Statistical modeling of microstructure and defect population effects on the fatigue performance of cast A356-T6 automotive components

With the increasing usage of cast aluminium–silicon alloys in the automotive industry, the fatigue performance of castings is becoming a major concern with regard to component integrity and reliability. Traditionally, the fatigue design of cast components has been based on extensive material property databases. The relationship between microstructure and the resulting fatigue life is primarily qualitative. In this study, an effort was made to quantify the fatigue performance of A356-T6 castings through both experiments and statistical modeling. The results indicated that, in spite of the consistent secondary dendrite arm spacing for all specimens, a wide range of microstructural and defect features were responsible for the fatigue behavior. Statistically, when a large pore or oxide film is present at or close to the specimen surface it can act as a favorable crack initiating site. However, in the absence of large pores or oxide films near the specimen surface, the scale of the micro-cells, comprised of eutectic surrounding the primary phase, determined the fatigue performance. On the basis of the experimental observations and probabilistic statistical analysis, a physical relationship was established between the casting-related microstructural features and the resultant fatigue life, including associated scatter, of cast A356-T6.