Modified Low-Cycle Fatigue Estimation Using Machine Learning for Radius-Cut Coke-Shaped Metallic Damper Subjected to Cyclic Loading

In this study, a coke-shaped steel damper that exhibits in-plane resistance is introduced as a passive damper. The double-coke damper presented in this study applies the concept of reduced beam sections to increase the ductility in the case of a prolonged earthquake. Multiplastic hinges are placed on each strip by setting the radius-cut section. The fatigue performance of the damper during earthquake loading is verified through a constant cyclic loading test. The results indicate that, as the number of plastic hinges inside the strip increases, the damper ductility increases, producing a stable hysteresis graph. In addition, a new equation that considers the damage index using parameters such as maximum strength and effective stiffness is proposed, and the experimental results are found to be in excellent agreement with the number of failure cycles obtained from the proposed model. By comparing the results of applying the proposed equation with the machine learning results, it is demonstrated that machine learning can be used for estimating the damper performance against the fatigue of the resistive cycle.