A Holistic Method for Fatigue Durability and Reliability Evaluations of Engineering Machinery and Structures

目前研究工程对象疲劳耐久性的损伤容限方法需要两阶段分析，即疲劳裂纹萌生和扩展寿命的估计。这种方法需要使用两种不同的理论和对疲劳裂纹起裂期的任意定义。该方法是根据2007-2012年资助项目开展的研究活动开发的，消除了一些不一致之处。然而，该方法的一个基本要素是材料性质Rho*，它被假定为裂纹尖端的有效半径或可以被认为是裂纹的最小材料缺陷的尺寸。目前，该材料常数是通过实验疲劳裂纹扩展数据间接确定的。作为2013-2018年探索基金（Discovery Grant）申请的一部分，该项目拟对选定材料进行系统的微观结构研究，目的是在不了解任何初步疲劳数据的情况下确定Rho*常数。我们自己的最新研究表明，Rho*常数可以根据断裂韧性、KIC和理论强度来评估。基于UniGrow模型和相关软件的最新疲劳裂纹扩展研究表明，仅通过疲劳裂纹扩展分析就可以估计出总疲劳寿命。因此，有必要建立一个与Rho*参数相当大小的裂纹疲劳扩展模拟模型。如果所提议的研究活动得到支持，将开发一种通用的疲劳分析方法。然而，当需要在估计工程对象的剩余疲劳寿命或应用损伤容限概念时进行可靠性和安全性决策时，确定性分析是不够的，因为所有输入数据实际上是分散的。因此，提出将确定性疲劳裂纹扩展模型与蒙特卡罗模拟方法相结合，对输入数据按规定的概率分布采样，进行多次疲劳裂纹扩展分析。如果分析成功，将为各种工程对象的安全疲劳寿命提供信息。

Contemporary damage tolerance methods concerned with fatigue durability of engineering objects require two-stage analysis, i.e. the estimation of fatigue crack initiation and propagation lives. Such approaches require using two different theories and an arbitrary definition of the fatigue crack initiation period. The methodology, developed as a result of research activities carried out within the current Grant 2007-2012, eliminates some of those inconsistencies. However, one of the basic elements of the approach is the material property Rho* assumed to be the effective radius of the crack tip or the size of the smallest material imperfection which could be consider as a crack. This material constant is at present determined indirectly from experimental fatigue crack growth data. It is proposed, as a part of the requested Discovery Grant 2013-2018, to undertake systematic microstructural studies of selected materials, with the aim of determining the Rho* constant without the knowledge of any preliminary fatigue data. Our own up-to-date studies have shown that the Rho* constant can be evaluated based on the fracture toughness, KIC, and the theoretical strength. The up-to-date studies of the fatigue crack growth based on the UniGrow model and associated software indicate that it is possible to estimate the total fatigue life based solely on the fatigue crack growth analysis. Therefore it is necessary to develop a model for simulating fatigue growth of cracks comparable in size with the Rho* parameter. If the proposed research activities are supported, a universal method of fatigue analysis will be developed. However, the deterministic analyses are not sufficient when reliability and safety decisions need to be carried out while estimating remaining fatigue lives of engineering objects or when applying the Damage Tolerance concept, because all input data are in reality scattered. Therefore, it is proposed to combine the deterministic Fatigue Crack Growth model with the Monte Carlo simulation method and to run multiple fatigue crack growth analyses based on the input data sampled according to their prescribed probability distributions. When successful the analysis would provide information on safe fatigue lives of various engineering objects.