Search

A quenched and tempered and a carbonitrided Fe-Cu-C steel were compared in fully reversed plane bending fatigue with notched and unnotched specimens at different densities. If the notches are characterized by elastic stress concentration factors Kt, the volume around the peak stresses diminishes with increasing Kt. Applying Weibull's survival statistics, the effect of highly-loaded volume on fatigue is quanti- fied to predict the fatigue performance of components. There are competing concepts with the same claim. For the relative stress gradient concept, it is shown that there is a loose, but systematic relationship between relative stress gradient and highly-loaded volume. With regard to the different heat treatments, the carbonitrided steel outperformed the quenched and tempered version up to the sharpest notches, although after carbonitriding all specimens were through hardened and free of compressive residual stresses. [ABSTRACT FROM AUTHOR]

The tooth-root load-bearing capacity of four different surface-densified powder metallurgical (PM) gears was successfully predicted based on the fatigue data of laboratory specimens. The influencing variables considered in the calculation were density, geometry, carbon content, and residual stresses. For separation of the variables of carbon content and residual stresses, two experimental series were investigated: one consisted of through hardening test specimens with various carbon contents, and a second involved case hardening. The casehardened samples had a carbon-content gradient in addition to a density gradient and surface residual stresses. Two different densification methods were applied: rolling and shot peening. The calculations were confirmed using pulsator testing results of the gears. Based on an applicable calculation, further insights are provided as to optimization of the density gradient and how to take residual stresses into account in such calculations. [ABSTRACT FROM AUTHOR]

Highlights: [•] An artificial neural network was utilized to predict endurance limits from carefully selected inputs. [•] The worst-case correlation factor of 0.9 indicated that the neural network has been well trained. [•] Comparison of predicted and experimental data confirmed the accuracy of the model. [ABSTRACT FROM AUTHOR]

From fatigue data for two high-to-medium-strength structural PM steels an approach has been developed to design Haigh diagrams and estimate the local endurance-stress amplitude at 107 cycles and 50% survival probability as a function of the stress ratio or the static mean stress. All stresses are considered local stresses KtδA and Ktδm where Kt is the stress concentration factor, δA is the nominal endurable stress amplitude, and δm is the nominal mean stress. The background is a functional relationship between the fatigue strength at a constant stress ratio R ≠ -1 and the fully reversed fatigue strength at R = -1. To limit experimental effort, the fatigue strengths of two PM steels were measured at four stress ratios: R = -5, R = -1, R = 0 and R = 0.5. The three relationships KtδA (R ≠ -1) = f [KtδA (R = -1)] are valid for notched and unnotched specimens of both steels. Thus, for R = -5, R = 0, and R = 0.5, the corresponding Haigh diagram can be derived. This approach should be applicable to real components and to other metallic materials. [ABSTRACT FROM AUTHOR]