Filters help eliminate dross in copper-base alloy castings.

Research on various copper-base alloys using three types of ceramic filters shows a dramatic reduction of dross inclusions in castings.

In both ferrous and non-ferrous metalcasting, employing the appropriate gating system can minimize the entrapment of nonmetallic inclusions. In many cases, however, the use of filters in those gating systems eliminates even finer inclusions, while improving the appearance, quality and properties of the castings.

In the nonferrous industry, filtration of aluminum alloys is common, while to a lesser extent filters are being used for magnesium and zinc-base alloys. However, little research has been done on the effects of filtering on castings of the family of manganese bronzes and silicon brasses and bronzes, which are susceptible to dross-related defects. The possible benefits of filter use have not been thought by copper-base foundrymen to justify the expense.

Thus, testing was undertaken to determine the usefulness of three types of ceramic filters in improving castings poured from various copper base alloys.

Setting Up the Tests

The principal filter types used in foundries are ceramic foam, which exhibit a random pore size distribution, and the extruded cellular ceramic type with straight pores of constant cross section.

Recently, a pressed strainer-core type filter (SC) has been introduced, which also exhibits straight pores of constant cross-section. These filters had never before been used in nonferrous applications.

The basic plan to test these filters was to pour four plate castings, each measuring 150 x 150 x 25 mm, from a single sprue and runner [ILLUSTRATION FOR FIGURE 1 OMITTED]. The three types of filters were incorporated into each mold, one on the ingate of each of three plates, with the fourth plate unfiltered. The riser of each plate was not placed above the ingate, but at the opposite end of the plate to show dross on the cope surface. The plates were molded in green sand, and care was taken to ensure that the metal could not flow around the edges of the filters.

Four alloys were chosen for the tests. Nickel-aluminum bronze (C95800) - a common alloy for marine casting applications - is the most dross-forming alloy. C86300 and C86400 manganese bronzes were also chosen. C86300 is the most prone to drossing in this family of alloys. Finally, silicon brass (C87500), with intermediate drossing characteristics, was selected.

After cooling, the plates were examined for dross-related defects, and an evaluation made of each filter's effectiveness. These examinations were made visually and with X-ray radiography, taking variables like alloy composition, pour temperature and filter fineness into account. Tensile and impact specimens were machined from the castings to determine whether filtering affects the castings' mechanical properties. In the case of the C95800 and C86300 alloys, cast-to-size test bars were poured with and without filters along with the plates.

Mold Filling Ability

It was necessary to judge whether filtering would hinder the filling of the mold cavity, and if the temperature of the metal and the fineness of the filter would affect this concern. Low and high superheat pour temperatures were selected for each alloy, and the various finenesses of filters were tested as well.

When pouring the C87500 alloy, there was no difficulty filling up the mold regardless of metal temperature or the type and fineness of the filter used. In the case of the two high-strength yellow brasses (C86300 and C86400), plates filtered with the fine SC filter could not be filled at 140F (60C) superheat. As for the C95800 alloy, plates with the fine ceramic foam filters couldn't be filled at both 140F and 212F (100C) superheat. In addition, plates with the fine SC filter couldn't be filled at 140F superheat.

All the plates could be filled using the extruded ceramic filters, irrespective of alloy, superheat and fineness. It was found that the superheat should exceed 212F to fill nickel-aluminum castings with fine ceramic foam filters, and if the SCs are used, the superheat should be above 140F.

Quality Evaluation

Upon inspection, all the unfiltered plate castings showed dross inclusions on the cope surface near the ingate and in between the gate and riser locations [ILLUSTRATION FOR FIGURE 2 OMITTED]. By contrast, the top surface of each filtered casting looked clean, and only a few dross spots were observed in each alloy, regardless of the filter type.

Examination of the transverse sections of the plates near different drossy areas indicated that the inclusions penetrated to a depth of 8-10 mm from the top surface of each alloy. In general, the depth of the dross was related to the intensity of dross on the surface, and relatively more dross was observed on castings poured with high superheats. When the sections through the thickness of the filters were inspected, it was seen that most of the dross was trapped in the top half of the filter [ILLUSTRATION FOR FIGURE 3 OMITTED].

In respect to mechanical properties, tensile specimens machined from filtered and unfiltered castings revealed no difference at all. Un-notched charpy impact specimens were also taken, making certain that some dross was included in the center of the specimens. Again, there was no difference in the impact values in the case of the two high-strength yellow brasses. With the other alloys, however, the impact toughness of the filtered castings was higher than that of the unfiltered, the difference for silicon brass being the most dramatic.

From these results, it is clear that filtering is beneficial to copper-base castings, regardless of which filter is used. However, foundries must consider the relationship of the temperature and type of alloy to the filter fineness, and also the purchasing cost of the various types of filters, before deciding to incorporate them in molding operations.