Manufactured alternatives to traditional molding media.

As foundries continually strive to improve casting proficiency, three alternative, manufactured molding media are available to eliminate many of the defects associated with traditional sand molding.

Since the dawn of metalcasting, foundrymen have worked to improve the strength and stability of their sand molds. Initially, they simply used natural sand - without any binder additions of clay or water - to build strong and durable molds. In the 1920s, synthetic molding - the use of bentonite clays and water in molds - was introduced, as foundrymen formed a stronger, naturally bonded mold. In the 1950s, nobake molding was developed, providing a more stable and durable environment to receive metal. In other developments through the years, additives like seacoal, cereals and cellulose materials have been added to the molding media to promote further stability, reduce thermal expansion and increase thermal conductivity.

However, these advances and additions to the media will never result in an ideal metalcasting mold, according to Jack Gentry, an industry consultant in Palm Springs, California. The problem isn't what is being added to the mold - it is the sand.

"Silica is not the ideal molding material," said Gentry. "Veining, scabbing defects and cracked molds are results of materials that are not thermally stable, as they expand and contract upon the addition of molten metal. Although oil-vine is an option and does relieve some of the expansion defects and reduce scrap, it still isn't perfect."

Table 1. Chemical Analysis of the Ortho-Silicate by Weight %

Material                                                    Weight %

Si[O.sub.2]                                                   50.14
MgO                                                           31.35
[Fe.sub.2][O.sub.3]                                           15.87
[Al.sub.2][O.sub.3]                                            1.6
CaO                                                            0.67
NiO                                                            0.14
Ti[O.sub.2]                                                    0.13
[Na.sub.2]O                                                    0.05
[K.sub.2]O                                                     0.03

The need for improved molding materials has been addressed with the development of manufactured sand alternatives that begin to alleviate the inherent problems - thermal expansion, thermal conductivity, durability and stability - that accompany traditional molding media. Three products' in today's market - Green Diamond, Cast-rite and Ceramacore - are manufactured alternatives for foundries looking to complement or replace their current sand system.

ORTHO-SILICATE

Developed as a product for the sand-blasting, roofing and road-surfacing industries, Green Diamond aggregate, manufactured by Green Diamond Products, Inc., Riddle, Oregon, was introduced to foundries in 1995 as an alternative to silica sand and olivine.

The alternative's chemical makeup, as shown in Table 1, is similar to natural olivine. It is a fused ortho-silicate that is formed from molten lateritic ore - a by-product of the ferro-nickel smelting process. The molten ore is water quenched to produce a consistent, man-made sand. After further processing and blending, the resultant 45-145 GFN ortho-silicate has these reported advantages oversilica and olivine sand:

* higher durability;

* lower bulk density;

* lower thermal conductivity;

* less thermal expansion;

* environmental friendliness.

At the heart of the ortho-silicate's advantages is its durability. Its development at high temperatures during the smelting process and the subsequent quenching produce an orthosilicate of low friability with a hardness value of 6.5-7 on the mohs hardness scale. These properties result in the reduced breakdown of the orthosilicate with repeated use, less dust generation and the reduced introduction of replacement ortho-silicate after each molding cycle.

In one example, 50, 70 and 100 mesh samples of both the ortho-silicate and olivine were screened and then tumbled in a 4-in. drum containing four 0.5-in. round steel rods for 2 hr. After a re-screening, the ortho-silicate showed 40% less breakdown than the olivine.

Another physical property benefit is the bulk density of the ortho-silicate, which is up to 20% lighter than olivine and 8% lighter man silica. This produces more molds and cores per ton of material.

The third advantage the ortho-silicate provides is a thermal conductivity lower than either olivine or silica. The lower conductivity produces a greater insulating medium that reduces chill and allows for lower pouring temperatures (30-50F lower) and a reduction of binder burnout. The net result is lower cost, as make-up sand, metal poured and binder additions are lessened.

In an experiment last year to test the insulating value of the ortho-silicate and the flow of molten metal, a mold was set up with three runners faced with three different materials - the ortho-silicate, silica sand and olivine - each bonded with sodium silicate nobake. Aluminum was poured down the sprue such that, after the gating system was full, the silica-faced strip was adjacent to the sprue, and the ortho-silicate and olivine strips were equidistant. As illustrated in Fig. 1, the ortho-silicate allowed the metal to flow more than 45% further.

