The effect of inclusion geometry according to forging ratio and metal flow direction on very high-cycle fatigue properties of steel bars

Abstract The objectives of the present paper are to clarify the effect of inclusion geometry according to the forging ratio and the metal flow direction on very high-cycle fatigue properties of steel bars and to propose an evaluation method for characteristic inclusion geometries observed at fracture origin. Specimens were made of steel plates rolled to three levels of thickness from steel bars of low alloy steel. Then the specimens were cut in both parallel and perpendicular directions to the metal flow of the steel plates. Fatigue tests were conducted in a long life regime by using an ultrasonic fatigue testing machine. The fatigue limit for 109 cycles decreased with the increase of the thickness of steel plate, i.e. a decrease of the forging ratio. The fatigue limit of specimens cut parallel to metal flow was higher than that of specimens cut perpendicularly to metal flow. Fish-eye fractures appeared on the fracture surface of most specimens. Especially, characteristic inclusions, which are clusters of oxides, sulfides and their combination, were mainly observed at the origin of fish-eye fractures in specimens cut perpendicularly from thick plates. In this paper, these characteristic inclusions were treated as the extended elliptical inclusion/crack from the point of view of fracture mechanics and the area parameter model. More careful inspection at the fracture origin revealed that the rough surface area called the optically dark area (ODA) appeared only around the edge of the short radius on the extended elliptical inclusion where the higher stress intensity factor is generated. This result suggests that the middle part of the extended elliptical inclusion dominates the formation of ODA and the very high-cycle fatigue property. Finally, based on the fracture mechanics approach, area L/W=4, which is the square root of the area of the surrounded part where length/width area parameter model.