PSI - Issue 71

Shohei Matsuda et al. / Procedia Structural Integrity 71 (2025) 4–9

5

economical producing of products with complex shapes, and ease of recycling. However, casting defects, such as gas and shrinkage pores, oxide films, and eutectic Si particles, are known to reduce the fatigue strength, causing large scatter. Therefore, to establish a fatigue limit prediction method applicable to a broad range type of cast aluminum alloys, it is necessary to understand the mechanism of fatigue strength determination by investigating the behaviors of the cracks generated from natural defects such as casting defects. Wang et al. reported that natural defects, such as pores and oxide films, observed at the fracture origin shorten the fatigue life and that the pores are more harmful. They also elucidated that the effect of the features of matrix structure, such as secondary dendrite arm spacing (SDAS), on fatigue strength becomes more pronounced in HIP treated defect-free specimens (Wang et al. 2001, Part I & Part II). Some studies employ the √ as a geometric parameter of defects to predict the fatigue limit (Ueno et al., 2012; Tajiri et al., 2014; Roy et al., 2011). Ueno et al. and Tajiri et al. modified the √ parameter model so that the prediction fits the experimental results. Roy et al. performed tension-compression, reversed torsion, and combined axial-torsional fatigue tests by introducing artificial defects into the specimens by electro-discharge machining (EDM) and compared the experimental results with the previously proposed prediction methods for multiaxial fatigue limit. Cast aluminum alloys contain many natural defects of different types, shapes, and dimensions. In addition, the grain sizes are significantly larger than those of structural steels, reaching several millimeters (Tajiri et al., 2014; Garb et al., 2018; Serrano-Munoz et al., 2020). Hence, it is easily imagined that various statistical factors, regarding the shape and size of the defects that act as the fracture origins as well as the size and crystallographic orientation of the grains surrounding the defect, will contribute to the scatter in fatigue limit and fatigue life. Thus, statistical analysis is essential for the fatigue strength design of cast components (Tebaldini et al., 2017). This study, focusing on the scatter of fatigue strength, aims to investigate the propagation and non-propagation mechanisms for the fatigue cracks emanating from defects that dominate the fatigue strength of Al-Si-Mg cast aluminum alloys. Another purpose is to examine the applicability of the √ parameter model to the present material. 2. Material and experimental procedure 2.1. Material The material used was a cast aluminum alloy JIS AC4C-T6 (equivalent to A356-T6). The average grain size measured by electron backscatter diffraction (EBSD) was about 0.5 mm, but large grains exceeded 1.5 mm. After degassing, the melt was cast into a metal mold, and the cast blocks were age-hardened by the T6 heat treatment. The specimens for fatigue tests were cut out from the lowest part of the cast block. Extraordinarily large pores were occasionally observed on the fracture surface of the specimens that failed in tensile tests and had a crucial influence on the monotonic stress-strain curves. The tensile strength and % elongation of the specimens with no defects identified on the fracture surface were 310 MPa and 13.3 %, respectively. The average Vickers hardness measured with a load of 2.00 kgf (19.6 N) on the cross-section of the cast block was HV = 109. 2.2. Experimental procedure To investigate the effects of artificial defects on the fatigue strength by minimizing the impact of casting defects, rotating bending fatigue tests were performed using small specimens with a diameter of 5 mm. An identical four-hole artificial defect of √ = 287 µm in size was drilled onto the surface of all specimens, as shown in Fig. 1. After introducing the artificial defect, the specimen surface was polished using emery paper of a grade from #150 to #2000 and buffed using an alumina suspension (particle size: 1 µm). The testing machine used was a Shimadzu rotating four point bending tester with a moment capacity of 15 Nm. In this study, the experimental fatigue limit σ w, exp is defined by the lowest value of the stresses under which a specimen endures without failure until the number of cycles N = 2 × 10 7 . After the fatigue tests, the presence or absence of a non-propagating crack at the artificial defects was confirmed for all specimens that had not broken.