Unexpected near-threshold fatigue crack growth and criterion to predict it by mesoscopic- modeling

在本研究中，采用K_<max>常数法对几种材料进行无闭合条件下的试验，观察到阈值应力强度因子（SIF）下降范围ΔK_< >，这是由于SIF最大值K较高。我们提出了一个简化的模型来解释这一现象，考虑到在断口表面观察到静态断裂的痕迹。同时，根据以往Al和Ti的数据，提出了预测这一现象的评价准则，并通过实验证明该准则对钢也是有效的。考虑到在实验室空气环境中进行的SLC（持续载荷开裂）试验的数据，我们还验证了环境对感兴趣的现象的影响很小。然后，我们提出了一种定量估计由于最大SIF高而导致阈值SIF范围减小的方法。根据碳钢、铝合金和钛合金的实验数据，利用遗传算法反求模拟参数。将候选参数集命名为“generation”，反复生成和评估，直到模拟再现实验数据。非常有趣的是，虽然这三种材料的测试条件不同，但得到的模拟参数似乎与材料无关。通过将预测值与实验测定的脆性碳钢阈值降低值进行比较，验证了该方法的有效性。最后，我们总结了我们对感兴趣的现象的评价标准，作为工程应用的评价图。

In this research, the decrease in threshold stress intensity factor (SIF) range ΔK_<th> due to high maximum SIF K, that is observed for several materials in the tests with the K_<max> constant method when closure-free conditions are realized. We proposed a simplified model to explain the phenomenon, considering the fact that marks of static fracture are observed on the fracture surface. At the same time, we proposed an assessment criterion to predict this phenomenon from the past data for Al and Ti and showed that the criterion is valid also for steels by experiments. We also verified that the environmental effect on the phenomenon in interest is small, considering the data of the SLC (Sustained Load Cracking) tests conducted in laboratory air environment.Then, we proposed a method to quantitatively estimate the decrease in threshold SIF range due to high maximum SIF. The parameters for the simulation were inversely determined from the experimental data of carbon steel, Al and Ti alloys with the aid of Genetic Algorithm. The set of candidate parameters named as a 'generation' were repeatedly generated and evaluated until the experimental data were reproduced by the simulation. Very interesting was the fact that though the test conditions for these three materials were different, the obtained simulation parameters seemed to be independent of material. The proposed method was validated by comparing predictions with experimentally determined values of the decrease in threshold on embrittled carbon steel. Finally, we summarized our assessment criterion for the phenomenon in interest as an assessment diagram for engineering application.