wall shear stress is converted into the wall friction velocity through a custom function. The wall friction velocity is compared with the threshold friction velocity. When the wall friction velocity is greater than the threshold friction velocity, it is erosion, and the formula is used. Calculate the erosion amount, where A ero is a constant coefficient, taking 7.0×10 -4 ; When the wall friction velocity is less than the threshold friction velocity, the deposition is carried out using the formula Calculate the deposition amount, where w f is the snow settling velocity, φ s is the mass concentration of snow, φ s =ρ s f, f is the volume fraction of snow, ρ s is the density of snow particles; then use the formula Calculate the change in snow thickness, where q s for q ero or q dep , γ is the maximum volume fraction of snow, which is taken as 0.62; finally, the changed snow thickness is superimposed on the initial snow depth of the building in step one to obtain the final distribution of snow, and the CFD numerical simulation of wind-induced snow drift combined with actual measurements is completed.

[0061] Embodiment 2: This embodiment simulates the numerical value of wind-induced snow drift without combining actual measurement and without considering the relative slip velocity of wind phase and snow phase. The specific method is carried out according to the following steps:

[0062] Step 1: Collect basic information needed for numerical simulation: The geometric characteristics of the specific building under study: a cube model with a side length of 1m, the geographical location of the building with a geomorphic parameter of 0.16, an average wind speed of 5m / s at a reference height of 10m, and a wind profile using the formula z and v0 represent the height of the reference point (1m) and the corresponding reference average wind speed (5m / s), respectively; α represents the landform parameter. For Class B landform, α is 0.16; the turbulent kinetic energy is calculated using the formula The turbulent energy dissipation rate is given by the formula Among them, C μ is an empirical constant, taken as 0.09; I and l are the turbulence intensity and turbulence length scale respectively, and the turbulence kinetic energy intensity is calculated using the formula The integrated turbulence scale is given by the formula Among them, z G 、z b, α', are all geomorphic parameters. In the Japanese AIJ standard, Class B topography corresponds to Class II flat land. The values ​​of the above three geomorphic parameters are 5m, 350m and 0.15 respectively.

[0063] The relevant values ​​are adopted through experience: snow particle diameter 0.00015m, no consideration of repose angle, snow density 150kg / m 3 , the calculated threshold friction velocity is 0.2 m / s, the leap snow particle velocity is 0.2 m / s, the ground roughness coefficient is 0.25, and the snow particle settling velocity is 0.2 m / s.

[0064] Step 2: Establish the grid model required for numerical simulation: In Ansys-Icem CFD, the model is established and the grid is divided according to the building geometry information and initial snow distribution pattern collected in step 1. The model calculation domain is determined, and the length, width, and height are 20m, 10m, and 8m respectively. The calculation domain is set to have a blocking rate of 1.25%. Then, a grid model is established in Icem CFD based on the model, and the number of grids is 3.6 million based on the model size. The grid model controls the structured grid quality at 0.8 and the unstructured grid quality at 0.8. Finally, the solved grid file is output in Ansys-Fluent format.

[0065] Step 3: Write the boundary conditions of the air term for numerical simulation: Use the geographical location information of the building, the average wind speed at the reference height, the wind profile, the turbulent kinetic energy, the turbulent intensity, the integrated turbulence scale, and the turbulent dissipation rate specification information collected in step 1 to write the boundary conditions of the air phase. Based on the existing technology, in Visual Studio, using C language and Fluent secondary development interface, write the exponential wind profile, turbulent kinetic energy, and turbulent dissipation rate, save the C language file, and put it in the same folder as the grid model file in step 2;

[0066] Step 4: Write the snow phase boundary conditions for numerical simulation: Use the snow particle density of 150kg / m collected in step 1 3 、Local gravity acceleration 9.81m / s 2 , the threshold friction velocity is 0.2 m / s, and the snow concentration empirical formula proposed by Pomeroy & Gray (1990) and Pomeroy & Male (1992) is adopted, that is, the formula where ρ s represents the density of snow particles, g is the acceleration due to gravity, which is 9.81 m / s 2 ,u p is the average velocity of snow particles in the saltation layer, u p =2.8u *t ,u * is the inlet friction velocity, and the friction velocity obtained by fitting based on the actual measurement by Beyers et al. is the wind speed u at a height of 10 m.10 The relationship is determined, The snow phase motion speed is 5m / s, the snow particle friction speed is 0.15m / s, and u *t is the threshold speed; take 0.15m / s; h s is the critical height of the transition layer and the suspension layer, The snow phase movement speed is 5m / s, the snow particle friction speed is 0.15m / s, and the snow phase volume fraction formula is written as the inlet boundary condition of the snow term in Visual Studio based on C language and Fluent secondary development interface. Save the C language file and put it in the same folder as the model grid file in step 2 and the C language file in step 3;

