Pressure drop data was collected for 44 cellular and 18 foam filters using water flow.Pressure drop correlations were developed for 100, 200 and 300 csi cellular iron filters and 10, 20 and 30 ppi foam iron filters. Pressure drop for these filters was dependent upon cells (pores) per square inch and filter thickness. Pressure drop decreases with decreases in cell count, and increases with increases in filter thickness. The combination of these characteristics plus the porosity determine the pressure drop curve for a filter.Accurate correlations with confidence levels generally above 90 percent were developed when the filter pressure drop data was divided by filter thickness, and grouped by cell count. Foam filters showed slightly more variability in pressure drop than cellular filters, but only for 10ppi filters. Even still, correlations for the 10ppi filters achieved an 87 percent confidence level, which is quite good. Future work will review this situation more closely.The pressure drop correlations were converted to a standardized description of filter flow characterstics, which involves computation of the Darcian and non-Darcian permeability coefficients. The power of these coefficients is that they are the standard for characterizing filter flow performance, and are not dependent upon the fluid medium used to collect the pressure drop data. More importantly, the pressure drop performance of the filter can be determined for any fluid flow using these coefficients, assuming that the fluid viscosity and density at the pouring temperature are known. These are the values used for computer simulation analyses.Computer simulations were conducted for several casting configurations using MAGMASOFT. Each configuration was run with the "newly" developed data, and re-run with corresponding filter data available in the current literature. The program was given the metal flow stream information, and asked to predict the fill time. In all cases, the predicted fill time for the case suing the newly developed data more closely matched the "real" fill time (measured in the foundry), than the predicted fill time when using the previous set of filter data. This validates the accuracy and usefulness of the data for computer simulations.As expected, filter pressure drop increased with increasing cells (or pores) per unit area of the filter. In addition, foam filters exhibit considerably higher pressure drop values than cellular filters.
展开▼