摘  要：根据试验要求，对叶片在样本较少情况下测定其疲劳极限σ-1和S-N曲线。通过振动疲劳试验的方法进行研究，试验采用升降法选择所施加振动应力的水平，采用异方差回归分析方法进行疲劳试验数据的分析。最后获得叶片材料的均值S-N曲线和置信度为95%、可靠度为99.9%的P-S-N曲线，并得到各自曲线上的疲劳极限σ-1。此方法保证试验的顺利进行并且具有较高的可靠性，对于小样本测定材料的S-N曲线具有较好的实用价值。

关键词：疲劳极限；小子样测试；S-N曲线；升降法；异方差回归分析方法

文献标志码：A       文章编号：1674-5124（2015）12-0124-04

Studies on small sample test method for blade S-N curve

ZHANG Bing1，2，　CHEN Liwei1， CHEN Yongxiang1， YAN Xudong2， ZHAO Shuaishuai2

（1. Beijing Institute of Structure and Environment Engineering，Beijing 100076，China；

2. Tianjin Aero-space Relia Technology Co.，Ltd.，Tianjin 300462，China）

Abstract: To determine the fatigue limit σ-1 and S-N curve of blades with few small samples as the test requirements. Vibration fatigue test， that is， the level of vibration stress was fixed via up-down method and the test data was analyzed through variance regression. The S-N curve of mean value and P-S-N curve with 95%-confidence coefficient and 99.9%-reliability of blade materials have been obtained， so as the fatigue limit σ-1 of each curve. The method ensures the smoothness and reliability of the test and has a guiding significance in determining the S-N curve of materials through small sample test.

Keywords: fatigue limit； small sample test； S-N curve； the up and down method； the variance regression analysis method