基于Monte Carlo模拟的概率疲劳S-N曲线修正  

Correction of probability fatigue S-N curves based on Monte Carlo Simulation

作　　者：王起梁[1] 赵永翔[2] 叶小芬[1] 

机构地区：[1]南车戚墅堰机车车辆工艺研究所有限公司,江苏常州213011 [2]西南交通大学牵引动力国家重点实验室,四川成都610031

出　　处：《机车车辆工艺》2015年第3期1-4,10,共5页Locomotive & Rolling Stock Technology

摘　　要：系统介绍了服从客观疲劳分散性规律、保证服役范围预测合理的概率疲劳sa-N曲线的科学原则与方法,包括保证高应力幅范围疲劳寿命均方差为正、科学确定概率疲劳sa-N曲线的极大似然法,以及把概率疲劳sa-N曲线合理延续到超长寿命范围的＂协同概率外推法＂和不完善疲劳可靠性数据的重构法。根据收集到的42CrMo合金结构钢旋转弯曲疲劳极限数据,应用Monte Carlo模拟法,同时考虑置信度和超长寿命范围,且控制样本数在5－15范围,从而实现了试验数据与统计参数的高精度重构还原。Following the objective fatigue dispersion rule, principles and methods of determining probability fatigue sa - N curve by means of ensuring the prediction rationality of service range, including a positive mean variance of fatigue life within high stress amplitude scope and the maximum likelihood method for proper determination of probability fatigue sa -N curve, ＂concurrent probability extrapolation method＂ and the reconstruction method for incomplete fatigue reliability data to properly extend sa - N curve to extra -long life span are presented based on the collected spiral bending fatigue limit data of42CrMo structural alloy steel and Monte Carlo simulation. In consideration of the confidence coefficient and extra - long life span, the number of sample is limited to 5 - 15, so that the highly precise reconstruction and reduction of test data and statistical parameters are realized.

关 键 词：疲劳 S-N曲线 MONTE CARLO模拟 

分 类 号：O346.2[理学—固体力学]