Low-cycle fatigue life modeling of similar and dissimilar carbon steel under rotary friction welding effect using adaptive neuro-fuzzy inference system (ANFIS)
摘要
Rotary Friction Welding (RFW) is pivotal in manufacturing components for the automotive and marine sectors. Despite its importance, the low-cycle fatigue life of dissimilar carbon steel joints, specifically C35 and C45, is not well-documented. This paper reports the effects of RFW parameters on the fatigue life of these steels through both experimental trials and advanced modeling techniques. Utilizing axial low-cycle fatigue tests on base and welded metals under various friction pressures, our findings reveal a clear relationship between increased friction pressure and enhanced fatigue strength coefficient. Unlike conventional studies, this research employs the Coffin-Manson equation alongside an Adaptive Neuro-Fuzzy Inference System (ANFIS) to refine fatigue life predictions significantly. This innovative approach not only elevates prediction accuracy but also offers a novel integration of empirical and ANFIS-based modeling. Our results confirm that higher friction pressures (25–35 MPa) substantially boost both fatigue life and weld strength. This research pioneers advanced modelling techniques in RFW and proposes new practical guidelines for optimizing welding conditions, thereby enhancing joint durability and reducing fatigue testing costs in industrial settings.