PhyBaLSL – Short-time procedure for the determination of the fatigue lifetime of metallic materials under service loading

Abstract

Because components are usually not loaded with periodical waveforms and constant amplitudes, but with temporally variable, cyclic loadings, fatigue testing under near service loading conditions is essential to optimize the component’s design. However, this kind of fatigue tests requires high efforts in material, time and costs, which can be significantly decreased by using efficient short-time procedures. In the present work, a modification of the physically based lifetime calculation method (PhyBaLLIT) for service loading conditions, i.e., PhyBaLSL, was used to estimate the Gassner curve of ductile cast iron EN-GJV-400 resulting from cyclic loading based on the Car Loading Standard (CARLOS). For fatigue tests, the first 5 000 cycles of CARLOS were used for periodically repeated loading intervals (LIs) with defined maximum stress σmax, CARLOS. In the presented work single step tests (SSTs) with constant σmax, CARLOS as well as load increase tests (LITs), in which σmax, CARLOS was stepwise increased after each LI, were performed. Because for PhyBaLSL the determination of the material’s cyclic deformation behavior is a prerequisite, measuring intervals (MI) with constant stress amplitude σa, MI were performed after each LI. The influence of MIs on the fatigue lifetime and cyclic deformation behavior was analyzed by using different σa, MI, showing no significant influence but higher resolution of fatigue-induced changes with increasing σa, MI. Based on the measurements obtained in the LIT and two SSTs, the Gassner curve could be determined with PhyBaLSL showing a good correlation to the lifetime data obtained from additionally conducted SSTs. Furthermore, the results from LITs enable a rough estimation of the expected fatigue strength at 2 × 106 cycles. In addition to the results obtained from cyclic deformation behavior determined in MI, the change in electrical resistance ΔR was measured in the specimens’ gauge length during the fatigue tests. This measurement does not require a constant stress amplitude σa and hence, can be used to characterize the fatigue behavior without using MIs. The presented results demonstrate that ΔR enables an adequate estimation of the Gassner curve by using PhyBaLSL.