Abstract Design wind pressures for components and cladding (C&C) are key design parameters. For mid- and high-rise buildings, design pressures for C&C are often evaluated from wind tunnel tests, especially when building geometries are irregular or surrounding environments are complex. Wind tunnel data acquisition and analysis approaches both contribute to uncertainties of predicted design wind pressure coefficients. Although there is still debate on the most appropriate probability distribution to fit the peak pressure data and the best method to obtain the design pressures, Gumbel distribution and the generalized least square method (also known as BLUE) are widely considered and adopted in engineering practices. Wind tunnel test durations and number of peaks used in the extreme value analysis also influence peak pressure predictions. Some recent studies investigated the effect of these factors for predicting design wind pressures on low-rise buildings, which typically have a model scale of about 1:30 to 1:100. It is not clear whether observations in these studies are directly applicable to mid- and high-rise buildings or stadiums, which typically have a much smaller model scale (~1:400). For this purpose, a long duration wind tunnel test was conducted by employing a model with a generic high-rise building geometry. The desired design percentile of extreme pressures is briefly discussed. The effect of test durations and number of peaks involved in the extreme value analysis was investigated. Lastly, this study investigated the influence of inconsistency between the averaging time of design wind speeds and the target reference periods, in which the design pressure coefficients were expected to occur, on the prediction of design wind pressures. The results suggest that an adequate combination of test durations and number of peaks used in the extreme value analysis reduces uncertainties significantly and results in appropriate design wind pressures for engineering applications.

中文翻译：

---

考虑gumbel分布压力系数预测中高层建筑局部峰值风压的不确定性

摘要 组件和覆层 (C&C) 的设计风压是关键的设计参数。对于中高层建筑，C&C 的设计压力通常通过风洞试验来评估，特别是当建筑几何形状不规则或周围环境复杂时。风洞数据采集和分析方法都会导致预测设计风压系数的不确定性。尽管对于拟合峰值压力数据的最合适概率分布和获得设计压力的最佳方法仍存在争议，但 Gumbel 分布和广义最小二乘法（也称为 BLUE）在工程实践中被广泛考虑和采用。极值分析中使用的风洞试验持续时间和峰值数量也会影响峰值压力预测。最近的一些研究调查了这些因素对预测低层建筑设计风压的影响，低层建筑的模型比例通常约为 1:30 到 1:100。目前尚不清楚这些研究中的观察结果是否直接适用于中高层建筑或体育场，它们通常具有更小的模型比例 (~1:400)。为此，采用具有通用高层建筑几何形状的模型进行了长时间风洞试验。简要讨论了所需的极压设计百分位数。研究了测试持续时间和极值分析中涉及的峰数量的影响。最后，本研究调查了设计风速平均时间与目标参考周期不一致的影响，其中设计压力系数预计会出现，根据设计风压的预测。结果表明，在极值分析中使用的测试持续时间和峰值数量的充分组合可以显着降低不确定性，并为工程应用提供适当的设计风压。