Experimental Study Of Non-Gaussian Peak Factor Of Low And High Rise Buildings Under Typhoon

At present,in terms of the study of non-Gaussian crest factor,wind tunnel tests are mostly used,and few researches on measured samples are involved.Moreover,the research process often corresponds to the single building type,and comparative studies of multiple building types are rarely conducted.In this paper,based on the measured data of coastal full-scale low-rise buildings and super high-rise buildings which have experienced many typhoons,the general situation of the measured data and the situation after experiencing typhoons are briefly introduced,and the processing method and data selection principle for the dead points of the data are introduced.This paper demonstrates in detail the derivation process of the crest factor of the Gaussian process,including the number and expectation of the random process exceeding the limit,and the rate of Gaussian process exceeding the limit.It introduces the derivation process of the Hermite moment model,briefly describes several calculation methods of nonGaussian crest factor based on the zero up-crossing theory,covering softening and hardening responses,and deduces the application scope of the model through the monotonicity premise of Hermite model.This paper makes a statistical analysis of the non-Gaussian characteristics and measured crest factors for two types of buildings.It is shown that there is large difference of non-Gaussian characteristics between the two types of buildings.The skewness and kurtosis of high-rise buildings are concentrated in the area of weak softening response,while the low-rise buildings are mostly concentrated in the area of high skewness and high kurtosis.The distribution of the measured crest factor shows obvious regularity.When it is in the strong softening area of negative skewness,the minimum crest factor is larger;when it is in the strong softening area of positive skewness,the maximum crest factor is larger.The area with larger measured crest factor is mainly affected by kurtosis,while the area with smaller measured crest factor is mainly affected by the absolute value of skewness.The value of the measured crest factor accounts for the largest proportion in the range of 3 to 6,and the value of highrise buildings is very concentrated.In this paper,the application ratio of various calculation methods in the measured samples and the error rate between the calculated crest factor and the measured crest factor are combined to give suggestions on the use of the method.The results indicate that the Sadek-Simiu method is more suitable for the softening response samples of high-rise buildings,while the modified crest factor method is recommended for the softening response of low-rise buildings,and the hardening crest factor method is better for the hardening response samples.