Abstract

Low-cycle fatigue (LCF) plays an important role in the evaluation of steel structures subjected to cyclic large stresses, such as the reactor pressure vessel. In LCF, fracture is caused by the propagation of cracks generated by large deformations. Therefore, it is important to predict how the cracks will grow and when the structures are broken. However, the conventional method for predicting the number of cycles of fracture (Nf) in LCF is incomplete. This is because it is not revealed whether the conventional counting method of the number of random amplitudes, Rain Flow Method, can be applied after crack initiated because the crack closure occurs. This means that the invalid cycles which do not contribute to the crack propagation due to the crack closure are counted by using Rain Flow Method. In other words, it is required to confirm whether Rain Flow Method is applicable to the cracked material and make a mechanism for precisely predicting the life in random LCF amplitudes by removing the crack closure.

In this paper, the validation of Rain Flow Method was investigated with following two steps. First, the relationship of the ΔVg, amp-Nf were obtained. Second, the prediction of the life in random LCF amplitudes was employed through ΔVg, amp-Nf, Rain Flow Method and Linear Cumulative Damage Rule (LCDR). In addition, the investigation of how to remove the crack closure was conducted by using the point of the change of slope of P-Vg, the analysis of dadN-Δδeff, and FE model.

This content is only available via PDF.
You do not currently have access to this content.