Getting to the core of thin-walled castings: through the addition of a sand additive/replacement to the cores, thin-walled castings with properties comparable to those with thicker walls and great weight were produced.

Economic and environmental pressures continue to force metalcasting facilities to look for ways to produce castings with thinner walls and at lighter weights but with the same or better mechanical properties. However, several key requirements exist to produce thin-walled castings with good mechanical properties:

* the molding materials and processes must provide a high degree of dimensional accuracy to maintain specified wall thicknesses;

* the combination of metal chemistry, fluidity, pouring process, gating system design, venting and molding materials must allow the mold to fill completely before the onset of solidification and the occurrence of misruns;

* the metal must solidify and cool without the formation of carbides, centerline porosity or other micro-structural deficiencies that would impair the casting's physical properties.

To date, the focus has been more on metal composition, inoculation practices and gating than on mold and core materials. However, improvements in the heat capacity and thermal conductivity of these materials may have the potential to reduce wall thickness more effectively than all of the other factors combined. Investigations have shown that the use of low-density alumina-silicate ceramic (LDASC) as a sand additive/replacement can dramatically reduce the heat capacity and thermal conductivity of cores, providing the potential for the production of thin walled castings in a variety of metals. In iron alloys, thin sections can be produced without the formation of primary carbides and with good mechanical properties. Thin-walled ductile iron castings can be made with microstructures and physical properties equivalent to heavier sections. Steel and stainless steel alloys can be poured in thin sections without misruns or hot tears.

Low-Density Alumina-Silicate Ceramics

LDASC has been commercially available for several years as additives that are successful for anti-veining and improving surface finish. They also have been used in coatings and as a primary constituent of both insulating and exothermic riser sleeves. LDASC typically is composed of 25-40% alumina and 55-75% silica. The material is inherently refractory with melting and softening temperatures ranging from 1,200-1,600C (2,200-2,900F). The material has thermal diffusivity properties comparable to other refractory materials, but its low density provides bulk heat capacity and thermal conductivity values that are much lower than typical foundry aggregates. Typical properties and a SEM photograph of the material are shown in Table 1 and Fig. 1.

[FIGURE 1 OMITTED]

Coremaking with LDASC

LDASC can be used as a sand additive /replacement to make cores using a variety of binders and production methods. The material can be viewed as a sand replacement because it accounts for a significant portion of the volume of the sand mix and has a significant effect on the bulk properties of the final core. But the small particle size and low density of the material present some special problems for coremaking.

The first hurdle to overcome is in the thought process. Most metalcasters simply don't think that a core should be light enough to float on water; nor do they think cores should contain more than 1-2% binder by weight. Therefore, it may be more helpful to think in terms of volume. If 100% LDASC was used to produce a core, it would have the same volume but only 25% of the weight of silica sand.

The same amount by weight of LDASC would produce about four times the volume of silica sand cores. In addition, the small particle size of the material requires proportionally higher binder levels because of the larger surface area.

But sometimes a core with LDASC can create special processing problems. First, because of the low density and small particle size, changes in the mixing equipment or cycle could be needed to adequately coat the material with binder. The coated material is less flowable than 100% sand and could be difficult to blow into a corebox. Its smaller particle size and sleeve lower permeability may make it more difficult to cure and may increase the tendency to plug corebox vents. Also, because the material has a melting point lower than silica sand, exposure to molten metal for extended periods of time could result in premature softening of the material and cause poor surface finish or core collapse.

In most situations, it is more effective to use a mixture of sand and LDASC. Even relatively small additions of the material can reduce core density and improve molten metal flow. The density and corresponding thermal properties of the blend can be tailored to control cooling rates based on section thickness.

For relatively heavy sections of 4-6 mm (0.15-0.25 in.), small amounts of the material can provide some improvements in metal flow and cooling rate with good surface finish and dimensional control. However, higher levels could result in softening of the material and cause surface roughness and possible dimensional changes.

For thinner sections, larger amounts of the material may be needed to improve metal flow and cooling rates. However, the lower heat exposure from the thinner section still results in good surface finish and dimensional accuracy. Using a core made with a high ratio of the material in a silica sand mold could help produce a thin-walled casting because the core would extract less heat than the mold and increase metal flow while slowing solidification.

Effects on Ductile Iron Castings

A series of specially designed thinwalled ductile iron castings were poured with varied levels of LDASC to evaluate the effects on the microstructure and hardness. Sixteen thinwalled manifold test castings were produced with different amounts of LDASC in both the cores and the facing mix on the mold surfaces.

