Midea, Anthony C

Abstract: The main objective of this project was to develop pressure drop curves for cellular and foam iron filters. The data are required to characterize the steady state flow through a filter at various flow rates. The resultant pressure drop data are valuable for improving the accuracy of fill time predictions and flow modeling in computer simulations. A total of 62 filters were tested for pressure drop using a water flow apparatus. 44 tests were conducted on cellular filters, and 18 tests were conducted on foam filters. Square and round filters were tested. Filter sizes ranged from 37mm to 55mm for the cellular tests, and 50mm tests were conducted for the foam filters. The pressure drop characteristics were consistent and repeatable. In general, the pressure drop depended upon the filter cell count and thickness, and not the geometric size. Pressure drop correlations were developed for 100, 200 and 300 csi (cells per inch) cellular filters, and for 10, 20 and 30 ppi (pores per linear inch) foam filters. Correlation accuracy was generally greater than 90\% with a confidence interval of 95\% (three sigma). In general, the pressure drop for a given filter type increased with cell count. Pressure drop values for foam filters are significantly higher than for cellular filters. Some variation in pressure drop was noticed for 10 ppi filters due to foam variability. However, the variation was considered minor due to the 87\% accuracy of the correlation. 20 and 30 ppi filters did not exhibit this characteristic, and correlations for both filter types exceeded 95\% accuracy. Foam filters will be analyzed in further depth in a follow-on project. The pressure drop data were converted to a standardized format, and Darcian and non-Darcian permeability coefficients were calculated. These coefficients were input to a physics based mold filling/solidification computer simulation program, and several simulations were conducted to compare the results with previous results using “generic” filter data. Using the new filter data, the user is able to more accurately predict the time to fill for the casting cavity.