[0067] Step 5: Access the program required for numerical simulation. The specific operation is as follows: Store the grid file of step 2 and the custom program of steps 3 and 4 in the same folder, open the Ansys-Fluent interface and select 3D, double precision, and parallel. Select the above folder directory under General Options-Working Directory, select the environment Set up Compilation Environment for UDF interface to open the software. Select User-Defined-Functions under the Ansys-Fluent main menu

[0068] - Under Compiled, select the C language files in steps 3 and 4 to load and compile, and obtain the inlet velocity and volume fraction of the air phase and snow phase;

[0069] Step 6: Set the model, material, and solver parameters of the numerical simulation software, use Ansys-Fluent to import the grid model, and perform wind and snow two-phase flow simulation based on the secondary development programming imported in step 5, as follows:

[0070] 1) Under General, in the Time column, select Transient;

[0071] 2) Select Mixture model in the Multiphase column under Models, select Realizable k-epsilon model under the k-epsilon two-equation model in the Viscous column under Models, and select Phase Interaction-slip in the Multiphase column under Models. At this time, the relative slip velocity between the wind phase and the snow phase is not considered;

[0072] 3) Select Fluid under Materials, add an air phase, set the Viscosity and Density, and create a new material snow phase. Set the snow phase Viscosity to be the same as air. Set the Density to 150kg / m according to step 1. 3 , particle size 0.00015m;

[0073] 4) Set the required fluid and solid in the grid calculation domain under the volume grid Cell Zone Conditions, and set the boundary conditions in the Boundary Conditions: the inlet boundary is the velocity inlet, the two sides and the top of the calculation domain are symmetric boundaries, the ground and the model surface are no-slip walls, and the outlet boundary is a free outlet;

[0074] 5) Connect the air phase boundary conditions of step 3 to Boundary Conditions-inlet (inlet): set the wind profile, turbulent kinetic energy and dissipation rate, and connect the volume fraction of the snow phase in step 4 to Volume Fraction under Boundary Conditions-inlet (inlet)-snow phase-Multiphase; the Fluent-Mixture model is calculated using the finite volume method. Since the wind-snow flow has only two phases and the sum of the integral fractions of the two phases is 1, only the volume fraction of the snow phase needs to be given, and the air phase is calculated by the software without setting;

[0075] 6) Select the SIMPLE algorithm in Methods under Solution.

[0076] 7) Under Solution, select Monitors-Residual and set the convergence accuracy of the continuity equation, momentum equation, and k-epsilon. In this example, the convergence accuracy is set to 1e -6 Accuracy.

[0077] 8) In the solver settings Solution, select Standard Initialization in Initialization to initialize at the entrance.

[0078] 9) Finally, in Run Calculation, set the number of calculation steps to 1000 and the precision to 10 -6 , start the calculation and obtain the convergence file of wall shear force, velocity and volume fraction of air phase and snow phase.

[0079] Step 7: Use the convergence file calculated in step 6 to post-process the wall shear force and velocity: After the Ansys-Fluent calculation converges, the file is saved in a Tecplot-supported format. Use Tecplot post-processing calculation to extract the calculated x, y, z components of velocity, pressure, and wall shear force, and use the Tecplot custom function Specify Equations to make a cloud map and contour map of pressure, velocity, and wall shear force within the calculation domain.

[0080] Step 8: Post-processing calculation of numerical simulation of snow deposition erosion: Based on the wall shear force calculated in step 7, the wall shear force on the bottom surface of the model is extracted through Tecplot. The wall shear stress is converted into the wall friction velocity through a custom function. The wall friction velocity is compared with the threshold friction velocity. When the wall friction velocity is greater than the threshold friction velocity, erosion occurs. The formula is used. Among them A ero is a constant coefficient, taking 7.0×10 -4 ; When the wall friction velocity is less than the threshold friction velocity, the deposition is carried out using the formula where w f where is the snowfall velocity, φ s is the mass concentration of snow (φ s =ρ s f, f is the volume fraction of snow, ρ s is the snow particle density) to calculate the erosion and deposition amount; based on the initial snow depth in step 1, the formula is used Calculate the thickness of snow change and the amount of superimposed snow change (where q is used uniformly) s Instead of q ero and q dep , γ is the maximum volume fraction of snow, which is taken as 0.62), the final distribution of snow is obtained, and the numerical simulation of wind-induced snow drift is completed.

