复合材料B基准值统计的多环境样本合并方法
Multiple condition samples merging method for B-basis value statistics of composites

为缩减获得复合材料设计许用值的试验成本,首先,介绍了复合材料手册-17-G最新推荐的一种基于回归分析的合并多环境样本的B基准值统计方法,重点阐述了在批间变异检验、环境样本间的方差等同性检验及正态性检验中加入工程经验判断的合理性及适用条件;然后,根据该多环境样本合并方法,设计开发了B基准值统计分析软件;最后,采用该软件对江苏恒神股份有限公司部分复合材料的力学性能数据进行了统计分析。结果表明:与传统的单环境样本统计程序相比,该多环境样本合并方法不仅显著提高了B基准值,而且克服了传统统计程序中对样本数量及批次数的限制。所得结论表明该方法具有较高的可操作性,适用于在缩减样本的情况下统计分析复合材料B基准值。 In order to reduce the test cost for obtaining the design allowables of composites, a regression analysis-based B-basis value statistical method merging multiple condition samples which was latest issued by Composite Material Handbook-17-G was introduced, and the rationality and applicable conditions for the addition of engineering experience judgment in the check of variation between batches, the equality check for square deviation between condition samples and the normality check were expounded emphatically. Then, the software for statistical analysis of B-basis value was developed based on the merging method of multiple condition samples. Finally, the software was used to statistically analyze the mechanical properties data of partial Jiangsu Hengshen Company Limited composites. The results show that comparing with the traditional single condition sample statistic program, the multiple condition samples merging method not only raises the B-basis value significantly, but also overcomes the limitation for number of samples and number of batches in traditional statistic program. The conclusions obtained show that this method has higher operability, and is applicable to the statistical analyses for B-basis values of composites in the case of reduced sample. 国家自然科学基金(11302099);中国博士后科学基金(2013M541658);江苏省自然科学基金(BK20130784)