A numerical simulation method for aerodynamics of multi-rotor vertical wind turbines

Abstract

The present invention discloses a method for numerical simulation of aerodynamics of a multi-rotor vertical wind energy generator, step one: establish a geometric shape file of a calculation area for calculation; step two: divide the rotating area and the non-rotating area, and establish a geometric shape file; step three: generate a grid for the entire calculation area, step four: establish fluid control equations and boundary conditions; step five: when solving the control equation of step four, it is necessary to introduce the corresponding turbulence model equation; step six: interpolate after discretizing the original model using the finite volume method; step seven: decompose the solution area; step eight: solve the matrix on each core using the corresponding matrix solution method. The advantages of the present invention compared with the prior art are: ensuring the accuracy of the flow field calculation around the multi-rotor vertical axis wind energy generator, being able to clearly display the vortex distribution around the rotor, still having a high parallel efficiency when there are more than 1000 cores, and greatly shortening the calculation time.

Description

Technical Field

[0001] The invention relates to the technical field of vertical wind energy generators, and in particular to an aerodynamic numerical simulation method for a multi-rotor vertical wind energy generator. Background Art

[0002] In the actual operation of multi-rotor vertical wind turbines, due to the mutual influence between the rotors, the eddy currents around the rotors are more complex. Compared with the horizontal axis wind turbine rotors, the rotors rotate faster, and the numerical solution process requires higher discretization of the rotating area and higher solution accuracy. There are two main problems in the numerical calculation of multi-rotor vertical wind turbines: the accuracy of the eddy current calculation and the calculation processing of the rotating area.

[0003] In the CFD solution process, it is mainly divided into transient calculation and steady-state calculation according to the time term. Transient calculation can more clearly show the changes of the flow field during the calculation process, but its calculation amount is generally large. Steady-state calculation mainly obtains the results when the flow field is stable. Its calculation amount is small, but the change process of the entire flow field cannot be observed.

[0004] For the existing technology, there are three main processing technologies for the rotating part, namely, the fixed rotor technology (Frozen Rotor Technique), the dynamic mesh technology (dynamic mesh) and the mixed plane method (Mixing plane Apporach). The representative method of the fixed rotor technology and the mixed plane technology is the multiple reference system method (Moving reference MRF for short). In this method, the rotating body does not actually move during the calculation process. The calculation is achieved by setting multiple reference systems. This method is mainly used for steady-state calculations. The calculation cost of this method is low, but its calculation accuracy is also low. In the dynamic mesh technology, the mesh actually rotates during the calculation. It is mainly used for transient calculations. Compared with the other two methods, it has higher accuracy and can accurately show the flow field change trend during the rotation process, but its calculation time is about 30 times that of the steady-state calculation time.

[0005] At present, the research on multi-rotor vertical wind energy engines is mainly focused on improving the accuracy of calculations. However, with the increase of the calculation area, the amount of calculation grids has increased significantly, and the scale of the problems that need to be solved has also become larger and larger. When using dynamic mesh technology to ensure calculation accuracy, the calculation speed becomes a big threshold, resulting in numerical analysis results that are either insufficiently accurate or the calculation time is too long. Numerical calculations cannot be used well to assist the design of actual wind energy engines, and the calculation cost is too high. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the above technical defects and provide a method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine, which ensures the accuracy of flow field calculation around the multi-rotor vertical axis wind turbine and can clearly display the vortex distribution around the rotor. It has a high parallel efficiency with more than 1000 cores and greatly shortens the calculation time.

[0007] In order to solve the above technical problems, the present invention provides a technical solution: a method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine, comprising the following steps:

[0008] Step 1: Create a geometry file of the calculation area used for calculation;

[0009] Step 2: Segment the rotating area and the non-rotating area, and create a geometric shape file that wraps the entire rotating area;

[0010] Step 3: Generate mesh for the entire calculation area, which is divided into two parts: the rotation area and the non-rotation area;

[0011] Step 4: Establish fluid control equations and boundary conditions;

[0012] Step 5: When solving the control equation in step 4, the corresponding turbulence model equation needs to be introduced;

[0013] Step 6: After the original model is discretized using the finite volume method, the rotation area and the non-rotation area need to be interpolated on the grid of the geometry file in step 2. The interpolation method uses the arbitrary grid interface method.

