Abstract:

Wind speed and turbulence are two closely related indicators that measure the properties of a wind profile. Obtaining an insight into the development of turbulence over a complex urban terrain can help deepen the understanding of the performance of urban wind farms. In this research, three building models with vertical scales 1:2000, 1:1000, and 1:500, respectively, were constructed. Using large boundary-layer wind tunnels and generating wind from two directions (northwest and southeast), the variation in turbulence with height over a complex urban terrain, and its dependence on the macroscopic terrain characteristics, were analyzed in a neutral flow simulation. Based on the experimental data obtained in the wind tunnel, the two model coefficients A and B were determined with respect to four types of boundary-layer roughness, under neutral flow or with turbulence varying with height at different vertical scales. In both cases, the average correlation of the proposed model was about 0.8. A close relationship between the wind profile index and the turbulence at different heights was observed. Based on the profile index α the turbulence at different heights could be predicted, so that the variation in turbulence with height over a complex urban terrain could also be quantified. Generally speaking, the turbulence decreases with altitude, and the maximum turbulence develops at the bottom. However, there are exceptions. It is common that the turbulence of the hole at the lowest measuring point is not the largest, which makes the shape of turbulence change with the height like a hook. The shape of turbulence varying with height can be summarized into four types. The height at which the maximum turbulence occurs is found to be concentrated in the range 0-0.2 h (where h is the dimensionless unit), which makes up more than 80% of the total number. Therefore, in the height range 0-0.2 h above the urban terrain, the wind direction and velocity of the airflow showed complex patterns, and turbulence is extremely developed, with an important impact on the diffusion of urban pollutants and the transfer of heat. Using the existing model, the main coefficient of the turbulence model corresponding to a given height could be determined, with high precision accuracy, according to the four kinds of boundary-layer roughness and the different vertical scales. The development along the height of the non-maximum turbulence intensity depended on the difference between the actual wind profile and the standard wind velocity at the same height, whereas the maximum turbulence level under the given urban topography occurred within the narrow range 0-0.2 h. The turbulence degree index β was used to characterize the variation of turbulence intensity with height. The exponential β of the turbulence intensity decreased with increase in the exponential alpha of the profile, regardless of the shape of the terrain (e.g., a ridge or a flat terrain). It was shown that the overall turbulence profile increases from the upwind and top areas to the leeward area, and increases gradually along the wind flow direction. Turbulence profile also has a strong dependence on the terrain and the wind path, and has the same flow characteristics as those over a simple terrain. At the same time, the shape of the β isolines of the three models did not show the same overlap. On the contrary, great differences were observed. This shows that in the past, when the wind tunnel simulations were carried out, the method of ensuring the number of thunderbolts simply by increasing the vertical scale was affected by a large uncertainty.

Key words: urban terrain, turbulence, profile index, vertical structure, wind tunnel simulation, Guangzhou