[0081] The snow flux per unit time shown in the simulation results of Example 1, the snow flux per unit time obtained in Example 2 without combining with actual measurement and without considering the relative sliding speed of the wind phase and the snow phase, and the snow flux per unit time curve of the classic experiment are compared, as shown in FIG. Figure 4 As shown, from Figure 4 It can be seen that the snow flux per unit time obtained by Example 2 without combining with actual measurements and without considering the relative slip speed of the wind phase and the snow phase is quite different from the curve obtained by the classical experimental method, while the simulation results of Example 1 are in good agreement with the results of the classical experiment, indicating that the simulation method of Example 1 is highly accurate.

Claims

1. A CFD numerical simulation method for wind-induced snow drift combined with actual measurements, characterized in that The method proceeds as follows: Step 1: Collect basic information needed for numerical simulation: collect the geometric characteristics of the building under study, the geographical location of the building, geomorphic parameters, average wind speed V0 at the reference height H0, wind profile specifications, turbulent kinetic energy specifications, turbulent intensity specifications, integrated turbulence scale specifications, turbulent energy dissipation rate specifications; at the same time, measure the snow particle density and snow particle diameter d of the snow at the location of the building s , angle of repose, wind speed, wind direction and initial snow depth of buildings, calculate threshold friction velocity, leaping snow particle velocity, ground roughness, building wall roughness, snow surface roughness and snow particle settling velocity w f ; Step 2: Establish the grid model required for numerical simulation: Use the building geometry information and initial snow distribution pattern collected in step 1 to establish the model and divide the grid to obtain the grid file; Step 3: Write the boundary conditions of the air term for numerical simulation: Use the geographical location information of the building location, the average wind speed at the reference height, the wind profile, the turbulent kinetic energy, the turbulent intensity, the integrated turbulence scale, and the turbulent dissipation rate specification information collected in step 1 to write the boundary conditions of the air phase. Then, in Visual Studio, based on the C language and the Fluent secondary development interface, write the exponential wind profile, turbulent kinetic energy, and turbulent dissipation rate program, save the C language file, and put it in the same folder as the grid model file in step 2; Step 4: Write snow phase boundary conditions for numerical simulation: Using the snow particle density, gravitational acceleration at the building location, and threshold friction velocity collected in step 1, using the snow concentration empirical formula and the snow phase motion velocity, snow particle friction velocity, and critical height information of saltation and suspension layer used in the formula, write the snow phase volume fraction formula in Visual Studio based on C language and Fluent secondary development interface as the inlet boundary condition of the snow term, save the C language file, and put it in the same folder as the model grid file in step 2 and the C language file in step 3; Step 5: Access the program required for numerical simulation. The specific operations are as follows: open the Ansys-Fluent interface and select 2D / 3D, double precision, and parallel. Under General Options-Working Directory, select the folder directory containing the grid file of step 2, the C language file of step 3, and the C language file of step 4. Select the Set up Compilation Environment for UDF interface in the environment to open the software; select User Defined-Functions-Compiled under the Ansys-Fluent main menu, select the C language files of steps 3 and 4 to load and compile, and obtain the inlet velocity and volume fraction of the air phase and snow phase; Step 6: Using Ansys-Fluent software, first set the model, material, and solver parameters, then import the grid model obtained in step 2, the inlet velocities and volume fractions of the air phase and snow phase obtained in step 5, perform wind and snow two-phase flow simulation, and obtain the convergence files of wall shear force, air phase and snow phase velocity and volume fraction; Step 7: Use the convergence file calculated in step 6 to post-process the wall shear force and velocity: Save the convergence file calculated by Ansys-Fluent in a format supported by Tecplot, use Tecplot for post-processing calculations, extract the components of velocity, pressure, and wall shear force in the x, y, and z directions in the three-dimensional space of the calculated domain, and use the Tecplot custom function Specify Equations to make a cloud map and contour map of pressure, velocity, and wall shear force within the calculation domain; Step 8: Post-processing calculation of numerical simulation of snow deposition erosion: Based on the wall shear force calculated in step 7, the wall shear force on the bottom surface of the model or on the horizontal and vertical midlines of the bottom surface is extracted through Tecplot, and the formula Convert the wall shear force into the wall friction velocity, where τ is the wall shear force and ρ is the air density; The wall friction velocity is compared with the threshold friction velocity. When the wall friction velocity is greater than the threshold friction velocity, the snow is eroded; when the wall friction velocity is less than the threshold friction velocity, the snow is deposited. The erosion amount and deposition amount are calculated respectively, and the thickness of the snow is calculated. Finally, the snow change amount is superimposed with the initial snow depth of the building to obtain the final distribution of the snow, and the CFD numerical simulation of wind-induced snow drift combined with actual measurements is completed. The calculation formula of erosion amount is: Among them A ero is a constant coefficient, u * is the inlet friction velocity, u *t is the threshold speed; The calculation formula for the deposition amount is: where w f where is the snowfall velocity, φ s is the mass concentration of snow, φ s =ρ s f, f is the volume fraction of snow, ρ s is the density of snow particles; The calculation formula for the change in snow thickness is: where q s for q ero or q dep ,γ is the maximum volume fraction of snow, ρ s is the density of snow particles.

