Fig. 1 (a) Discovery of high-density coagulation of micropores in regions not deeper than 30 μm from the casting surface, and resultant drops in fatigue life. (b) Specimen with almost no micropores had at least double the fatigue life. Both were observed under CT, and only the internal pores are shown.
We have repeatedly visualized the fracture phenomena of various materials, and made many discoveries that disagree with conventional theories. Below is an example. We discovered hydrogen micropores coagulating at high densities several tens of micrometers below the surface of aluminum alloy die cast (Paper 2
). The pores were newly found defects not detected by conventional observation methods. Figure 1 shows the relationship between the presence of the pores and fatigue life.
Fig. 2 In-situ observation of fatigue crack initiation from the high-density coagulation of micropores in regions not deeper than 30 μm from the casting surface. The cracks developed most easily from micropores that aligned horizontally at high densities.
Figure 2 shows the process of fatigue crack development from a cluster of pores (diameters < 10μm) located beneath the surface (Paper 2
).The fatigue crack was about 20 μm long when it first appeared, and developed to about 30 μm. As shown in Fig. 3, the spatial distribution and morphological effects were statistically analyzed to understand the geometrical conditions predisposing to fatigue crack generation(Paper 3
).The ductile fracture of metal materials is common knowledge, even described in textbooks(Review paper 2,Paper 1
).However, the actually tested materials showed completely different phenomena.
Fig. 3 Statistical analysis of fatigue crack initiation from pore clusters, using a multiple regression analysis model. The mean diameter of two adjacent pores (the mean distance between Parameter a2and the casting surface (Parameter a4)controls the initiation of fatigue cracks.
Fig. 4 4D visualization of the ductile fracture process of 2024 aluminum alloy. The indwelling pores grew and were the main cause of the fracture.
Figure 4 shows that the high-density indwelling hydrogen pores, which were produced during the manufacturing process, grew and coagulated rapidly, and induced ductile fracture. The conventionally known fracture mechanisms act as a secondary and supplementary cause.
Fig. 5 Retroactive tracking of coordinates of the ductile fracture surface of 2024 aluminum alloy (left). Resultant identification of dimple patterns caused by hydrogen pores on the surface (right: yellow).
We identified 500,000 fracture surface coordinates, and traced the displacement of dispersed grains retroactively, to analyze in which part of the pre-load specimen the fracture surface was located (Fig. 5). Based on the results, the dimples on the fracture surface were classified into those originating from hydrogen pores and those formed by the conventionally known mechanisms. The region that resulted from hydrogen pores accounted for 39% of the entire fracture surface. However, the damage initiated from the indwelling pores occurred first, and thus the effects of the indwelling pores on the ductile fracture were larger than the area ratio. The material control of structural metal materials has involved controlling the kind, size and distribution of dispersed grains. However, the results described above suggest that factors that have not been noticed, such as hydrogen solubility and the number, size and spatial distribution of pores, may be the dominating factors in such fractures. For example, conventional tissue control guidelines state that reducing the number and size of dispersed grains is effective. However, this results in increasing the size of pores. Our findings are interesting in that they suggest the effectiveness of a material control method exactly the opposite to that of the conventional common knowledge.