考慮風向性對局部設計風載重的影響

Estimation of local wind loads considering wind directional effects

李紹緯(Shao-Wei Li)

指導教授 ： 羅元隆

淡江大學/工學院/土木工程學系碩士班/碩士(2020年)

https://doi.org/10.6846/TKU.2020.00491

摘要

本研究首先以台北測站為對象，調查該測站於1961年至2019年間所有颱風期間平均十分鐘風速紀錄，並依每15度角為一個風向進行分類，並將風速轉換為實際建物高風速。取每個颱風在各角度下的最大風速值以及不分風向下的最大風速值，繪製其機率分布、累積機率分布，同時擬合甘保分布以觀察其趨勢。結果發現採集到的極值風速資料不全然屬於甘保分布趨勢，後續分析則可分別採用甘保或廣義極值分布的假設來作比較。接下來，本研究以淡江大學第一號大氣邊界層風洞實驗室進行高層建築模型的氣動力風壓試驗，量取模型表面長時間的風壓分佈歷時。風洞試驗採用我國規範中的鄉鎮型地況作為模擬實驗流場。高層建築模型以真實CAARC大樓建築物作為1/400縮尺模型。每15度風攻角下進行實場約莫六十個小時的長時間風壓量測。擬合風壓係數後發現同樣不全然屬於甘保分布或廣義極值分布。最後藉由羅和Kasperski (2018)[1]所提及的乘冪積分方法，針對不同極值分佈型態與風向角變化與否，計算出七種不同組合的設計風載重進行比較。
結果顯示，方向性的影響造成各風向的風壓載重設計值均呈現下降分布；以最佳非超越機率和78%分析之結果對比，兩者計算出之風載重設計值相差不大，誤差值落在5%左右。以甘保分布作為極值風速及極值風壓係數分布的假設與以廣義極值分布為主的假設結果做比對，兩者的誤差值有至少20%左右的差距，廣義分布之假設設計風載重值皆大於甘保分布之假設設計風載重值，且兩者的趨勢走向相同。以全風向及分風向極值風速及風壓係數為主的假設，大部分風載重結果誤差值差距40%以上甚至於達到100%。全風向之設計風載重值皆大於分風向之設計值，但分風向角之設計值受權重影響較為顯著。以風向角權重方面去做比較，統計出來之權重與均權(角度總數分之一)作結果比對，兩者之設計風載重值趨勢起伏相近，但沒有明顯的單一偏大的趨勢。以是否折減風速去做比較，先做折減風速的設計風載重比用權重折減的設計風載重值來的大，但始終坐落在現在使用之設計值與分風向角甘保設計之設計值之間。在所有之比較分析中，以分風向角下甘保假設的設計風載重為最小，以其做設計較不保守但在合理範圍；以廣義極值假設之全風向角設計值最大，以其做設計則是最為保守。而目前所用之78%非超越機率之全風向設計風載重計算結果則在兩者之間偏向保守。

關鍵字

風向性 ； 非超越機率分布 ； 甘保分布 ； 廣義分布 ； 年破壞超越機率

並列摘要

This research aims to investigate the effects of wind directions and extreme distribution types on design wind loads. In the first phase, the records of the Taipei Meteorological Station were adopted for the extreme wind speed observation. It was found that in different wind sectors, the Gumbel distribution may not always be the best fitting model for extreme distribution type. In the second phase, the long-term pressure measurement of a scaled CAARC building was also conducted for the identification of extreme distribution types for different locations. In the third phase, the convolution iteration method to estimate design wind loads was adopted to evaluate the effects of wind direction changes and the distribution types. In total, there are seven combination scenarios for discussions.
Results showed that the consideration of wind direction changes generally reduced design wind loads at the range of around 5% in comparison with the Cook and Mayne method. The consideration of the generalized extreme distribution (GEV) altered significantly from the Gumbel model. The difference was found at least 20%; the design wind loads based on GEV were larger than that based on Gumbel. Considering different probabilities of wind directions gave little differences in design wind loads. Among all the seven scenarios, the case with different wind directions under Gumbel assumption produced the smallest design wind loads; while the case with unique wind direction under GEV assumption produced the largest design wind loads. The conventional Cook and Mayne method is in between and tends to conservative values.

並列關鍵字

wind direction ； non-probability ； Gumbel distribution ； GEV distribution

參考文獻

[1] Lo, Y.L , M.Kasperski (2018)“以極值分析理論推估最佳設計風壓係數”.

[2] M.Kasperski , Lo, Y.L (2017)“Directional design of a high-rise building”, Proc. 9th Asia-Pacific Conference on Wind Engineering Auckland, New Zealand 3-7 December 2017.

[3] W.H.Snyder(1972)“Similarity criteria for the application of fluid models to the study of air pollution meteorology”, Boundary Layer Meteorology , Vol.3 , pp.113-134.

[4] J.E.Cermak(1975)“Application of fluid mechanics to wind engineering”, A freeman-scholar lecture , Journal of Fluids Engineering , ASME , Vol.97 , pp.9-38.

[5] Nakamura, Y., Ohya, Y. (1984)“The effects of turbulence on the mean flow past two dimensional rectangular cylinders”, J. of Fluid. Mech., Vol.149, pp.255-273.