[0014] Step 7: Decompose the solution region and discretize the corresponding discretized matrix to different numbers of cores;

[0015] Step 8: Use the corresponding matrix solution method for each matrix on the kernel to solve the discretized linear matrix.

[0016] Preferably, the geometry file in step 1 includes the geometry file of the vertical axis wind turbine, the geometry file of the entire region that the flow field of a single wind turbine may affect, and the geometry files of the inlet and outlet for wind inflow.

[0017] Preferably, in step 2, for the rotor rotating around the axis, the dividing interface should be cylindrical.

[0018] Preferably, in step three, the entire calculation area is divided into structured hexahedral grids, and the rotating area grids and the non-rotating area grids are separated by the cylindrical surface in step two and form corresponding surface grids.

[0019] Preferably, the control equation in step 4 is an incompressible Newtonian fluid, which should satisfy the Navier-Stokes equations for incompressible fluids.

[0020] Preferably, in step five, a large eddy simulation filter is used for filtering, a Sparat-Almaraz turbulence model is introduced for calculation, and an eddy delay separation method is used.

[0021] Preferably, in step six, a driven surface and an active surface are set on the surface that divides the rotating area and the non-rotating area, the active surface maintains the rotation speed, and the driven surface remains stationary, and interpolation is performed according to the area proportions of the driven surface and the active surface grids in the grid area respectively.

[0022] Preferably, in step seven, a decomposition method based on the amount of calculation and the amount of grid is used when solving the regional decomposition.

[0023] Preferably, in step eight, a multi-grid method is used to solve the flow field for each matrix.

[0024] The advantages of the present invention compared with the prior art are: 1. The present invention uses a cylindrical surface to divide the rotating area and the non-rotating area, and uses actual rotation calculated for the rotating area, thereby improving the calculation accuracy;

[0025] 2. Solve the NS equations based on large eddy simulation for calculation, introduce the SA turbulence model and use the delayed separation method to reduce the error of eddy distribution and obtain high-resolution calculation results;

[0026] 3. Distribute the matrices of different cores to the computational grid according to the computational load and grid quantity to obtain higher parallel efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is an overall flow chart of an aerodynamic numerical simulation method for a multi-rotor vertical wind energy generator of the present invention.

[0028] Figure 2 It is an example diagram of the shape file of the calculation area in step one of the specific measures of the aerodynamic numerical simulation method of a multi-rotor vertical wind turbine of the present invention.

[0029] Figure 3 This is an example diagram of the interface of dividing the rotating area and the non-rotating area of ​​a multi-rotor vertical wind power generator aerodynamic numerical simulation method of the present invention.

[0030] Figure 4 This is an example diagram of the computational grid distribution of a method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to the present invention.

[0031] Figure 5 This is an example diagram of the calculation results of a method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to the present invention.

[0032] Figure 6 It is a calculation case eddy flow distribution diagram in the numerical calculation of the aerodynamic numerical simulation method of a multi-rotor vertical wind energy generator of the present invention.

[0033] Figure 7 It is a streamline distribution diagram of a numerical calculation case of a multi-rotor vertical wind turbine aerodynamic numerical simulation method of the present invention.

[0034] Figure 8 It is a parallel efficiency diagram of a numerical calculation case of a multi-rotor vertical wind turbine aerodynamic numerical simulation method of the present invention. DETAILED DESCRIPTION

[0035] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0036] refer to Figure 1 :A method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine includes the following steps:

[0037] Step 1: Establish the geometry description file of the entire solution system: During the establishment process, it is necessary to establish the geometry file of the entire calculation area, mainly including: the geometry file of the vertical axis wind turbine, the geometry file of the entire area that the flow field of a single wind turbine may affect, and the geometry file of the inlet and outlet of the wind inflow, see Figure 2 .

