Abstract
High load capacity of carburized gears originates mainly from the hardened layer and induced residual stress. On the other hand, surface decarburization, which causes a nonmartensitic layer, and inclusions such as oxides and segregation act as latent defects which considerably reduce fatigue strength. In this connection, the authors have proposed a formula of strength evaluation by separately quantifying defect influence. However, the principal defect which limits strength of gears with several different defects remains unclarified. This study presents a method of inferential identification of principal defects based on test results of carburized gears made of SCM420 clean steel, gears with both an artificial notch and nonmartensitic layer at the tooth fillet, and so forth. It clarifies practical uses of presented methods, and strength of carburized gears can be evaluated by focusing on principal defect size.