The next advantage, thermal expansion, is best illustrated with Fig. 2. Independent lab test results showed a constant linear change in expansion for the ortho-silicate that was lower than either silica or olivine. For foundries, this result would reduce expansion defects without requiring the addition of cellulose or cushioning materials.

The last advantages of the orthosilicate are demonstrated after the mold has been poured. The shakeout of an ortho-silicate mold or core is better than a silica or olivine sand mold, as it easily falls apart with vibration. The ortho-silicate also performs well at shakeout on nobake molds. Since ortho-silicate is a glass, the adherence of the binder to the grains is substantially reduced after the material has cooled. This results in easy mechanical and thermal reclamation for the ortho-silicate after casting.

Environmentally, the ortho-silicate is an inert product without free silica, so the 50% Si[O.sub.2] present is locked up in the matrix of the grain in a form of silicate, reducing the risk of silicosis. In addition, the heavy metals in the ortho-silicate from the smelting process are below the levels set by the EPA for toxic compounds.

Despite the advantages listed above, the ortho-silicate does have one limitation when compared to silica: its sinter threshold is 2650F (1454c), which allows - at its highest temperature - manganese steel to be poured.

The cost of the ortho-silicate, as produced, is less than olivine, and, there fore, that market is being targeted. However, the ortho-silicate, as illustrated in the test results, isn't limited to replacing olivine. The angular and sub-angular shape of the grain is marketed as a replacement sand for all forms of molding and coremaking.

CARBON SAND

The oldest of the three manufactured molding mediums described here is Cast-rite, a round grain carbon sand that was introduced to metalcasters in the early 1960s by Gentry and has been marketed since 1989 by American Colloid Co., Arlington Heights, Illinois. This product is derived from petroleum fluid coke that has been calcined at 2200F (1204c) and can be used by itself or in mixtures with silica sand for cores and molds. The reported advantages it provides over traditional molding media include both physical and chemical properties at room temperatures.

The physical property advantages relate to the product's round grain and its low bulk density. The round grain characterizes this carbon sand as different from its predecessors that were crushed carbon shapes (electrodes) too porous to be used as the sole molding media. Because of its grain shape and 75 GFN, the sand resists crushing and promotes good flowability. This is especially important in core blowing, as the round grains will blow easier and consequently improve the density of the core.

In terms of weight, at 70 lb per cu. ft, the sand is 25% lighter than silica, thus reducing the workload on personnel and equipment, while increasing the molds and cores made per lb of material. For comparison, 1 ton of zircon sand has a volume of 12 cu. ft, 1 ton of olivine is 19 cu. ft, 1 ton of silica is 20 cu. ft and 1 ton of carbon sand is 28 cu. ft.

The chemical property advantages the sand provides to molds includes low thermal expansion and high thermal conductivity. Avoiding the sharp inversion thermal expansion that silica sand undergoes, carbon sand eliminates the defects that accompany expansion (such as veining and scabbing), as well as permitting hard ramming and high-pressure molding to further promote smoother finishes and dimensional accuracy. Carbon sand is equal to zircon sand in terms of thermal expansion as illustrated in Fig. 3, and is superior to silica sand.

In terms of thermal conductivity, the sand doesn't provide heat sink capabilities, but it will conduct heat faster and further than silica sand to promote improved surface metallurgy and soundness through faster solidification.

For core production, the carbon sand is usually blended with silica sand. When blended at 50% by weight, carbon sand eliminates the need for corewash because carbon is not wetted by molten metals. In addition, the sand can be used with all conventional binders, molding processes, equipment and types of sand for use in iron or nonferrous castings. Carbon-hungry steel will absorb carbon sand unless it is coated with a wash to prevent direct contact. In addition, carbon sand can be blended with any other sand in any ratio.

When pouring molten iron, the chemical inertness of carbon sand is another advantage. The formation of iron silicates during casting and the resultant penetration can be eliminated by replacing silica with carbon sand.

For molding applications, the performance of the product with various ratios of southern (calcium) and western (sodium) bentonite was measured in the early 1960s. Twelve batches were prepared for testing - six with 100% carbon sand and six-blended 50/50 by weight with silica sand. Analysis of the basic green sand tests revealed a logical pattern of results except in the case of hot compressive strengths as shown in Fig. 4.

In additional production data from the 1980s, the durability of the product was highlighted. Molds were made with jolt-squeezers and poured with 1375-2650F (746-1454c) iron, bronze and aluminum over an eight-month period. The sand system was turned 2-3 times daily during the test period, and except for additions of bentonite and water, the pile was maintained as originally established. Under these conditions, grain size remained close to its original size.