2. A CFD numerical simulation method for wind-induced snow drift combined with actual measurements according to claim 1, characterized in that The steps for solving the grid file described in step 2 are: determine the model calculation domain according to the geometric characteristics of the building in Ansys-Icem and set the calculation domain to meet the blocking rate of less than 3%, then establish a grid model based on the model in Icem, and then set a reasonable grid size based on the model size, adjust the grid skewness to no more than 0.95 between triangles and tetrahedrons, adjust the grid aspect ratio, adjust the grid flattening degree between 0-1, and finally output the solved grid file in Ansys-Fluent format.

3. The CFD numerical simulation method for wind-induced snow drift combined with actual measurements according to claim 2 is characterized in that, in order to ensure smooth convergence of calculation and accuracy of simulation, The reasonable grid size refers to that the grid model controls the structured grid quality to be between 0.6 and 1, and the unstructured grid quality to be between 0.3 and 1.

4. A CFD numerical simulation method for wind-induced snow drift combined with actual measurements according to claim 2, characterized in that The grid aspect ratio is: the grid aspect ratio in the flow core area is less than 5:1, and the grid aspect ratio in the boundary layer is less than 10:

1.

5. A CFD numerical simulation method for wind-induced snow drift combined with actual measurements according to claim 1, 2, 3 or 4, characterized in that The specific process of the wind and snow two-phase flow simulation described in step 6 is as follows: 1) Use Ansys-Fluent software and select Transient in the Time column under General; 2) Select Mixture model in the Multiphase column under Models, select Realizable k-epsilon model under the k-epsilon two-equation model in the Viscous column under Models, select Phase Interaction-slip in the Multiphase column under Models, and set the sliding velocity algebraic equation; 3) Select Fluid under Materials, add air phase, set Viscosity and Density, and create a new material snow phase. Set the snow phase Viscosity to be consistent with the air phase and set Density and particle size according to step 1. 4) Under Cell Zone Conditions, set the mesh calculation domain to the required fluid and solid, and set the boundary conditions under Boundary Conditions: the inlet boundary is the velocity inlet, the two sides and the top of the calculation domain are symmetric boundaries, the ground and the model surface are no-slip walls, and the outlet boundary is a free outlet; 5) Connect the air phase boundary conditions in step 3 to Boundary Conditions-inlet: set the wind profile, turbulent kinetic energy and dissipation rate. Connect the volume fraction of the snow phase in step 4 to Volume Fraction under Boundary Conditions-inlet-snow-multiphase. The Fluent-Mixture model is calculated using the finite volume method. Since the wind-snow flow has only two phases and the sum of the integral fractions of the two phases is 1, only the volume fraction of the snow phase needs to be given. The air phase is calculated by the software and does not need to be set. 6) In the Methods section of the Solver Settings, select the SIMPLE algorithm. 7) In the Monitors-Residual selection under Solution, set the continuity equation, momentum equation, and k-epsilon convergence accuracy, where the k-epsilon convergence accuracy is set to 1e -6 Accuracy; 8) In the Initialization under the Solution setting, select Standard Initialization and initialize at the entrance; 9) Finally, in Run Calculation, set the calculation time step and number of calculation steps, start the calculation, and obtain the convergence file of wall shear force, air phase and snow phase velocity and volume fraction.

Images