PSI - Issue 68

J.A. Ziman et al. / Procedia Structural Integrity 68 (2025) 1159–1165 J.A. Ziman et al. / Structural Integrity Procedia 00 (2025) 000–000

1160

2

1. Introduction The relation between the load amplitude and the corresponding number of cycles to failure N f is usually given in terms of so-called S-N curves. As the generation of these curves involves a large amount of experimental testing effort according to DIN 50100, it is a time- as well as cost-consuming procedure. A reduction of the experimental effort can be achieved by the development of new lifetime prediction methods (LPM), which are able to describe the fatigue behaviour of metallic materials with a demonstrably reduced number of specimens. These methods combine non destructive testing (NDT) methods, digitalized measurement technology, and signal processing with conventional destructive testing to significantly increase the information content regarding fatigue properties, Starke (2019), Weber et al. (2023) and Teng et al. (2020). As a result, a process-oriented assessment of the induced damage instead of a lifetime-oriented analysis is provided. In particular, electrical resistance measurements, as well as temperature-based methods are suitable for a reliable in-situ characterization considering dynamic loading. Besides geometric changes (specimen cross-section and length), the electrical resistance depends on the electrical resistivity, which, according to Matthiessen (1865), belongs to temperature-dependent and temperature-independent portions, Wu et. al. (2020, 2023). Because of the increased energy input per time unit, an increased testing frequency leads to an increased energy dissipation in terms of heat. Therefore, it is necessary to consider temperature effects on the electrical resistivity via a temperature-correction for a reliable estimation of the fatigue behaviour. This calculated temperature-corrected electrical resistance can be used as an input variable in the LPM StressLife, which was originally developed for tests at 5 Hz, Starke (2019) and Weber et al. (2023), and is now being used and validated at test frequencies up to 260 Hz for the first time. 2. Methods and experimental setup 2.1. StressLife method As part of the investigations, the LPM StressLife is applied to fatigue specimens made from the ferritic-pearlitic steel SAE 1045 (C45E, 1.1191) in a normalised condition (austenitizing temperature T aus = 860 °C, austenitizing time t aus = 45 min) at higher test frequencies (80 and 260 Hz). The procedure of the LPM is shown schematically in Figure 1. At first, the stress amplitude σ a is given as a function of the average material response M per load step of a load increase test (LIT) in a Morrow-equivalent plot. Separating M into a predominantly elastic and mainly plastic portion enables the calculation of the weighted hardening exponent n total in order to take sufficient account of both the slope of the elastic and the plastic portion. The coefficients B and C as well as the exponents b and c can be calculated according to the equations (1) to (4) using M after half of the number of cycles to failure N f /2 of two constant amplitude tests (CAT). These parameters allow M to be separated according to equation (5) into a predominantly elastic (according to Basquin (1910)) and predominantly plastic portion (according to Coffin (1954) and Manson (1953)). In this context B is representing the fatigue strength coefficient, b the fatigue strength exponent, C the fatigue ductility coefficient and c the fatigue ductility exponent. As a result, M of a further CAT can be estimated. An allometric regression of all CAT data points provides the remaining input variables K’ representing the cyclic hardening coefficient and n’ in terms of the cyclic hardening exponent. Based on these input variables, a StressLife-curve can be calculated by using the equation (6), Starke (2019) and Weber et al. (2023). (1)

(2)

(3)

(4)