Factors Affecting Microshrinkage Formation in Thin Wall Ductile Iron
Jonathan W. Woolley, Roxana Ruxanda, Marian Liliac, Masami Fukumoto, Doru M. Stefanescu and Christoph Heisser
American Foundry Society Transactions, Volume 112
ABSTRACT
Recent research on thin wall ductile iron (DI) castings has demonstrated their excellent static mechanical properties which make them a material of choice for automotive applications. However, to reach the full potential of thin wall DI castings it is necessary to identify the metallurgical and processing conditions necessary to obtain sound castings with the desired microstructure. The paper presents an analysis fo the effect of process variables on microshrinkage formation in thin wall DI gravity cast in to resin bonded sand molds.
The mold design has a critical influence on the formation of microshrinkage, since this determines the extent to which liquid metal can be fed to each plate. Accordingly, three different patterns with horizontal plates were used to physically simulate feeding through a riser, feeding from an adjacent section, and no feeding. Other variables included the pouring temperature (1357 to 1430 C), the carbon equivalent (4.58 to 4.96), the cooling rate (in the range of 5 to 25 C/s), and the casting modulus (modulus = volume/area 1to 3.4 mm). The influence of these variables on the formation of microshrinkage has been evaluated in DI plates of nominal thickness ranging from 2 to 6 mm.
It was concluded that microshrinkage can be easily avoided with appropriate feeding either from a riser or from adjacent sections, under the processing conditions for thin wall DI castings used in this study. The experimental results demonstrate that, with proper feeding, it is possible to cast sound horizontal plates with dimensions of 2x60x100 mm.
Computational experiments with the Magmasoft software correctly predicted the absence of microshrinkage in plates fed through a riser 75Dx70 mm. These plates had thickness of 2, 4, 6, and 8 mm. The other dimensions were 40x95. Successive increase in the length and width of the plates did not result in microshrinkage prediction until the plates reached a length of 190 mm and a width of 160 mm. This is probably optimistic and validation experiments have yet to be run.
For the independent plates that were not fed it was found that the microshrinkage increases with plate thickness, or conversely is smaller at higher cooling rates. The wider 60 mm plates, typically exhibited more shrinkage than the narrower ones (40 mm).
