As-cast variations in mechanical testing of cast iron and its use to design tests for parameter studies

摘要

When making casting trials to improve properties of the cast iron products as a function of deliberately altered process parameters such as raw materials, additives or processes; output parameters such as microstructure data, thermal or mechanical properties need to be measured. Since variations in the intended casting process, sample preparation and test procedure will affect the results; these variations may overshadow the change due to the studied and altered parameter of interest. This work aims at investigating the variation in the mechanical test results of as cast test samples and uses these data to predict how many samples are needed to attain a specified confidence interval for a specific measured output property such as the ultimate tensile stress, yield stress, elongation and Brinell hardness. These data can be used to state with a specified confidence whether a change in a specific output parameter is significant or not and in fact depends on the altered input parameter. The investigation has been performed in both grey and ductile irons. The work shows that the confidence can be substantially improved by casting several samples in parallel and evaluate all of them. The work also discusses the important effect of microstructure formation during cooling and the effect of this on requirements on the geometry of the casting.