This production scenario also re-emphasized that the product performs best if used in conjunction with silica sand. From a 15% carbon sand/85% silica sand weight ratio to a 50/50 disbursement, the molds and cores will benefit from the properties of the carbon sand. Beyond a 50% mixture of carbon sand, the molds and cores have diminishing returns, as the benefits do not outweigh the costs of extra carbon sand.

Table 2. Analysis by Weight % of Carbon Sand

Material                                                    Weight %

Carbon                                                        93.72
Water                                                          0.1
Volatile                                                       0.15
Ash                                                            0.12
Nitrogen                                                       0.41
Sulfur                                                         5.5

Once the molds have been poured, the carbon sand can be mechanically reclaimed fully, and to some degree, thermally reclaimed. Carbon sand can be reclaimed thermally in a system having air control so that sufficient sand and organics are burned to maintain desired reclaimer temperature, without fuel gas burners.

Another process limitation for the product is that many automatic molding systems are programmed to work with silica sand, therefore the weight difference with carbon sand results in molds that aren't fully compacted. If a foundry is going to use carbon sand, an adjustment of equipment is needed.

Environmentally, carbon sand doesn't cause silicosis, because there isn't any silica in the product, but it is considered a nuisance dust (like silica), so facemasks are required. In addition, if a 100% carbon sand mold was utilized, the addition of seacoal would not be needed to counteract thermal expansion, thus eliminating benzene and formaldehyde emissions. If carbon sand is blended with silica to produce a mold, then the seacoal can be added at the same ratio, resulting in reduced emissions. Lastly, the sulfur in carbon sand, as shown in the analysis of Table 2, is tightly bound within the grain so as not to cause an environmental problem.

SYNTHETIC MULLITE

The newest molding media alternative is Ceramacore. Developed in the 1980s by Carbo Ceramics, New Ideria, Louisiana, as a proppant for the gas and oil industry, the synthetic mullite's potential use in metalcasting was recognized in a 1993-94 Dept. of Energy study that evaluated 16 different granular media as alternatives to silica sand. The evaluation concluded that the synthetic mullite line showed the best combination of properties as compared to silica sand.

With this recognition, two different forms of the synthetic mullite - 311 and 411 - were introduced to lost foam metalcasters as a backing material for polystyrene foam patterns. Over the last three years, through testing and foundry applications, several reported advantages have been detailed for the 311 product's 30/50 mesh (30 [TABULAR DATA FOR TABLE 3 OMITTED] GFN) and the 411 product's 30/60 mesh (41 GFN) over silica sand and olivine as a backing material:

* higher flowability;

* higher permeability;

* lower thermal expansion;

* lower bulk density for 311 when compared to silica and 411 when compared to olivine;

* reclaimability.

Tests and foundry practices have proven the synthetic mullite's spherical shape provides it with higher flowability and tighter compaction around foam patterns - while using less energy - when compared to the more angular products of the competition.

The second advantage for the synthetic mullite is its gas permeability capabilities. In its original use, the product was developed to allow oil and gas to flow through it. For the purposes of lost foam, this permeability allows the gas from the dissolved polystyrene to dissipate quickly through the backing material, eliminating many of the gas-related defects associated with the process.

The third advantage is the low thermal expansion of the synthetic mullite. Comparable to zircon sand for this property, the synthetic mullite has been used to pour molds up to 3000F (1649C) without any expansion or thermal related defects.

A perceived limitation to the synthetic mullite's implementation in lost foam is its price tag, which compares to zircon sand. This cost perception, however, is countered by the product's reclaimability and bulk density, which provide foundries a return on their investment. Both products are highly reclaimable, showing virtually no break-down through repeated use. The result is that the cost is absorbed through perpetual use of the products. This reclaimability is due to the construction of the products to withstand pressure - 10,000 psi for the 311 and 12,500 for the 411 - and temperature - 2900F (1593C) for the 311 and 2750F (1510C) for the 411.

Additional absorption of cost is with the bulk density. The 311 product has a bulk density comparable to silica, while the 411 is comparable to olivine. For one cu. ft of material, 311 weighs 95 lb, silica weighs 100 lb and 411 weighs 113 lb, so foundries are receiving more material per lb as compared to zircon and olivine.

