摘要
完成了铁道车辆 LZ5 0车轴钢的概率机械性能试验研究 ,拓宽了确定有限疲劳可靠性数据良好假设分布的统一方法 ,比较了 6种常用分布 (即三参数 Weibull、两参数 Weibull、正态、对数正态、极大值和极小值分布 )对试验数据的描述效果 ,综合分析了他们的拟合优度、失效机理的一致性和尾部预测的安全性。从预测的安全性角度 ,选择了极小值分布为最佳统计模型 ,提出了给定可靠度和置信度下基于极小值分布的机械性能参数估计方法 ,并有效地估计了材料的概率机械性能参数。
Experimental study on the probabilistic mechanical properties was performed on the LZ50 axle steel for railway vehicles.The unified approach for determining appropriate statistical models of the limited fatigue reliability data was developed to analyze the present test data.The six commonly used distributions,namely three parameter Weibull,two parameter Weibull,normal,lognormal,extreme maximum value,and extreme minimum value,were compared in fits of the data from the three considerations of goodness-of-fit,consistency of failure mechanism and safety in tail region predictions.From the viewpoint of safety,the extreme minimum value distribution was selected as an optimal model of the data.A method based on this distribution was proposed for the estimation of mechanical properties under given reliability and confidence levels.The method has been well used for the estimation of probabilistic parameters of the mechanical properties of the present material. 9 tabs,2 figs,20 refs.
出处
《交通运输工程学报》
EI
CSCD
2003年第2期11-17,共7页
Journal of Traffic and Transportation Engineering
基金
国家自然科学基金项目 (50 0 750 73)
全国优秀博士学位论文作者专项资金项目 (2 0 0 2 0 67)
铁道部科技研究开发计划项目 (2 0 0 1 J0 1 6)
关键词
铁道车辆
车轴
LZ50钢
机械性能
可靠性
对数正态分布
railway vehicle
axle
LZ50 steel
mechanical properties
reliability
lognormal distribution