Fatigue life prediction and reliability assessment of ductile iron castings using optimized mold design

Abstract

Metal castings are often questioned for their service life and reliability due to a lot of uncertainties associated with the process. It is important to virtually analyze these aspects of a cast product before it actually put into service. This paper presents a new approach to predict the fatigue life and reliability of ductile iron castings through coupled simulations and experiments. In this work, fatigue specimens are considered as sample cast products which are produced using an optimized mold design for minimum porosity. Radiographic examination is done to observe the quality of cast specimens. Experimental load-controlled fatigue testing of selected specimens is done and compared with simulated fatigue lives in fe-safe (Dassault Systems UK Limited 2015). Reliability of cast specimens under fatigue is computed using stress-strength model by considering the variability in both stress and strength. A safe load-induced stress is estimated to obtain infinite life of ductile iron parts under consideration. Furthermore, the reliability models are developed by fitting probability distributions to reliability estimates and determining the distribution parameters. It is concluded that the mold design optimization using casting simulations can lead to nearly defect free castings which can serve for the similar fatigue lives as their sound counterparts. Moreover, the probability of survival of the cast components for an infinite life can be computed using the classical stress-strength reliability model based on which safe loading conditions can be determined for almost any engineering application.