风压系数特性对风荷载超越概率和结构失效概率的影响OA
Effects of Wind Pressure Coefficient Characteristics on Wind Load Exceedance Probability and Failure Probability of Structure
为明确风压系数样本大小、参数估计方法、概率分布类型等对风荷载超越概率和结构失效概率的影响规律,本文定量分析了风压系数特征变量对结构风荷载超越概率和结构失效概率的影响,采用蒙特卡洛方法和可靠度校准方法开展了风荷载超越概率和结构失效概率估计研究.结果表明:低重现期设计风荷载较高重现期设计风荷载而言,对风压系数样本量敏感性更低,低重现期和高重现期设计风荷载超越概率的平均值和标准差都会随着风压系数样本量的增加而逐渐变小;低重现期时,在57%与78%两种特殊分位值下的设计风荷载超越概率的平均值和标准差都大于随机分位值下的结果,在小样本时,建议风压系数采用78%分位值计算设计风荷载;在样本量足够的情况下,依据Gumbel分布拟合风压系数来进行结构设计会更加安全;采用L矩法进行风压系数概率分布参数估计,在小样本情况下对于Gumbel分布和广义极值(GEV)分布均具有较快的超越概率收敛能力;与风速相比,风压系数的样本量对风荷载超越概率影响更小,建议风洞试验采样及相关极值统计的风压系数样本量不小于10;在相同样本量下,风荷载与恒荷载的比值越大,结构失效概率平均值和标准差越高,结构越趋于危险;在确定的风荷载与恒荷载比值下,采用极大似然法(MML)且样本量大于10时,对结构失效概率和可靠度的估计结果偏差较小.研究成果可为风压系数的采样设置、概率统计拟合以及设计风荷载的可靠度校准提供参考.
The influence of sample size,parameter estimation methods,probability distribution types on exceedance probability of design wind load,and structural failure probability is not yet clear.For evaluating the effects of wind pressure coefficient characteristics on the exceedance probability and structural failure probability,the Monte Carlo method and reliability calibration method are used to estimate these probabilities.The results indicate that low return period design wind loads are less sensitive to the sample size of wind pressure coefficients compared to those with high return periods.Both the mean and standard deviation of the exceedance probability decrease gradually as the sample size increases.During the low return period,the mean and standard deviation of the design wind load exceedance probability at two special percentile values of 57%and 78%are greater than those at random percentile values.For small samples,it is recommended to use the 78%percentile value of the wind pressure coefficient to calculate the design wind load.When the sample size is sufficient,fitting the wind pressure coefficient based on the Gumbel distribution is safer for structural design.Applying the L-moment method to estimate the probability distribution parameters of wind pressure coefficients has a rapid convergence ability for both Gumbel and GEV distributions in small sample sizes.Compared with wind speed,the sample size of wind pressure coefficient has a smaller impact on wind load exceedance probability.It is recommended that the sample size for wind tunnel test and related extreme value statistical wind pressure coefficient should not be less than 10.Under the same sample size,the larger the wind-to-dead load ratio,the higher the mean and standard deviation of structural failure probability,making the structure more dangerous.Under a specific wind-to-dead load ratio,the maximum likelihood estimation method is used,and the sample size greater than 10 has little impact on the estimation of structural failure probability and reliability.The research results can provide a reference for wind pressure coefficient sampling,probability statistical fitting,and reliability calibration of design wind load.
李寿科;苏培林;李友新;杨泽佳;邓慧;陈宁;刘敏
湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201湖南科技大学 结构抗风与振动控制湖南省重点实验室,湖南 湘潭 411201
建筑与水利
风压系数概率分布样本量风荷载超越概率结构失效概率可靠度
wind pressure coefficientprobability distributionsample sizewind loadexceedance probabilitystructural failure probabilityreliability
《建筑钢结构进展》 2026 (4)
30-37,8
国家自然科学基金(52378510)
评论