PSI - Issue 68

Rita Dantas et al. / Procedia Structural Integrity 68 (2025) 901–907

903

Rita Dantas / Structural Integrity Procedia 00 (2024) 000–000

3

Thus, it is commonly observed a direct relation between the sensitivity of a material to strain rate and the frequency e ff ect present in fatigue results (Krupp and Giertler , 2022; Hong et al , 2023). The sensitivity of materials to strain rate and, consequently, to frequency e ff ect is highly influenced by the lattice crystal structure. Certain structures seem to be more prone to frequency e ff ect than others, depending on intrinsic characteristics such as the number of slip planes, planar density and interatomic spacing (Hong et al , 2023). This dependency is explained by the fact that plastic deformation at atomic level is a mechanism of dislocations movement. Thus, the resistance to plastic deformation is constrained by the capacity of dislocations to move along a crystal structure (Callister and Rethwisch , 2018; Hertzberg , 1996). However, the impact of lattice structure can be balanced and even vanished by the influence of ultimate strength. Materials with higher strength exhibit less plastic deformation, which enhances the resistance of dislocation move ments even at ultrasonic frequencies (Hong et al , 2023) . Moreover, Furuya et al (2019) states that frequency e ff ect also depends on the mechanism of failure. For the cases where internal crack initiation is dominant, which is verified in high strength steels, the frequency e ff ect is negligible. On the other hand, factors external to the material can also have an impact in fatigue data, such as the technology of testing or the specimens geometry. Regarding the first topic, it is important to notice that ultrasonic fatigue tests are performed under displacement control, while more conventional technologies of testing to evaluate high-cycle fatigue (HCF) and VHCF regime are usually in load or stress control. As consequence, when the crack initiates, the displacement applied can in reality induce a nominal stress lower than the target value (Zimmermann , 2019). Moreover, the temperature increase due to the high frequency of testing can also influence the material’s fatigue behaviour. Nonetheless, this factor can be easily discard and even avoid due to the cooling system present in ultrasonic systems or to the pause / pulse loading configuration (Zimmermann , 2019; Mayer, 2016). Nevertheless, the ultrasonic fatigue testing is usually characterized by smaller specimens to ensure the range of amplitude stress desired to test. As consequence, the risk volume is smaller in ultrasonic specimens than in the con ventional specimens tested in resonance or bending machines. Thus, the probability of volume defects and consequent fatigue failure also reduces (Furuya et al , 2002; Mayer, 2016). Thedi ff erent factors that influence frequency e ff ect are summarized in Fig. 1 and sorted by material or test typology causes. The material causes are related between each other, since the strain rate sensibility is highly determined by lattice structure as well as material strength. Therefore, the frequency e ff ect can be more or less relevant for materials with di ff erent properties and characteristic, which leaded to the development of several works regarding this topic (Mayer, 2016). In case of mild and low carbon steels is observed a considerable influence of testing frequency in the experimental fatigue results (Guennec et al , 2014). On the other hand, the high strength steels (martensitic steels and martensitic stainless steels) and high alloys steels are generally not a ff ected by frequency e ff ect. This can be explained by the small plastic deformation present on these materials during fatigue tests (Furuya et al , 2002).

Fig. 1. Main causes of frequency e ff ect in experimental fatigue data

Regarding aluminium alloys, the frequency e ff ect is not so relevant, since they are not very sensitive to strain rate. Finally, there are some works about nickel based alloys and titanium alloys, but the conclusions are not very consistent and the frequency e ff ect in these materials require further studies (Mayer, 2016; Zimmermann , 2019).