These advantages in lost foam are what attracted Ashland Chemical Co., Dublin, Ohio, to the product in July of last year, when the company signed an exclusive marketing agreement with Carbo Ceramics. But the supplier's interests spanned beyond lost foam to possible implementation into traditional sand casting. The initial steps the company has taken with the synthetic mullite have been in the coreroom.

While the synthetic mullite can be used with all types of core binders and systems, it must be used in combination with other specialty sands. Due to the particle size of the synthetic mullite's 311 product (30 GFN), it is used in combination with existing specialty sands to achieve the desired casting surface finish. The 411 product (41 GFN) - which has the higher grain fineness for an improved surface finish - would be an option, but it absorbs too much of the binder at that level and loses its cost competitive advantage to zircon.

In foundry production tests in the last year, the 311 synthetic mullite has shown to work best for coremaking in conjunction with zircon sand at a 30% synthetic mullite/70% zircon sand ratio. The testing Ashland is performing on the synthetic mullite in sand core molds is still in preliminary stages. However, the product's reclamation capabilities, as described above, are encouraging. Figure 5 illustrates the synthetic mullite's green strength characteristics.

Environmentally, the product poses no hazards. They are compounds of alumina silicates that are chemically inert and pH neutral. In the crystalline phases, the 311 product is primarily made of mullite and beta-cristobalite and the 411 is primarily made of mullite and corundum. In addition, neither product has any quartz silica. The basic properties of the synthetic mullite are listed in Table 3.

At this point in development, the synthetic mullite's limitations in lost foam include cost perception and the higher energy requirement to fluidize the product in a sand cooler. For green sand applications, limitations include grain fineness and cost.

Presently, the synthetic mullite is being targeted as a replacement to high-end specialty sands, such as zircon. An aggregate of higher grain fineness also is being developed to pursue the broader market as a replacement for silica sand.

RELATED ARTICLE: Eagle Foundry Reclaims Ortho-silicate Successfully

Eagle Foundry Co., a 140-employee iron and steel shop in Eagle Creek, Oregon, produces 5-1500 lb manganese steel castings for the mining and rock crushing industries. Up until early 1997, due to the reactive nature of manganese steel, the foundry poured these castings on its sodium silicate olivine molding line. However, Eagle Foundry was looking to upgrade the consistency and reclaimability of this nobake line, so they turned to the ortho-silicate aggregate.

"The chemistry of olivine varies depending on what part of the country it was mined from. In some cases, we received inferior grades," said Jeff Bailey, vice president of technical services. "The ortho-silicate provides consistency because of how it is manufactured, so the characteristics of our molds remain constant."

The smelting process from which the aggregate is derived produces the same consistent by-product every time. The only variable for the aggregate is within the grain fineness range chosen by the foundry.

The other advantage the Eagle foundry received, said Bailey, was the cost reduction in terms of dry attrition reclamation and purchase. Eagle Foundry reduced its addition of new material, after the molds underwent reclamation, from 35%-15%, when it switched its nobake line to the aggregate. Bailey also said that the aggregate ran at 100% reclamation, with satisfactory results, longer than the olivine could.

In addition, the foundry saved cost because the aggregate was less dense than olivine. Since foundries purchase the material by weight and not volume, the comparable prices of the olivine and aggregate resulted in more molds per tons of media purchased for Eagle Foundry.

"The aggregate improved the consistency of our molds and reclamation practices and saved us costs," said Bailey.

RELATED ARTICLE: Citation Uses Synthetic Mullite in Its Large Lost Foam Molds

Citation Foam Casting Co., a 115-employee lost foam shop in Columbiana, Alabama, implemented Ceramacore 411 on its large horizontal lost foam casting line in September 1997. The synthetic mullite is the backing material for its 250-700 lb iron casting patterns.

The foundry had previously used olivine on this lost foam molding line, but the aggregate's flowability, durability and thermal expansion characteristics were attractive benefits that forced a change.

"The synthetic mullite is spherical, so in our vibration process it flows extremely well and is energy efficient," said Brian Harvey, research and development manager. "We were able to eliminate a great deal of molding energy."

The foundry also expressed notable improvements in dimensional precision and overall part quality. In addition, the durability of the synthetic mullite, said Harvey, is unchallenged. Although the large line at Citation is low volume, the foundry turns the sand over 3-4 times per day, and except for loss of the aggregate on shakeout, the foundry hasn't had to replace any lost sand with the synthetic mullite.

"For lost foam casting, it is the Utopia," said Harvey. "If you wanted to develop a molding media, it would have to be spherical, durable and thermally stable, and the aggregate is those things."