我國耐風設計規範中，建築物頂層角隅處之振動加速度極值的計算，是基於順風向振動與橫風向振動及扭轉向振動不相關，橫風向振動與扭轉向振動完全相關，且各方向振動加速度為高斯分布之假設估計而得。本研究將考慮三個方向振動加速度之相關性，以及非高斯分布之尖峰因子，以修改規範公式，其中非高斯分布之尖峰因子可藉由Hermite多項方程式求得。本研究先利用CFD模擬地況B風場下，深寬比為1與0.5及高寬比為6、5與4的矩形建築之風壓歷時，將結果與風洞資料比較，發現模擬之結果整體趨勢表現上跟風洞實驗是一致的，但是兩者風力頻譜值大小仍有差距，因此後續加速度計算仍使用風洞實驗資料。頂層角隅處之振動加速度極值估計之準確性與形心處三個方向振動加速度之相關係數、Skewness及Kurtosis有關，本文使用兩種方法估計頂層角隅處之振動加速度極值，方法一假設形心處三個方向振動加速度呈高斯分布，方法二根據風洞實驗資料，實際估計形心處三個方向振動加速度之機率分布。將上述兩種方法所得結果作比較後，可以發現當角隅處總振動加速度之歷時為軟化(Softening)過程時，方法二所得結果會大於方法一；當角隅處總振動加速度之歷時為硬化(Hardening)過程時，方法二所得結果會小於方法一。另外由於形心處橫風向與扭轉向加速度相關係數通常遠小於1，方法一所得結果會小於規範值；即使當橫風向與扭轉向頻率為1且風速越小和建築物高度越大的情況下，形心處橫風向與扭轉向加速度相關係數較接近1時，方法一所得結果依舊小於規範值。本文最後同時考慮各方向加速度之相關係數及非高斯特性，提出頂層角隅處振動加速度極值之估計公式。

In Taiwan's wind-resistant design codes, the calculation of the extreme acceleration of vibration at the corners of the top of the building is based on the assumptions that the alongwind acceleration is not related to the crosswind acceleration and the torsional acceleration, the crosswind acceleration is completely related to the torsional acceleration and all accelerations follow the Gaussian distributions.In this study, we will consider the correlation of the accelerations in three directions and all accelerations follow the non-Gaussian distribution to modify the current formula. The peak factor of the non-Gaussian distribution can be obtained by Hermite polynomial equation. In this study, CFD was used to simulate the wind pressures of rectangular buildings .The results were compared with wind tunnel datas, and the simulation results were found to be overally consistent with wind tunnel experiments, but there is still a gap between the wind power spectrum values of the two, so the subsequent acceleration calculations still use wind tunnel experimental datas. The accuracy of the extreme value of the accelerations at the top corner is related to the correlation coefficients, Skewness, and Kurtosis of the three-direction
v
accelerations at the centroid. This study uses two methods to estimate the extreme value of the accelerations at the top corner.Method 1 assumes that the acceleration in three directions at the centroid is Gaussian distribution. Method 2 According to wind tunnel experimental data.The probability distribution of acceleration in three directions at the centroid is actually estimated.After comparing the results obtained by the above two methods, it can be found that when the duration of the total acceleration at the corner is a Softening process, the result obtained by the Method 2 is greater than that of the method 1.When the duration of the total accelerations at the corner is Hardening process, the results obtained by Method 2 will be less than Method 1.In addition, since the correlation coefficient between the across-wind direction and the torsional acceleration at the centroid is usually much smaller than 1, the result obtained in method 1 will be less than the TDC value.Finally, this study considers the correlation coefficients and non-Gaussian characteristics of acceleration in all directions at the same time, and proposes an estimation formula for the extreme value of accelerations at the top corner.

參考文獻
[1]Architectural Institute of Japan (2004), Recommendations for Loads on Buildings.
[2]Asami, Y. (2000), “Combination method for wind loads on high-rise buildings”, Proceedings of the 16th National Symposium on Wind Engineering, Tokyo, Japan,pp.531-534 (in Japanese)
[3]Asami, Y. (2002), “Correlation method for wind load combinations”, Document presented at Subcommittee on Wind Loading, Architectural Institute of Japan.
