基于损伤力学模型的概率疲劳寿命预测方法

The Prediction of Probability Fatigue Life Based on Damage Mechanics Model

基于金属结构危险点复杂应力状态下的疲劳损伤演化模型,利用材料的p-S^N曲线等对其部分损伤参数和材料参数进行概率化处理,提出了一种通过随机抽样计算危险点概率裂纹形成寿命的方法,讨论了损伤演化模型参数的基本概率特性以及相关性概念,采用拉丁超立方抽样方法实现了概率损伤参数的随机抽样,由此进行不同采样点的疲劳损伤累积,继而得到危险点的疲劳裂纹形成寿命。将损伤演化模型以UMAT子程序的形式嵌入ABAQUS软件系统中,给定结构模型及疲劳加载历程,即可用于复杂结构的疲劳寿命分析。结合铝合金2024-T3直耳片的疲劳试验,对其危险点的疲劳损伤及裂纹形成寿命概率分布进行了数值计算,初步表明了新算法模型的合理性。

A method of predicting probabilistic crack propagation life through random sampling is proposed,which is based on fatigue damage evolution model under complex stress state for metal structures. Some damage parameters and material parameters are regarded as random variables through p-S - N curve. The probability properties of damage parameters are discussed. Latin hypercube sampling is used to generate random samples of these parameters.Fatigue damage accumulation process is simulated for different samples to get crack initiation life. When structure and load are determined,fatigue life for complex structure can be obtained by embedding damage evolution equation in ABAQUS. The probabilistic properties of fatigue life for aluminum alloys 2024-T3 straight lugs are obtained through this method which coincide well with fatigue test results,which shows the validity of the method.