论文部分内容阅读
疲劳裂纹数密度和裂纹扩展速率是描述材料疲劳损伤程度、预测材料疲劳寿命的两个重要参数。关于光滑试件表面疲劳裂纹数密度演化及疲劳裂纹扩展规律的定量化描述的研究 ,人们一直非常重视 ,业已提出了一些模型 ,但其适用范围均受到限制。因此 ,目前尚未有被广泛接受的裂纹演化定量模型。研究用 BP神经网络描述疲劳裂纹数密度演化、裂纹扩展速率演化规律 ,结果表明 :该方法合理可行 ,解决了传统描述方法确定长短裂纹分界点的困难 ,同时 ,也避免了显式模型中参数的物理意义不十分明确 ,实际应用时难以确定的问题。
The fatigue crack density and crack propagation rate are two important parameters to describe the degree of material fatigue damage and predict the fatigue life of the material. The research on the quantitative description of the evolution of the number of fatigue cracks and the fatigue crack growth on the smooth specimen has drawn great attention. Some models have been proposed, but their application scope is limited. Therefore, there is not yet a widely accepted quantitative model of crack evolution. The results show that this method is reasonable and feasible, which solves the difficulties of the traditional description method in determining the demarcation point of the long and short cracks. At the same time, it also avoids the problem that the parameters of the explicit model Physical meaning is not very clear, difficult to determine the actual application of the problem.