There is increasing interest in the lost foam casting technique because of its ability to produce near-net-shaped components of high complexity. The idea is to first make a prototype of the part to be cast in foam. This is then used as a pattern that can be placed in a box and surrounded by sand. Finally, metal is poured such that it smoothly replaces the foam by melting and/or evaporating it.

The stiffness of the foam makes it possible to cast parts having thin walls or other fine-scale features, and since the foam does not have to be removed at the end of the casting process, parts can be made that require fewer gaskets to assemble. Furthermore, because the foam pattern holds the sand (mold) in place there is little need to use binders in the sand, which means that the sand doesn’t have to be disposed of and can be used again. All these features make the lost foam process highly attractive to manufacturers.

Unfortunately, one rarely gets a free lunch and lost foam casting is no exception. For the process to be successful there must be a high degree of process control. Foams must have the proper characteristics and be coated with just the right material, and pouring sprues and gates for delivering metal to the mold must be carefully arranged. Metal pour temperatures must be sufficiently high to prevent premature solidification. And finally, the filling pattern of a mold should be such that metal fronts do not merge in a way that traps liquefied foam material, which could cause internal defects in the cast part.

To help casters address some of these difficult problems the computational fluid dynamics (CFD) program FLOW-3DÒ has been outfitted with special modeling capabilities to simulate the lost foam process. Using these models, it is possible to simulate the filling of a lost foam mold and the subsequent solidification of the metal. An extra feature in FLOW-3DÒ is the capability to predict where folds or other defects associated with trapped foam products are likely to be located.

The purpose of this paper is to demonstrate the usefulness and accuracy of lost foam predictions made with FLOW-3DÒ by presenting a direct comparison between experimental and computational results. The example chosen for this comparison is described in the next section. Subsequent sections present the comparisons with an emphasis on how the computational results can be used to understand why things happened as they did. This last point is most important, because it offers the user direct evidence and insight into how a casting could be improved.



Predicting Defects Lform