Distribution Characteristics Of Wind Pressure On The Cladding Of Tall Building

Wind load is the major consideration in the load design of cladding and its connections. Most of thewind load designs of cladding are based on the Gaussian assumption. However, more and more studiespresent that the local pressure fluctuations in the separated and reattached region of bluff body are found tobe non-Gaussian. The Gaussian assumption will underestimate the wind load of cladding and cause thecladding to unsafety. It is a general way to estimate peak pressure by a fixed peak factor. In this way, it willeither underestimate or overestimate the wind load. Thus, it has both theoretical and practical significancesthat studying the probability distribution characteristics of structure to estimate efficiently the accuratewind loads on the cladding.Based on the wind tunnel experiment of the Guangzhou West Tower which is 432 meters high, thispaper investigates the pressure distributions on the structure. The influences of turbulence and surroundings,especially the unbuilt East Tower, on the pressure fluctuation and RMS of pressure are also investigated inthis paper. Wind pressures for cladding design are presented. The main contributions and results of thisstudy are summarized as follows.Firstly, atmospheric boundary layer is simulated in the wind tunnel. Aim at the shortcoming of deltaspire in turbulence simulation, a new shape of spire which has curve side is presented in this paper. Thenew type of spire solves the problem that turbulence is too low in the boundary layer, which has troubledthe wind engineering researchers for a long time. The boundary layer with the same velocity profile butdifferent turbulence profile is simulated for study the influence of turbulence on the Guangzhou WestTower and the other special structures.Secondly, the pressure distribution on the Guangzhou West Tower is studied. The results show that thesurroundings including the Guangzhou East Tower present shielding effects on the mean pressure, but havesignificant influence on the RMS of pressure. Compared with the isolated case, the surrounding buildingshave significant influences on the RMS of pressure on the leeward face. The maximum suction appears onthe small radius curved surface and the RMS of pressure has an increase of 15.4% due to the interference ofthe East Tower.Thirdly, considering the stochastic properties of the wind pressure fluctuations, a number of timehistories of the wind pressure fluctuations are needed to carry out statistical studies. The results show that strong non-Gaussian characteristics are found in the separated region of the structure. Then the Gammadistribution is applied to fit the pressures and estimate the peak pressures, and the results are compared withthe Gauss based results. The results reveal that the minimum peak pressure coefficient somewhere in theseparated region estimated by using Gamma distribution assumption is -1.96, but only -1.46 is found ifusing Gaussian assumption. This indicates that the Gaussian assumption will underestimate the wind loadsand cause the cladding to unsafety. The shortcoming of the Gamma distribution assumption is also foundand discussed. After analyzing the merits and flaws of the Gaussian distribution and the Gammadistribution, a more accurate and efficient method is proposed to estimate wind pressures for claddingdesign.