| The southeast coasts of China are the typhoon-prone area,where a large number of houses are damaged or even collapsed due to typhoon landfall,bringing great economic and property losses to the local area.Among them,low-rise buildings share most of the losses,especially for light steel structure workshops,as they prevail due to its light weight,plate reusable and other advantages are widely used.For light steel structure workshop,doors,windows and roof slabs are easy to be destroyed in typhoon.After break of these components,rainwater intrusion will cause further losses,and they may also be the source of debris damage to adjacent buildings.Therefore,evaluation of vulnerability of doors,windows and roof slabs of light steel structure under wind load is of great significance.But most of the existing studies are just aimed at a specific closed or open state.The purpose of this paper is to make a consecutive analysis of failure of doors and windows by means of numerical simulation technology,and then to analyze the overall loss ratios of doors,windows,roofs and enclosures under different open conditions,and determining the wind induced vulnerability curve for this type of structure.The research content of this paper is as follows:The opening sequence and failure probability of doors and windows under wind loading are investigated.Firstly,the feasibility and reliability of the CFD(Computational Fluid Dynamic)numerical simulation technology in this paper are verified by three industry-recognized low-rise building model measured data,including the simulation of the external and the internal wind pressure coefficient.Then simulation of wind pressure of a light steel structure workshop under closed state with two wind directions is performed,and the mean pressure coefficients of difference zones of the wall are obtained.The failure probability of doors and windows under various mean wind speeds between 10-80 m/s is calculated by Monte Carlo simulation method when appropriate resistance models of window and door are adopted.If the failure occurs to door or window,it is then assumed that opening status is generated,and a building model with opening is re-established to obtain new average external wind pressures.The failure probability of the new model with opening is calculated for the remaining doors and windows,and this loop is repeated until all doors and windows are destroyed or no new opening positions are generated.The opening sequence and the internal pressure change of the building at each status under two wind directions and the failure probability of the doors and windows in the closed or each opening states are obtained.Finally,the failure probability for each failure states at various wind speeds is synthesized,and the comprehensive failure probability curve of doors and windows is obtained.The overall failure probability of roof panel and envelope of light steel structure factory is studied.Having determined the sequence of opening of doors and windows and the average internal and external wind pressure coefficient of the building associated with various damage states,the failure probability of the roof panel under different nominal wind speeds is calculated by Monte Carlo simulations by adopting a roof panel resistance model and failure criterion.Then,the loss rate and average loss rate matrix of doors,windows,roof and the envelope structure for different predefined damage levels are established,and according to the procedure for calculating the exceedance probability curve of each damage levels.The over-failure probability curves of doors and windows,roof panel and envelope structure as a whole are given for two wind directions under closures and each opening states.In addition,the overrun failure probability curves of the whole envelope structure are synthesized for different nominal wind speeds in each state.The overall loss ratio curve of the retaining structure is obtained by weighting the mean damage ratios of different components of the building.Some suggestions for improving the wind resistance of building are put forward. |