摘要
采用Hermite矩模型可将非高斯时程表示为高斯时程的非线性函数,建立了非高斯时程和高斯时程之间的一一对应关系,也建立了非高斯峰值因子和高斯峰值因子之间的一一对应关系,为非高斯峰值因子、极值的计算奠定了理论基础。介绍了软化时程、硬化时程和偏斜时程的Hermite矩模型变换理论,明确了高阶矩模型的单调变换范围;在此基础上,研究了软化时程非高斯峰值因子简化计算式的理论误差。结果表明由简化计算式得到的非高斯峰值因子略大,其误差均小于20%。利用非高斯峰值因子的简化计算式,计算了平屋盖表面典型测压点的非高斯峰值因子和风压极值。分析结果表明:绝大多数测压时程样本属于软化时程,极少数样本属于硬化时程或偏斜时程;利用非高斯峰值因子的简化计算式,需要考虑测压时程的随机特性,取多个时程样本峰值因子的平均值作为非高斯峰值因子的代表值。
The non-Gaussian wind history can be expressed as the non-linear function of a Gaussian time history by the Hermite moment models. The one-to-one mapping between the non-Gaussian and the Gaussian time histories is established while the one-to-one mapping between the non-Gaussian and the Gaussian peak factors is produced. The moment-based Hermite models provide a method to formulate the non-Gaussian peak factor and extreme wind pressure.The Hermite models of softening,hardening and skewed histories were introduced in this paper while the monotonic limits of high-order softening models were clarified numerically. The theoretical error of reduced formula for nonGaussian peak factor was investigated. It is indicated that the non-Gaussian peak factor resulted from the reduced formula is slightly larger than the theoretical result and that the relative error is less than 20%. Using the reduced formula of non-Gaussian peak factor,both the peak factors and extreme values of wind pressure on a flat roof were estimated. It is demonstrated that most samples of wind pressure histories belong to the softening histories and a few samples are the hardening or skewed histories and that the mean value of non-Gaussian peak factors is preferred when the randomness of measured samples are considered.
出处
《建筑结构学报》
EI
CAS
CSCD
北大核心
2015年第3期20-28,共9页
Journal of Building Structures
基金
国家自然科学基金项目(51378061)
教育部博士点基金项目(20120009110037)