부산대학교 도서관

Since in superlong fatigue failure with Nf≥108, the average fatigue crack growth rate is much less than lattice spacing (-0.1Å or 0.01nm or 10-11--12m/cycle), we cannot assume that crack growth occurs cycle by cycle in the early stage of fatigue process.In this paper, possible mechanisms for extremely high cycle fatigue are discussed. A special attention was paid to a newly found particular fatigue fracture morphology in the vicinity of fracture origin (nonmetallic inclusions) of a heat treated alloy steel, SCM435. The particular morphology looks a dark area inside fish-eye mark by optical microscopic observation. Specimens with short fatigue life of N=-105 do not have such dark area in fish-eye mark. SEM and AFM observations revealed that the dark area has a rough surface quite different from usual fatigue fracture surface in martensite lath structure. The predictions of fatigue limit by the √area parameter model are -10% unconservative for fatigue limit defined for N=107. Thus, the fatigue failure for N≥108 is presumed to be caused by a mechanism which induces breaking or releasing of fatigue crack closure phenomenon in small cracks. The breaking or releasing crack closure mechanism is presumed to be caused by environmental effects such as hydrogen embrittlement coupled with extremely high cycle fatigue. Some indirect evidences to support this hypothesis are shown.