Abstract

This article investigates the dependence between the method of statistical data processing and the necessary amount of experimental studies required for empirical determination with a given level of reliability of the impact surface dent visual detectability threshold in composite aeronautical structures. Two methods for determining the minimum detectable damage size are considered, one of which is based on using a binomial distribution, and the other, on bootstrap modeling. A comparative analysis of the effectiveness of the two methods is carried out. As a result, it is shown that the bootstrap method gives significantly less scatter of the estimates and allows reducing the number of experts involved by more than by half with the same accuracy of the estimates.

中文翻译：

---

确定建立复合航空结构中几乎可见的冲击破坏尺寸所需的最少检查次数

摘要

本文研究了统计数据处理方法与经验确定所需的实验研究之间的依赖关系，该研究需要以复合航空结构中撞击表面凹痕视觉可检测性阈值的给定可靠性水平进行经验确定。考虑了两种确定最小可检测破坏大小的方法，一种基于使用二项式分布，另一种基于自举模型。对这两种方法的有效性进行了比较分析。结果表明，自举方法可以大大减少估计的分散性，并且在估计精度相同的情况下，可以将涉及的专家人数减少一半以上。