No abstract

To better understand the effects of cast defects on mechanical properties, cast aluminum alloys with various porosities were used. Porosity in the cast samples was created during the casting process, and to clearly identify the porosity effects on the mechanical properties, artificial defects (porosity-like tiny holes) were created mechanically. The tensile properties for the cast aluminum alloys appear to be attributed to the area fraction of the porosity on the fracture surface (namely, the defect rate, DR), although there were different trends because of the different stress concentrations: the ultimate tensile strength and 0.2 pct proof strength were linearly related to DR, while a non-linear correlation was detected for fracture strain. Even in Al alloys with small amounts of defects, significant reductions in the fracture strain were observed. These results were verified using tensile tests on specimens containing artificial defects. The effects of artificial defects on the tensile properties were further investigated using numerous tiny holes, created in several formations. The artificial defects (several small holes), lined up at perpendicular (90 deg) and 45 deg directions against the loading direction, made significant reductions in the tensile properties, even though only weak defect effects were observed for the 90 deg loading direction. No severe defect effects were obvious for the specimen with a tiny defect of /0.1 mm, because of the lower stress concentration, compared to the microstructural effects in the cast Al alloys: the grain boundaries and the second phases. Such phenomena were clarified using tensile tests on cast samples with differently sized microstructures. There were no clear defect effects on the yield strength as the defect amount was less than 10 pct, and microstructural effects were not detected either in this case. Failure characteristics during tensile loading were revealed directly by in-situ strain observations using high-speed cameras.

A modified linear-elastic method of analysis was proposed for the estimation of fatigue lifetimes of biaxially loaded components containing cracks. A constitutive relation of the crack growth rate, d a/ d N, was used as a basis for the analysis. The modifications included a correction for elastic-plastic behavior, consideration of Kitagawa effect, and consideration of the development of crack closure in the wake of a newly formed crack. A unified evaluation for the d a/ d N data measured by Brown and Miller in different biaxial stress states was made by use of the proposed method. It was found that the difference in elastic-plastic behavior accounted for the effect of biaxial stress on d a/ d N.

No abstract