The Probability Characteristics Of Wind Pressure And Calculation For Wind Load

Estimation of the extreme wind pressure has been one of the focus of the wind engineering research, early extreme wind pressure estimation was mainly based on the assumption which wind pressure followed the Gaussian distribution. As people de epened the cognition and research in characteristics of the wind pressure distribution, and they gradually realize the universality and importance of non-Gaussian distribution on wind pressure, and early assumption of the extreme value calculation no longer set up, the error in the estimation of extreme value of wind load was inevitable, so it was significant to explore a more reasonable algorithm. The design of structure envelope on the surface of high-rise building mainly depended on the est imation of extreme wind load, therefore, it was necessary for us to analysis the distr ibution of wind pressure and calculate the extreme value of wind pressure on the surface of high-rise building.Based on the relationship between the generalized extreme value distribution and the generalized Pareto distribution and autocorrelation coefficient of the sample data of wind pressure, we got the calculation method of extreme wind pressure; through the samples of the surface wind pressure obtained by the study of the field measur ement and wind tunnel test of high-rise buildings by the coastal, the results are co mpared by using the method with the improvement of Gumbel method, improvement of peak factor method and Sadek_Simiu method, it was showed that the method in this paper was suitable for high-rise buildings in estimation of extreme wind pressure.Firstly, the paper introduced the field test a nd wind tunnel test of high-rise building in southeast coast Xiamen. The southeastern coast of China region was the main region for the impact and landing by typhoon, with the development of economy of our country and the more quantities of the high-rise building in southeast coastal city, it was needed to do some in-depth study for the wind resistant design for high-rise buildings. Based on the comparison between field test and wind tunnel test of Xiamen high-rise building’s wind pressure, the distribut ion and probability characteristics needed to be statistical analysis. The result showed that two kinds of su rface wind pressure distribution showed no difference in distributing characteristics. but there were big differences in statistic al, that mainly due to the wind field around buildings was more complex than wind tunnel test, which caused the non-Gaussian characteristics of wind pressure in field was more obvious.Secondly, two kinds method for extreme wind pressure were intr oduced in detail respectively, and explains the advantages and disadvantages of each method. I mproved the Gumbel method used the conversion between different intervals and mi nimum observation interval which was got by the autocorrelation coeff icient to break the limitation of sample size. The minimum observation interval was used to scree ning the extreme sample data for calculating the extreme distribution parameters so as to achieve the desired extreme value. but,because of the conversion between the i ntervals,the estimation for extreme distribution parameter in long interval was sens itive to high tail part of the extreme distribution in short interval,so it was necessary to get more accurate judgment for high tail part. Based on the relationship between the generalized Pareto distribution and the extreme value distribution, the desired e xtreme and distribution were obtained.Finally, through the analysis of the wind pressure sample on the surface of the high-rise building which was measured by the field test and wind tunnel test, the results are compared by using the method with the improvement of Gumbel method, improvement of peak factor method and Sadek_ Simiu method, it was showed that the method in this paper showed better performance.