以RANS與DES紊流模式預測高層建築群地下室與行人層風場

Wind field of a high-rise building predicted by RANS and DES: basement and pedestrian level with multiple surrounding buildings

李柏勳(Po-Hsun Lee)

指導教授 ： 朱錦洲 ； 張建成

國立臺灣大學/工學院/應用力學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202102760

摘要

台灣位處在歐亞大陸與太平洋之全球最大陸地與海洋間交界位置，有特徵強烈的季節盛行風，特別是在冬季時，風速強勁且風向集中。而台灣係一高度城市化之島嶼，市區內高樓林立且城區高速擴張，在季風吹拂下高樓群之周圍風場之相關研究便顯得重要。由於高樓的排列會形成類似峽谷地形般之街谷，造成風流經縮口會有加速的現象。由於縮口造成的高風速陣風，在高樓周圍步行區域中經常導致人員不適，在高樓地下室之設備也經常由於車道的氣體引流導致風壓過高致使故障。
由於計算流體力學之日漸完善，且計算機性能高速發展，多建築群之周圍紊流場模擬變得可行。紊流計算流體力學之理論中，以雷諾平均法、大渦理論為大宗，現行之主流紊流模型包含雷諾平均納維爾史托克方程(RANS)、分離渦模式(DES)、大渦模式(LES)等等，這些不同的紊流模型都有其各自的應用場景與特性。本研究於台北某大樓之樓頂與地下室車道入口設置共三處測站，紀錄平均風速、陣風風速與風向等參數，以作為模擬之條件設置之參考，並利用測量結果與兩種紊流模式進行比對，比較風速之平均與變異數特徵與大樓尾流區的紊流特徵。本研究將延續先前研究，進行其他條件之流況探討，了解風場中不同條件對於設備之影響。本研究亦針對地面層風場進行舒適度分析，使用量測－模擬擬合的方式來預測行人層風場強度與機率，以了解特定位置步行區之行人風舒適度。
本研究使用大氣邊界條件之指數半經驗公式作為入流條件，搭配二至三棟的高層建築模型設為壁面無滑移條件，並將下游區設置一壓力出口(pressure outlet)，以進行數值模擬。在RANS與DES對於尾流區特徵之表現上，發現尾流渦度以DES較為破碎，RANS則較為平滑。以高層建築群為模型，在高雷諾數(〖10〗^6~〖10〗^8)下，RANS與DES之頂樓平均風速模擬值差異則有限，兩模式模擬值相對量測值有偏低的現象。在地下室之風場表現上發現，相比先前研究的方法，以全時開啟排氣通道與排氣扇可以有效降低設備上的風壓，是一以現存已有之設備即可改善的方法。在行人層舒適度上，由於高速的風經常吹拂，以三種舒適度標準進行評估後，發現該街谷中特定區域不適合人員長時間逗留。此外，以蒲福風級將該區域之平均風速繪製成圖後發現，不同風向會有不同的形態，推測與建築物排列及馬蹄渦協同下洗沖流之效應有關。