[4]Chen X. and Hung G. (2009), “Evaluation of peak resultant response for wind-excited tall buildings” Eng. Struct. 31 858-868.
[5]Chen X. and Kareem A. (2005), “Coupled dynamic analysis and equivalent static wind loads on buildings with three-dimensional modes”, J. Struct. Eng. 131(7) 1071–1082.
[6]Clough, Ray W. and Penzien, Joseph (1993), Dynamics of structures.
[7]Davenport, AG. (1964), “Note on the distribution of the largest value of a random function with application to gust loading. [J]．ICE Proceedings，28(2)：187-196.
[8]Ernesto Heredia-Zavoni (2010), “The complete SRSS modal combination rule”, Earthquake Eng. Struct. Dyn. 40 1181–1196.
[9]Gong, K. and Chen, X. Z. (2014), “Estimating extremes of combined two Gaussian and non-Gaussian response processes”,Journal of Structural Stability and Dynamics Vol. 14, No. 3 1350076 .
[10]Grigoriu, M. (1995) , “Applied non-Gaussian processes”,
[11]Holmes, J.D. (2002), “Effective static load distributions in wind engineering”, Journal of Wind Engineering and Industrial Aerodynamics 90 91–109.
143
[12]Kasperski, M. and Niemann, H. J. (1992), “The LRC (Load-Response-Correlation) Method: A General Method of Estimating Unfavorable Wind Load Distributions for Linear and Nonlinear Structural Behavior”, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 43, pp.1753–1763.
[13]Melbourne, W.H. (1975), “Probability Distributions of Response of BHP House to Wind Action and Model Comparisons”, J.Ind.Aerod. , 167-175.
[14]Rice, S.O. (1945) , “Mathematical analysis of random noise”,
[15]Robert H. Kraichnan (1969), “Diffusion by a Random Velocity Field”, Dublin, New Hampshire.
[16]Simiu, E. and Scanlan, R.H. (1996), “Wind Effects on Structures: Fundamentals and Applications to Design”, 3 rd Edition, New York, John Wiley and Sons.
[17]Solari, G. and Pagnini, L.C. (1999), “Gust Buffeting and Aeroelastic Behaviour of Poles and Monotubular Towers”, Journal of Fluids and Structures.
[18]Tamura, Y., Kim, C.Y., Kikuchi, H. and Hibi, K. (2014), “Correlation and combination of wind force components and responses”, Class notes for advanced school on wind engineering, Taipei, Taiwan.
[19]Tamura, Y., Ohkuma, T., Kawai, H., Uematsu, Y. and Kondo, K. (2004), “Revision of AIJ Recommendations for Wind Loads on Buildings”, Proceedings of the 2004 Structures Congress.
[20]Wen, Y.K. (1990), Structural Load Modeling and Combination for Performance and Safety Evaluation.
[21]Yang, C.Y. (1985), Random Vibration of Structures.
[22]內政部營建署，「建築物耐風設計規範及解說」，2015。
[23]洪明德，「建築物總風力載重不確定性之分析」，碩士論文，國立
144
台灣科技大學營建工程系，（民國87年）。
[24]呂良正，楊永斌，「以微震量測探討鋼筋混凝土構造建築物之基本振動周期」，內政部建築研究所委託研究報告，(民國88年)。
[25]高士哲，「非穩態風速與矩形建築風載重機率模式之建立及可靠度分析」，博士論文，國立台灣科技大學營建工程系，（民國102年）。
[26]顏靖偉，「方形斷面建築純量風力載重效應之組合」，碩士論文，國立台灣科技大學營建工程系（民國103年）。
[27]劉尚豪，「矩形斷面向量風力載重效應之組合」，碩士論文，國立台灣科技大學營建工程系（民國104年）。
[28]鄭啟明、吳重成、楊承翰，「高層建築耐風設計風力頻譜與風載重之修訂研究」，內政部建築研究所研究報告，(民國96年)。
[29]方富民，「建築結構所受風力之數值模擬與風洞實驗比較驗證研究」，內政部建築研究所委託研究報告，(民國97年)。
[30]林建宏，「建築物氣彈力反應數值模式建構與風洞試驗研究」，內政部建築研究所委託研究報告，(民國100年)。