The manifold casting (Fig. 2) is about 15 in. (381 mm) long with a 2-in. (51 mm) internal diameter. The wall thickness is approximately 0.079 in. (2 mm) with three heavier flanges that are 2 x 2 x 0.5 in. (51 x 51 x 13 mm). The castings were poured two-on with a single gate into the flange nearest the sprue. The cores and molds were produced with a phenolic urethane nobake system with 1% binder by weight on the sand. Binder levels were adjusted for blends of the material and lake sand with 8% binder by weight used with 100% LDASC.

[FIGURE 2 OMITTED]

The molds were poured at 2,350F (1,288C) with ladle-treated ductile iron containing 3-5% carbon and 2.4% silicon. The metal was then inoculated with 75% FeSi during treatment and again just prior to pouring. Some cold-shuts were present on two castings but all others were complete,

Following shakeout, the castings were sectioned through each of the three flanges. The sections from the middle flange (located 5.5 in. from the gate) were ground and polished for metallographic examination. Photomicrographs were taken in the thin sections 180 degrees from the flanges and showed the effects of reduced solidification and cooling rates with increased levels of LDASC. Higher amounts of the material in the mold facing reduced the percentage of carbides in the thin section and increased nodule size and ferrite levels.

The same trend was apparent with higher amounts of the material in the core. The casting produced with a 100% LDASC mold and core was completely free of carbides and had a microstructure typical of a much heavier section. The heavy sections had lower levels of carbides than the thin sections and were carbide-free with lower levels of the material.

The Rockwell hardness of the castings was checked in the heavy sections of the second flange and in the 2-mm section opposite it and converted to Brinell by ASTM B10. The casting hardness decreased dramatically in the thin sections with increasing LDASC levels.

The effects were less dramatic but still significant in the heavy sections. Both the casting microstructures and hardness values showed significant improvements with increasing levels of the material in contact with the casting.

The 25% LDASC blends appeared to have minimal affect on structures and properties, but both the 50% blend and 100% LDASC had significant effects. When the results are compared to a total LDASC percentage (half the facing percentage plus half the core percentage), a consistent picture of effects is noticeable.

Effects on Steel Castings

Trials also were conducted with stainless steel alloy CF-SM (AISI Type 316). Fluidity spiral molds made with silica sand and faced with sand/LDASC blends were used to quantify the effects on flow distance. The flow distance increased dramatically, indicating the potential to reduce misruns and casting thickness, as well as pour equivalent castings at lower temperatures.

An impeller casting (Fig. 3) with a history of misrun vanes was selected for the trial. In an attempt to eliminate misruns, production castings had typically been poured at temperatures of 1,620-1,635F (2,950-2,975F), which is 60 degrees higher than normal. Despite the unusually high pouring temperature, the scrap rate was excessive. To evaluate the effects of LDASC in eliminating misrun vanes, cores were made with 100% LADSC and a 50/50 sand/LADSC by volume blend with silica sand. An ester cured phenolic resin was used. Resin percent ages were 6% for 100% LDASC and 2.2% for the blends.

[FIGURE 3 OMITTED]

All of the cores that contained LDASC eliminated misruns, even at normal pouring temperatures. The casting produced with 100% LDASC had rougher surface finishes caused by the softening of the material in high heat areas. But the 50/50 blend eliminated misruns, and combined with a suitable coating, provided an acceptable surface finish.

Uncharted Territory

The use of LDASC as part of a total foundry process can dramatically improve metal flow, reduce misruns in thin sections and slow solidification rates in thin-walled castings. It makes the production of thin-walled gray and ductile iron castings possible without the formation of iron carbides and with physical properties comparable to much heavier sections. The blends can be tailored to specific casting section thickness and geometry.

able 1. Typical Properties of LDASC

Appearance:               Off-white to gray free flowing powder
Particle Distribution:    10-350 microns
Mean Diameter:            120-130 microns
Average Wall Thickness:   5-10% of diameter
Bulk Density:             0.35-0.45 g/cc (about 25 lbs./[ft.sup.3])
pH:                       7.0-9.0
Thermal Conductivity:     0.15-0.25 W/mC
Melting Temperature:      1,200-1,6000 (2,192-2,912F)
Specific Heat:            0.6-0.8 J/gC

For More Information

"Process Variables Affecting the Erosion Resistance of Phenolic-Urethane Coldbox Cores," C.M. Henry, R. Showman, D. Kahles and M. Nikolai, AFS Transactions, paper No. 03-144.

"Evaluations of Sands for the Coremaking Process: A Practical Approach, " J. Werling, AFS Transactions, paper No. 02-030.

About The Author

Ralph Showman is a senior staff scientist and Ronald Aufderheide is a senior product manager for Ashland Casting Solutions, Dublin, Ohio.