高层建筑标准模型绕流数值模拟

本文选用Standard ke 、RNG ke和Realizable ke 3种高雷诺数湍流模型，对CAARC标准高层建筑模型在不同网格划分方案、不同风场环境及不同风向条件下的流场特征进行了模拟，并用风洞试验数据验证模拟效果，得到以下结论:(1)建筑物边界层网格模型的细密程度会对模拟结果产生明显影响，当选用最细密的网格模型进行模拟时，RNG ke模型与风洞试验数据的均方根误差要比选用最稀疏的网格模型时小50%~70%，其余两种模型则大约相差40%~50%。(2)在B类、D类两类风场环境下，RNG ke 模型所得平均风压系数的变化区域及变化趋势基本一致，无明显差别，而其余两种模型在不同风场环境下的变化趋势及变化范围均有较大不同，差别主要表现在建筑物的侧风面和背风面上。(3)0°风向下，RNG ke模型在建筑物侧风面和背风面处的平均误差要比其余两种模型小10%左右；90°风向下，RNG ke模型在侧风面处的平均误差比其余两种模型小大约5%。总体来说，在多种模拟条件下，RNG ke模型所得结果的准确性、稳定性及规律性均要高于其余两种模型。

Numerical Simulation of Flow around a Standard High-Rise Building Based on Three Turbulence Models

In this paper, Standard ke, RNG ke and Realizable ke are three turbulence models of high Reynolds number are used to simulate the CAARC model under different meshing, different wind fields environment, and different wind directions. The numerical simulation results were compared with wind tunnel test data at home and abroad, the conclusions are as follows: (1) The precision of the building boundary layer grid model has an important influence on the simulation results. When the best grid model is selected for simulation, the RMSE of RNG ke model is equal to that of wind tunnel test data, and the root mean square error of RNG model is 50% - 70% less than that of sparse grid model, and the error of other two models is about 40% - 50%. (2) In the case of B-type and D-type wind farms, the variation area and trend of the average wind pressure coefficient obtained by RNG ke model are basically the same, and there is no obvious difference, while the variation trend of the other two models is in different wind farms. The change range of environment is also very different, the difference is mainly reflected in the crosswind side and leeward side of the building. (3) On the leeward side, the average error of the model is about 90% smaller than that on the leeward side, and about 5% lower than that of the other two models. Generally speaking, under different simulation conditions, the accuracy, stability and regularity of the results obtained by RNG ke model are higher than those of the other models.