關鍵字

高層建築群 ； 計算流體力學 ； 紊流模擬 ； 行人層 ； 馬蹄渦 ； 風舒適度標準

並列摘要

Taiwan is located at the junction of the world's largest land and ocean between the Eurasian continent and the Pacific Ocean. It has a strong seasonal monsoon, especially in winter, the wind speed is high and the wind direction is concentrated. Besides, Taiwan is a highly urbanized island. There are many high-rise buildings in the urban area and the it is expanding rapidly. Research on the wind field around the high-rise buildings under the blowing of the monsoon becomes important. Because the arrangement of tall buildings forms a street valley similar to the terrain of a canyon, the wind accelerates when passing through the narrow mouth. In the pedestrian zone around the tall buildings, that is, the pedestrian zone, under high wind speed gusts, it often causes a phenomenon of personnel discomfort. Equipment in the basement of high-rise buildings often fails due to excessive wind pressure caused by air drainage from the driveway.
With the improvement of computational fluid dynamics and the exponential growth of computer performance, simulation of turbulence field around multi-building groups became feasible. In the theory of turbulence computational fluid dynamics, the Reynolds average method and the large eddy theory are the main ones. The current mainstream turbulence models include the Reynolds average Navier-Stokes equation (RANS), detached eddy simulation (DES), large eddy simulation (LES). These different turbulence models have their own advantages and disadvantages. In this study, a total of three measuring stations were set up on the roof of a building in Taipei and at the entrance to the basement driveway to record the average wind speed, gust speed, wind direction as a reference for setting the simulation conditions. The measurement results and two turbulence models are used to conduct the study. The study compares the characteristics of the mean and variance of wind speed and the turbulence characteristics of the wake area of the building between two turbulence models. This study continues the previous research and discusses the flow conditions to understand the impact of different conditions on the equipment in the wind field and the pedestrian wind comfort near the buildings. This study also conducts a wind comfort analysis on the ground level, and uses the measurement-simulation fitting method to predict the strength and probability of the pedestrian level wind field to understand the wind comfort in the pedestrian zone at a specific location.
In this study, the exponential semi-empirical formula of atmospheric boundary conditions is used as the inflow condition, and the walls of two to three high-rise building geometry models and the ground are set as no-slip conditions, and a pressure outlet is set in the downstream outlet area for numerical simulation. In the performance of RANS and DES on the characteristics of the wake area, it is found that the wake vorticity is relatively broken with DES while RANS is relatively smooth. Taking a high-rise building group as a model, under high Reynolds numbers (〖10〗^6~〖10〗^8), the difference between the simulated values of the average wind speed on the top floors between RANS and DES is limited, and the simulated values are biased relative to the measured values. The phenomenon that the predicted value be lower than the measured value is found. On the performance of the basement wind field, it is found that comparing with the previous research method, opens the exhaust channel and turns on the exhaust fan at all times can effectively reduce the wind pressure on the equipment, which is a method that can be improved with existing equipment. In terms of pedestrian level wind comfort, due to winds often gusting, after evaluating the three comfort standards, it is found that the specific area in the street valley is not suitable for long stays. In addition, after plotting the average wind speed in this area with Beaufort wind scale, it is found that different wind directions will have different forms, which is presumably related to the arrangement of buildings and cooperation between horseshoe vortex and downwash.

並列關鍵字

high-rise buildings ； computational fluid dynamics ； turbulence simulation ； pedestrian level ； horseshoe vortex ； wind comfort criteria

參考文獻

[1] P. Panagiotou, P. Kaparos and K. Yakinthos, "Winglet design and optimization for a MALE UAV using CFD," Aerospace Science and Technology, vol. 39, pp. 190-205, 2014.

[2] J. Y. J. Kuo, D. A. Romero, J. C. Beck and C. H. Amon, "Wind farm layout optimization on complex terrains – Integrating a CFD wake model with mixed-integer programming," Applied Energy, vol. 178, pp. 404-414, 2016.

[3] T. Glatze, C. Litterst, C. Cupelli, T. Lindemann, C. Moosmann, R. Niekrawietz, W. Streule, R. Zengerle and P. Koltay, "Computational fluid dynamics (CFD) software tools for microfluidic applications – A case study," Computers Fluids, vol. 37, pp. 218-235, 2008.

[4] J. S. Jayakumar, S. M. Mahajani, J. C. Mandal, P. K. Vijayan and R. Bhoi, "Experimental and CFD estimation of heat transfer in helically coiled heat exchangers," Chemical Engineering Research and Design, vol. 86, pp. 221-232, 2008.

[5] V. Knop, A. Benkenida, S. Jay and O. Colin, "Modelling of combustion and nitrogen oxide formation in hydrogen-fuelled internal combustion engines within a 3D CFD code," International Journal of Hydrogen Energy, vol. 33, pp. 5083-5097, 2008.