[0038] Step 2: Segment the rotating area and the non-rotating area, and create a geometric shape file that wraps the entire rotating area: In this process, you need to create a geometric shape file for the entire rotating area. For a rotor that rotates around its axis, its segmentation interface should be cylindrical. Figure 3 .

[0039] Step 3: Divide the rotating area and the non-rotating area, and create a geometric shape file that wraps the entire rotating area: The entire calculation area is divided into a structured hexahedral grid. The rotating area grid and the non-rotating area grid are separated by the cylindrical surface in step 2, and the corresponding surface grids are formed. Figure 4 .

[0040] Step 4: Establish fluid control equations and boundary conditions: The entire calculation control equation is an incompressible Newtonian fluid, which should satisfy the incompressible fluid Navier-Stokes equations (hereinafter referred to as NS equations)

[0041] Navier-Stokes equations:

[0042]

[0043] Where: v refers to the velocity field, p refers to the fluid pressure, is the kinematic viscosity, μ is the dynamic viscosity, and ρ is the fluid density

[0044] The boundary conditions are as follows for the inlet:

[0045]

[0046] Where: U inlet is the inlet velocity, p is the pressure

[0047] For export use:

[0048]

[0049] Where: U outlet is the outlet velocity, p is the pressure

[0050] Step 5: Introduce the corresponding turbulence model equation for the control equation of step 4

[0051] In order to better solve the NS equation in step 4, a large eddy simulation (hereinafter referred to as LES) filter is used for filtering, and the filtered equation is:

[0052]

[0053] Where: is the velocity field after filtering, is the filtered pressure field, v0 is the kinematic viscosity, is the filtered strain rate tensor,

[0054] The Spalart-Allmaras (hereinafter referred to as SA) turbulence model is introduced into the above-mentioned filtered equation for calculation.

[0055]

[0056] Length scale In the turbulence model, the following conditions must be met:

[0057] Δ=max(Δx,Δy,Δz)

[0058] Where: d w Distance from finger to wall

[0059] In order to overcome the grid-induced separation problem in the SA model, an eddy current delayed separation method is used, in which the length scale is modified. The length scale is represented by a function,

[0060]

[0061] Where: f d is a shielding function, which takes a value of 0 at the boundary layer and 1 at the edge of the boundary layer. d It is possible to clearly distinguish whether the computational unit is inside the boundary layer, f d Defined as:

[0062]

[0063] Where κ = 0.41 is the Von Karman constant.

[0064] Step six: After discretizing the original model using the finite volume method, it is necessary to interpolate the rotating and non-rotating areas on the grid of the geometric shape file in step two. The interpolation method uses the arbitrary mesh interface method (AMI for short): set the driven surface and the active surface on the surface that divides the rotating and non-rotating areas. The active surface maintains the rotation speed, while the driven surface remains stationary. Interpolation is performed based on the area ratio of the driven surface and the active surface grid in the grid area. This interpolation method ensures the continuity of the values ​​of the rotating and non-rotating areas, so that the entire calculation can converge, while ensuring the accuracy of the dynamic mesh calculation.

[0065] Step 7: Decompose the solution area and discretize the corresponding discretized matrix to different numbers of cores. The appropriate decomposition method greatly improves the parallel efficiency: when decomposing the overall calculation area, the decomposition method based on the amount of calculation and the amount of grids is adopted. The amount of grids inside the rotating area is large. Due to the existence of interpolation, the overall amount of calculation is also higher than that of the non-rotating area. The number of decomposed cores in the rotating area accounts for a large proportion, and the number of decomposed cores in the non-rotating area is relatively small.

[0066] Step 8: Use the corresponding matrix solution method to solve the discretized linear matrix for each matrix on the kernel: Use the multi-grid method to solve the flow field for each matrix. After reaching the predetermined convergence condition, as the calculation time increases, the overall calculation flow field gradually stabilizes. The main parameters to be solved include the velocity field and the pressure field. Finally, the flow field changes at a specific calculation area at a certain time and the flow field diagram of the streamline changes can be obtained. See Figure 5 .

