PSI - Issue 53

Rainer Wagener et al. / Procedia Structural Integrity 53 (2024) 151–160 Author name / Structural Integrity Procedia 00 (2019) 000–000

157

7

At this point it must be concluded that the execution of a high-quality numerical fatigue approach for additively manufactured metallic structures is a very challenging task. On the one hand, the manufacturing technology leads to a very complex microstructure which may be predicted only in a statistical manner. On the other hand, it is still not possible to estimate the resulting types, positions and sizes of the defects before the manufacturing. Based on a good compromise and experiences it is possible to estimate different probabilities of occurrence of defects due to the exposure strategy depending on the part orientation. The conventional fatigue approaches require the knowledge of each relevant feature regarding its geometric dimensions, in order to estimate its influence on the fatigue. In case of additively manufactured structures, at least during the design state, when only the nominal geometry and the exposure strategy is known, is not possible to consider these relevant features in advance, since every part will at least slightly differ from the other one. 5. Fatigue approach based on Representative Structural Elements With respect to the different local fatigue related features within an additively manufactured metallic part, the strain-based local fatigue approach seems to be a good starting point to develop a fatigue approach based on Representative Structural Elements, introduced in Wagener et al. (2019) and specified in detail Wagener et al. (2020), and Wagener and Chiocca (2021). To describe the cyclic material behavior the so-called “cyclic stabilized” stress strain curve at the half of the fatigue life according to Ramberg and Osgood (1943), eq.1, and the strain-life curve, according to Basquin (1910), Coffin (1954), Manson (1965), Morrow (1965), eq. 2, are used, well-knowing that neither the stress-strain behavior is constant nor the strain-life curve is successfully validated for all metallic materials, as can easily be seen from the discussion by Endo and Morrow (1969), Sanders et al. (1977), Wong (1984), Wigant and Stephens (1987), Stephens and Koh (1998), Fatemi et al. (2005) and Wagener (2007). In the case of the stress strain curve more complex methods are available, but due to the assumptions of the cyclic stabilized stress-strain behavior and independency of the load-time history their benefit is questionable. , = , + , = + � ′ � ′ ( 1 ) , = , + , = ′ ∙ (2 ∙ ) + ′ ∙ (2 ∙ ) ( 2 ) To optimize the consideration of the impact of the load-time history on the stress-strain curve, it should be derived from the Incremental Step Test, which was introduced by Landgraf et al. (1969). Setting the number of cycles to failure equal to the damage sum D = 1 enables to describe the development of the stress-strain behavior depending on the damage progress by deriving the properties of stress-strain curve for each spectrum run, eq 3. , = , + , = + � ′ ( ) � ′ ( ) ( 3 ) Reducing the number of relevant properties to describe the fatigue behavior, requires a continuous Wöhler-line from the Low Cycle Fatigue (LCF) up to the Very High Cycle Fatigue (VHCF) regime to prevent possible changes of the fatigue approach during the design process in case that the load conditions change between LCF and HCF/VHCF loads. The Fatigue Life Curve by Wagener and Melz (2017, 2018), eq. 4, enables the combination of strain- and stress(force)-controlled test results. Following this, the SN-curve is derived under strain-control where plasticity dominates the fatigue behavior and under stress-control where macroscopic elastic material behavior occurs, because there should be no difference between stress-and strain-controlled fatigue test results in the elastic regime.