| Three linear shrinkage models were evaluated for their ability to accurately predict linear shrinkage in polycarbonate and a poly(butylene terephthalate) (PBT). Of the three linear shrinkage models, the corrected residual in-mold stress (CRIMS) model reflected the influence of processing conditions, overpredicted the in-flow linear shrinkage of polycarbonate by 10%, and provided poor predictions of cross-flow shrinkage. Unexpectedly, the uncorrected residual stress model underpredicted the inflow linear shrinkage of PBT. An examination of underlying causes showed that input flow rate and pressure and thermal properties produced small (∼1%) changes in linear shrinkage predictions, that three-dimensional heat transfer provided a 8 to 10% decrease in the cavity's cross-sectional area, and that volumetric shrinkage did not significantly affect the linear shrinkage predictions for polycarbonate. The results suggest that linear shrinkage data and the ability to correctly predict residual stresses produced the discrepancy between measured and predicted shrinkage for polycarbonate. Unexpectedly low volumetric shrinkage, a low solidification temperature, and adhesion of the polymer to the mold walls may have caused the underprediction of linear shrinkage for PBT. |
