Reducing Porosity in Aluminum Lost Foam Castings Through Computer Simulation

The lost foam casting (LFC) process offers unmatched versatility for producing near-net shape complex components for the automotive industry. Castings made by the LFC process can replace numerous individually-manufactured parts through part consolidation, have little or no draft, and have minimum machining stock. However, the filling behavior of lost foam castings is very different from that of open-cavity castings, so methods to reduce casting anomalies are not intuitive. Casting simulation codes have been in use for the past three decades and they are reasonably successful in predicting porosity in steel and other ferrous castings. The prediction of porosity in aluminum alloys is difficult because of long freezing ranges and lack of accurate thermo-physical property data. Also, for lost foam castings, large thermal gradients are created by slow mold filling, which further complicates porosity prediction. Both accurate thermo-physical property data and a modified solidification subroutine have been developed so that the temperature distribution at the end of filling and macro-porosity formed during solidification can be accurately predicted in lost foam aluminum castings. Model predictions have been validated with experimental cooling curves and foundry trials. Simulation results have shown that the gating system design for the lost foam casting is crucial in establishing thermal gradients for subsequent solidification. Experiments with commercial lost foam castings validated simulation predictions.