文献类型：期刊文献

中文题名：基于正态随机样本确定结构用木质材料强度标准值的方法

英文题名：Determination method of characteristic strength for structural wood materials based on normal distribution and random sampling

作者：钟永[1,2] 武国芳[1] 任海青[1] 孙正军[2] 江泽慧[2]

第一作者：钟永

通信作者：Jiang, Zehui

机构：[1]中国林业科学研究院木材工业研究所,北京100091;[2]国际竹藤中心,北京100102

年份：2018

卷号：39

期号：11

起止页码：129-138

中文期刊名：建筑结构学报

外文期刊名：Journal of Building Structures

收录：CSTPCD;;EI(收录号：20190306397473);Scopus(收录号：2-s2.0-85060005541);北大核心:【北大核心2017】；CSCD:【CSCD2017_2018】；

基金：“十三五”国家重点研发计划(2016YFD0600904,2016YFC0701505);“十二五”国家科技支撑计划(2015BAD14B0502)

语种：中文

中文关键词：结构用木质材料;置信度水平;抽样样本;正态分布;参数法;非参数法;强度标准值

外文关键词：structural wood materials;confidence level;sample;normal;parametric method;non-parametric method;characteristic strength value

分类号：TU366.2

摘要：强度标准值是结构用木质材料能否应用于建筑结构中的重要力学性能指标之一,应结合置信度水平要求、抽样样本的试样数来确定。GB 50005—2003《木结构设计规范》中尚未建立结构用木质材料强度标准值的取值方法和预估计算公式,在计算强度标准值时往往直接采用国外标准中的强度标准值取样系数或强度标准值取样顺序数。为了建立和完善我国结构用木质材料强度标准值的确定方法,合理确定其强度标准值指标,以正态分布生成的随机抽样样本为算例,分别采用参数法和非参数法,开展了不同样本强度变异系数情况下的置信度统计分析。基于参数法分析结果,确定了强度标准值取值系数与抽样样本试样数、置信度的关系;基于非参数法分析结果,确定了强度标准值取样顺序数与抽样样本试样数、置信度的关系。基于回归分析结果,分别提出了强度标准值取值系数、强度标准值取样顺序数的拟合公式,算例分析表明拟合公式计算结果与实际结果吻合良好。
Characteristic strength is a key mechanical property index for structural wood materials used in building structures, which should be determined according to requirements of confidence level and number of specimens contained in sample. Due to lack of determination method and predicted calculation formula for characteristic strength in current GB 50005-2003 ‘Chinese code for design of timber structures＇ , the coefficient or the order statistic value of characteristic strength is usually directly copied from foreign standard. In order to develop and improve the determination method for characteristic strength and to reasonably determine the characteristic strength, the confidence values under different coefficients of variation were analyzed by parametric and non-parametric methods in which the normal random sample was selected. Based on the analysis results by parametric method, relationships between the coefficient of characteristic strength and confidence level, number of specimens was obtained; while, based on that by non-parametric method, relationships between the order statistic value, confidence level and number of specimens were also obtained. The fitting formulas of the coefficient of characteristic strength and of the order statistic value were constructed. The analysis results of an example indicate that the calculation results by the fitting formulas are in good agreement with the actual results.