[0067] The present invention has been completely experimentally tested and the numerical calculation of a small dual-rotor three-rotor vertical axis wind energy generator has been completed.

[0068] Summary of some experimental results:

[0069] The small eddy current in the middle of the rotor of the vertical axis wind turbine performs well. Figure 6 .

[0070] The calculation streamlines of the vertical wind turbine are evenly distributed and clear, and the overall calculation resolution is high. Figure 7 .

[0071] This calculation was tested on the supercomputer "Tianhe-2" and the parallel efficiency was excellent. Figure 8 .

[0072] In the specific implementation of the present invention, 1. The present invention is based on the NS equation of large eddy simulation, and the eddy current performance of the calculation result in the overall calculation is relatively accurate, the simulation result is highly reliable, the eddy current details are fully displayed, and the overall calculation result has a high resolution.

[0073] 2. The present invention adopts a dynamic grid method for the rotating area, and the overall calculation adopts a transient calculation method, which can clearly show the development of the flow field at each stage, and the calculation is more accurate. The dynamic grid is interpolated using an arbitrary grid interface method. This calculation method ensures the continuity of the results of the rotating area and the non-rotating area, and ensures the accuracy of calculating the range of the rotating area.

[0074] 3. The whole set of calculation methods proposed in the present invention can have good parallel scalability on the supercomputer "Tianhe-2", and when the total number of cores reaches 1,000 cores, there is still a parallel efficiency of more than 50%, and its parallel efficiency is extremely high.

[0075] 4. Compared with the existing simulation methods for rotating fluids, the present invention can achieve high-resolution, large-scale parallel numerical calculations for vertical wind turbines with complex multi-rotors.

[0076] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine, characterized in that: The following steps are involved: Step 1: Create a geometry file of the calculation area used for calculation; Step 2: Segment the rotating area and the non-rotating area, and create a geometric shape file that wraps the entire rotating area; Step 3: Generate mesh for the entire calculation area, which is divided into two parts: the rotation area and the non-rotation area; Step 4: Establish fluid control equations and boundary conditions; Step 5: When solving the control equation in step 4, the corresponding turbulence model equation needs to be introduced; Step 6: After the original model is discretized using the finite volume method, the rotation area and the non-rotation area need to be interpolated on the grid of the geometry file in step 2. The interpolation method uses the arbitrary grid interface method. Step 7: Decompose the solution region and discretize the corresponding discretized matrix to different numbers of cores; Step 8: Use the corresponding matrix solution method for each matrix on the kernel to solve the discretized linear matrix.

2. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: The geometry files in step 1 include the geometry files of the vertical axis wind turbine, the geometry files of the entire region that the flow field of a single wind turbine may affect, and the geometry files of the inlet and outlet for wind inflow.

3. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step 2, for the rotor rotating around the axis, the dividing interface should be cylindrical.

4. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step 3, the entire calculation area is divided into structured hexahedral grids. The rotating area grid and the non-rotating area grid are separated by the cylindrical surface in step 2 and corresponding surface grids are formed.

5. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: The control equation in step 4 is for an incompressible Newtonian fluid, which should satisfy the Navier-Stokes equations for an incompressible fluid.

6. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step five, a large eddy simulation filter is used for filtering, and the Sparat-Almaraz turbulence model is introduced for calculation, and the eddy delay separation method is used.

7. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step six, a driven surface and an active surface are set on the surface that divides the rotating area and the non-rotating area. The active surface maintains the rotation speed, while the driven surface remains stationary. Interpolation is performed based on the area ratio of the driven surface and the active surface meshes in the mesh area.

8. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step 7, the decomposition method based on the amount of calculation and the amount of grid is used to solve the regional decomposition.

9. The method for numerical simulation of aerodynamics of a multi-rotor vertical wind turbine according to claim 1, characterized in that: In step eight, the flow field is solved for each matrix using the multi-grid method.

Images