Numerical Simulation and Analysis of the Arrangement of Wind Direction Meter for Wind Turbine

. This paper performs CFD numerical analysis on the wind direction variation characteristics of the upper part of the platform under the running condition of the whole machine, and analyses whether the current location of the equipment is suitable for placing the wind direction measuring equipment and finding a better position in the area. The vicinity of the reference point is the area where the airflow is relatively intense, and the wind direction is complex and changeable. In-depth CFD research on this flow phenomenon can obtain an intuitive wind direction state and inherently subtle changes.


Introduction
The wind direction meter is a high-precision measuring instrument for measuring the wind direction of the wind turbine in the wind field and is used for measuring the deviation angle between the wind direction and the central axis of the impeller [1]. The controller adjusts the rotation axis of the fan impeller according to the wind direction deviation signal through the yaw system to match the wind direction. The purpose of this paper is to maximize the efficiency of wind energy utilization, reduce the impact of wind direction changes on wind turbine load, achieve better power output, and improve unit efficiency. [2][3][4][5].
The wind direction meter is mounted on the top weather frame of the nacelle cover, which is susceptible to interference from blade rotation. Therefore, the actual power curve of the unit is seriously affected, so that the output power of the unit cannot meet the design requirements, and the utilization of wind energy and the efficiency of the unit are reduced [6][7][8]. The purpose of this paper is to use numerical simulation to select the best wind direction arrangement position on the nacelle cover to minimize the interference of the blade rotation on the wind direction meter. Ensure the safe and economic operation of the unit [9][10][11][12].

Model description
First, based on the analysis of a working condition, the effectiveness of the calculation process and method is investigated. Make an alternative layout point for analysis and comparison.
According to the provided three-dimensional digital model, the mesh model required for CFD calculation is established.
This project uses the relevant geometric functions of CATIA and Hypermesh to complete the geometric cleaning of the original digital modules of the main components blades, hubs, towers and nacelles, and assembles the components into a complete machine model. Based on the previous geometric cleaning, the Hypermesh related surface mesh creation function is used to divide the surface mesh. We used the trimming grid function in STAR-CCM +. The volume mesh required for the calculation is generated for the outer flow field area and the blade rotation area of the wind turbine, and the boundary layer of the wind turbine surface is simulated using the grid in the form of a prism.

Establishment of flow field model
The outer flow field is established by the boundary of the flow wall with the blade, the hub, the nacelle and the tower surface. The blades and hub are rotating mechanisms throughout the analysis, creating a rotating grid in this area. The 3D geometric model of the whole machine is shown in  The calculation domain of the whole machine is divided into two parts, the first part is the rotating area of the wind wheel, and the second part is the flow area of the external flow field. The enclosed area of the whole machine and the blade wind wheel is shown in the figure below.

Face meshing
The surface mesh is divided by the triangle mesh type. The division requirements are: no free edges, no repeating elements, the minimum angle of the mesh is greater than 20°; the specific settings of the mesh size are shown in the following table.

Body grid
The volume mesh adopts the unstructured hybrid grid layout. The specific parameters of the boundary layer and multiregion encryption are: when the total thickness of the boundary layer is 1mm, 2 layers, the growth ratio is 1.3。 The size of the encryption area is 100-200 near the cabin, and the tail The flow area is 0.6-1m. After repeated debugging, more than 9.4 million body meshes are finally generated, of which the blade rotation area is 3.4 million.

Environmental parameters
The environmental conditions for the simulation calculation are shown in the following table and  There is no need for vents and vents in the computer compartment. The flow rate at the vent is 10,800 cubic meters per hour.

Calculation parameters
1. Calculate the gas flow in the domain according to the incompressible, constant flow calculation, using a doublelayered wall function, 2. The inlet is the speed import boundary condition, and the wind shear is considered when the inlet speed of the hub is 8.96 m / s. Free flow boundaries are used around the exit and flow fields. 3. The convergence accuracy of the calculation is 0.001, and the second-order upwind difference format is adopted. 4. Transient calculation: The time step is 0.025s, the maximum number of iteration steps in the unit step is 20, and the total physical time is 60 seconds.

Transient results
The Fig 8 shows the real-time streamline diagram of the whole machine. It can be seen from the figure that the blade is the main source of interference in the transient, followed by the platform. The position near the reference point is the position where the airflow is relatively sharp, and the indepth CFD study is very necessary.  Fig. 9, and the data to be optimized is output (pink area).

Comparison and selection
The data of all wind speed monitoring points is converted into the wind direction angle, and the wind direction fluctuation curve is better than the position of L ref, R ref by comparing the wind direction real time curve.
The following Fig 15-22 is a real-time curve of wind direction for all monitoring points in the optimized area.            From the comparison results, the selected wind direction fluctuation is more stable than L ref, R ref , and the swing amplitude is smaller. It can also be seen that the flow direction of the wind is 8° or -8°, and the advantages of the multiple groups and the position of the most advantageous are unchanged, indicating that the change of the flow direction does not affect the distribution of the advantages and disadvantages of the wind direction fluctuation. The most advantageous coordinates are shown in Table 3 below, and the relative positions are shown in Fig. 27.    From the comparison of real-time wind direction results, the wind direction fluctuations in the two states with or without vents are basically the same, and the addition of vents has little effect on the wind direction near the reference point.

Conclusions
1. From the transient calculation results (speed, vortex dynamic map), it can be seen that the vicinity of the reference point is the region where the airflow is relatively intense. It is necessary to conduct in-depth CFD research on this flow phenomenon. 2. By comparing the real-time wind direction curve, the best advantage of wind direction fluctuation can be obtained, and the selected point wind direction fluctuation is more stable and the swing amplitude is smaller. 3. The change of the flow direction does not affect the distribution of the advantages and disadvantages of the wind direction fluctuation. In the case of incoming wind directions of 8° and -8°, the advantages of the multiple sets and the most advantageous positions remain unchanged and tend to be away from the surface of the nacelle. 4. With the change of the incoming flow direction, the fluctuation of the wind direction near the reference point in a single cycle (vibration shape, amplitude) does not change much, and the change is only the fluctuation trend in the large period of the oscillation, and the flow direction is opposite. The fluctuation trend in the big cycle of the shock is also reversed. 5. The wind direction fluctuations in the two states with or without vents are basically the same, and the addition of vents has little effect on the wind direction near the reference point. The above results are obtained based on the analysis of the two wind direction conditions of the same wind speed. It is possible to calculate more wind speeds and different wind down characteristics for verification, and then make a conclusion after global comparison and optimization.