PSI - Issue 68

Robert Szlosarek et al. / Procedia Structural Integrity 68 (2025) 1173–1180 Robert Szlosarek et al. / Structural Integrity Procedia 00 (2025) 000–000 and strains to the cyclic stress-strain curve. The equation consists of the cyclic strength coefficient ′ and of the cyclic strain hardening exponent ′ , both being material-related parameters. Strain-based fatigue tests are used to determine these parameters by analyzing the stress-strain hysteresis loop of a dataset of tests at half of the lifetime. Performing the evaluation at half of the lifetime is established due to the assumption that the parameters are more or less constant, which means cyclic saturation occurs. The same assumption is quite common in performing the fatigue analysis. However, nonconstant cyclic deformation curves are observed in certain materials. This can happen due to a softening or hardening of the material. Hence, it is questionable to determine the parameters at the half of the lifetime as well as assuming the parameters as constant for the fatigue analysis. This leads to inaccurate and sometimes disappointing computational outcomes. To overcome this issue the study presents a solution to consider nonconstant cyclic behavior in the fatigue analysis under constant total strain amplitudes. Möller et al. (2017) show a method how to take into account a cyclic softening or hardening by evaluating the reversal points of the hysteresis loop and the slope of the corresponding connection line. Korschinsky et al. (2024) present experimental data to the softening of EN AW-1050A H24 aluminum alloy. They recommend to use a damage dependent material model to describe the transient behavior of the material. In Möller (2020) a method is described to use linear functions of the material parameters to consider the cyclic softening of the material. In addition, a computational framework is given to implement these functions in the lifetime calculation. The present study was done with the material data of an experimental metastable, austenitic X2CrNiCuN17-6-4 steel as it is introduced by Hauser et al. (2023). This material undergoes a phase transformation during cyclic loading. Thus, the martensite content increases with each cycle. In Nitzsche et al. (2024) it is shown that the phase transformation leads to a massive cyclic hardening of the material. For example, the test with the highest total strain amplitude of 0.8 % shows a stress amplitude of about 550 N/mm² at the beginning and more than 1100 N/mm² just before the fatigue crack arises. In order to consider this hardening behavior, it is mandatory to use transient material parameters. In detail, a method will be presented to describe the values of ′ and ′ in dependency of the damage with the aim of representing the stress amplitude in the calculation quite equal to the experiments. Nomenclature fatigue strength coefficient & , fatigue ductility coefficient fatigue strength exponent fatigue ductility exponent ′ cyclic strength coefficient ′ cyclic strain hardening exponent damage value 2. Methods Within research the damage model of Smith, Watson and Topper (1970), the so called PSWT damage model, is used to perform lifetime calculation. The PSWT damage model is based on the energy density '()* = , !"# ∙ $," ∙ . (1) 2 !"# maximum stress " stress amplitude ! mean stress $," total strain amplitude Youngs modulus number of cycles & load cycles to failure & ,

1174