摘要
基于ＬａｎｇｅｒＳＮ模型的ＰＳＮ曲线，提出了设计ＳＮ曲线的概率估计方法和ＡＳＭＥ法的可靠性评价方法。ＰＳＮ曲线由文中提出的广义极大似然法估计。方法可应用于处理具有双随机性特征的ＳＮ数据，并可推广应用于处理应力控制成组法疲劳试验和极大似然法试验得到的ＳＮ数据。对由１Ｃｒ１８Ｎｉ９Ｔｉ不锈钢管道结构焊接头低周疲劳试验得到的虚拟应力幅裂纹萌生寿命数据的研究表明，方法有效地估计了数据的统计特征。可靠度为０．９９９９的设计ＳＮ曲线估计结果，除试验数据的分散性外，已相当程度地考虑了其它不确定性因素的影响。ＡＳＭＥ法中，基于２０倍裂纹萌生寿命降低系数的ＳＮ曲线估计结果较基于２倍虚拟应力幅降低系数的估计结果要保守得多。后者在６６６．６１ＭＰａ虚拟应力幅的寿命估计结果相当于０．９９９５２２可靠度水平，在２０９２．１８ＭＰａ虚拟应力幅相当于可靠度水平。这说明在低虚拟应力幅可能给出危险的估计。反之，在高虚拟应力幅可能给出过余保守的估计。造成过余保守或危险估计结果的原因是现有ＡＳＭＥ法没有考虑裂纹萌生寿命分散性随加载水平降低而增大的一般规律。因此，有必要在设计ＳＮ曲线估计中引入概率方法。
A probabilistic evaluating approach of design S N curve and a reliability assessment approach of the ASME code based evaluation are presented on the basis of Langer S N model based P S N curves.The P S N curves are estimated by a so called general maximum likelihood method in this paper.This method can be applied to deal with the virtual stress amplitude crack initial life data which have a characteristics of double random variables.Investigation of a set of the virtual stress amplitude crack initial life (S N) data of 1Cr18Ni9Ti austenitic stainless steel welded joint reveals that the P S N curves can give a good prediction of scatter regularity of the S N data.Probabilistic evaluation of the design S N curve with 0.9999 survival probability has considered various uncertainties,besides of the scatter of the S N data,to an appropriate extent.The ASME code_based evaluation with 20 reduction factor on the mean life is much more conservative than that with 2 reduction factor on the stress amplitude.Evaluation of the latter in 666.61MPa virtual stress amplitude is equivalent to 0.999522 survival probability and in 2092.18MPa virtual stress amplitude equivalent to survival probability.This means that the evaluation in the low loading level may be non conservative and in contrast,too conservative in the high loading level .Cause is that the reduction factors are constants and the factors can not take into account the general observation that scatter of the N data increases with the loading level decreasing.This has indicated that it is necessary to apply the probabilistic approach to the evaluation of design S N curve.
出处
《核动力工程》
EI
CAS
CSCD
北大核心
1999年第3期236-243,共8页
Nuclear Power Engineering
关键词
设计S-NF曲线
概率估计
疲劳
ASME法
可靠性评价
Design S N curve Probabilistic evaluation ASME code Reliability assessment P S N curve General maximum likelihood method
