Abstract:In view of the non-Gaussian characteristics of local wind pressure on the typical low rise building with a slope angle of 18.4°, a wind tunnel test has thus been carried out under the landforms of A, B and C. A research has been carried out on the distribution characteristics of time history probability density of wind pressure coefficient at typical measuring points, followed by an analysis of the distribution laws of non-Gaussian area of wind pressure under different geomorphic categories and wind directions, with a discussion of the relationship between non-Gaussian characteristics of wind pressure and roof flow field. The results show that, compared with Gaussian distribution and lognormal distribution, GEV distribution and gamma distribution can better fit the probability distribution of wind pressure coefficient in the vulnerable area of the roof, but the fitting effect of long trailing tail area is undesirable with a large skewness and kurtosis; the type of landforms and wind direction angle have significant influence on the non-Gaussian distribution of wind pressure of the low building roof. At the oblique wind direction, the Gaussian wind pressure area is divided into three sections along the direction of incoming flow, along which the area moves with the change of terrain categories. The non-Gaussian features of wind pressure is positively related to the mutual correlation coefficient of wind pressure space.