基于CFD技术的陡峭山体风场模拟方法研究

The application of CFD techniques on the wind field simulation over steep mountains:A method study

为更好将计算流体力学(CFD)技术应用于陡峭山体的风场模拟研究,文中对其中两个关键技术问题进行了研究,一是适宜的CFD解域顶高问题,二是入流边界条件设置方法问题。研究采用了香港国际机场的多普勒雷达径向速度观测资料,通过CFD模拟结果与观测结果对比验证展开研究。对于解域顶高,一般环境流体力学或建筑风工程领域要求顶高越高越好,例如5倍于地面障碍物的垂直高度,究其原因,主要是为了获得地面障碍物周边的整个流场结构,避免顶边界过低影响障碍物顶部流场的准确性。通过数值试验发现,若模拟关心的区域为较低海拔高度时,不必完全拘泥于该要求,对一座高度为1000m量级的山体进行模拟,顶高为3000和6000m的两组试验在边界层中低部给出的模拟结果没有显著差别。进行了3组边界条件设置方法的对比试验,结果表明单纯采用廓线法不足以充分描述来流信息,二维插值法尽管提供了更高分辨率的边界数据,但其结果总体上逊于0维插值法。研究表明,CFD可以很好地描述山地激发的涡旋/波动脱体运动,比单纯采用中尺度模式效果要好得多。在利用CFD工具研究复杂山地风场时,应本着"实用主义"的精神设置物理模型、参数和选取适当的方法、边界条件,以达到模拟精度和计算量的优化平衡,并足以用于研究所关心的具体问题。

Two technical issues related to the application of Computational Fluid Dynamics （CFD） on the wind simulation over steep terrains are discussed in this paper. The first issue is on how to appropriately set the top height of the simulation domain, and the second one is on how to appropriately set the inlet flow boundary conditions. The observed radial velocity data collected from a Doppler radar that is deployed to the northeast of the Hong Kong International Airport are used to validate the CFD sim- ulation results. For the first issue, some previous studies on environmental fluid dynamics or structure wind engineering point- ed out that a higher top height, such as 5 times of the vertical height of the obstacle, will lead to a better simulation since a higher top can help to avoid negative impacts of an inappropriately low top boundary on the flow field around the obstacles. However, in the current study, it is showed that if the focus area of the study is in a low elevation area, it is unnecessary to set a high domain top height. The numerical tests show that for a mountain with an elevation of around 1000 m, the difference in the simulated wind fields in low layers is not significant between the two tests in which the top heights of domain are set to be 3 times and 6 times of the obstacle height respectively. For the second issue, three numerical tests are carried out to compare the performances of three methods, namely the single profile （SP） method, the zero-dimensional interpolation （ZDI） method, and the two-dimensional interpolation （TDI） method. The results show that applying the SP to the whole inlet boundary is not sufficient to describe the inlet information, and the simulation results are the worst among the three ones. Surprisingly, though the TDI method can provide higher resolution data on the inlet boundary, the simulation results obtained by the TDI method are generally poorer than those obtained by the ZDI method. This study shows that CFD can provide better simulation results on vortex/wave shedding processes triggered by steep terrains compared to mesoscale models. When using CFD on wind simulation over a complex terrain, a ＂practical oriented＂ setting on physical models, parameters and boundary conditions is encouraged to use in order to get an optimized compromise between the simulation accuracy and the computation load.