How green sand systems are impacted by core sand dilution.

As more core sands enter your green sand system, an understanding of their rebonding characteristics can help you combat brittle sand woes.

Over the years, new binder systems have been developed that have advanced coremaking technology. Not only do today's binders cure faster, but they also collapse faster, enabling the casting of thinner wall components. This faster collapsibility, however, has presented foundries with new problems as higher levels of core sand make their way into the green sand system.

The problem, as seen on the shop floor, is the difficulty in rebonding these core sands. As a result, molding sands can become friable and brittle. This concern has increased further with higher core sand-to-molding ratios and the ever-shortening of available mulling times in high-speed molding sands.

Brittle sand can't maintain sharp mold edges and is prone to sand inclusions caused by the mold being washed away by the metal stream. Broken molds and poor draw in deep pockets are other problems encountered with brittle sands. The addition of high quantities of core sands to molding sand seems to have several effects:

* The increased core sand going into the green sand system results in higher clay bond requirements to maintain the active clay level in the system. This clay addition requires increased mulling time, which can not be accomplished effectively in many high production sand systems.

* The distillates of various binders can have a chemically based detrimental effect on the clay bonding.

* The production of lustrous carbon also can contribute to the problem by physically interfering with rebonding. Thus, when a phenolic urethane coldbox binder collapses readily and most of the core sand ends up in the green sand system, it produces a distillate that can be chemically harmful, and the lustrous carbon formed can leave a residual layer that hinders rebonding.

While varying opinions exist on how dilution affects green sand, this article attempts to provide an overview of several studies and present sources for further review.

Increased Clay & Mulling Times

Clay and other additives must be added to bring core sand grains up to system clay levels. As core sand dilution increases, the level of rebond additives necessary exceeds the levels needed to replace the clay destroyed by the heat of the metal. New rebond clays are more difficult to mull than the clays already in the return sand that have had the benefit of cumulative mulling. This excessive influx of new material and the respective absence of cumulative mulling are one cause of brittle sand.

The type and amount of core binder affect the green properties developed in the system sand. To combat the effects of core sand dilution, some foundries add new sand to their green sand system even if the core sand dilution is greater than the amount needed to maintain the system sand at a constant volume and cleanliness level. They feel that the new sand grain surfaces can be rebonded more efficiently than the surfaces of the core sand grains, which may be coated with residuals that hinder rebonding. Meanwhile, other foundries reduce or eliminate new sand additions if the core sand dilution is great enough to maintain the molding sand system at a constant volume and cleanliness level. The thought process behind this practice is to attempt to minimize new material additions and maximize cumulative mulling.

Core Sand Distillates

In his 1977 research, "Effect of Coldbox, Hotbox and Corning Cores on the Properties of Bentonite Bonded Molding Sands," D. Boenisch, Aachen Univ. of Technology, Aachen, Germany, conducted extensive work on the chemical effects of condensates that occur when the pyrolysis products of core binders, seacoal and organic materials condense in the mold.

He theorized that the distillates or condensates from the core sand make it difficult to develop bond from the clay. His work showed deterioration in some of the green sand that had been penetrated by the distillates. He found that the coldbox sands were the most destructive of the core binder systems, followed by hotbox and shell. The degradation was most evident in the reduction of green and wet tensile strengths.

Boenisch also studied the effects of salts on the activation of bentonite in molding sand mixes. The salts, present in some water, lower the wet tensile strength of the molding sand by electrolytic interaction with sodium bentonite. The higher the solubility of the salt, the greater the reduction in wet tensile. Many of the organic binders yield varying amounts of metal salts in the burnt return sand and these salts remain in the system sand. The levels of salt additions used in this study, however, were higher than could be expected for a recirculating green sand system. Due to core input, the sand volume in the system will climb and sand must be discarded periodically. The continual addition of new material, with no salts present, will prevent the accumulation of salts to an appreciable level.

Reclaimed Core Sand

In a report on "The Effect of Reclaimed Core Sand from Redset and Betaset Bonded Cores on the Properties of Green Sand," P. Naystroem studied the effects of reclaimed core sand on green sand properties. An acidic coldbox binder and an alkaline coldbox binder were compared to a commercial hotbox binder. The hotbox sand and alkaline shakeout sand did not affect the wet tensile strength. The acidic coldbox sand, however, resulted in considerable reduction in wet tensile strengths. Addition of soda ash ([Na.sub.2]C[O.sub.3]) to overcome the acidity was ineffective. Unburned core butts, whether acidic or alkaline, adversely affected the wet tensile strength.

Core Sand Blends for New Sand Additions

In their 1979 research on "The Influence of Nobake Core Sand Contamination on the Properties of Green Molding Sand," R. Naro, J. Plummer and T.J. Zeh experimented with core sand additions as new green sand replacements. They found nobake sand could replace up to 50% of the new sand and still not appreciably affect green sand properties. Because they used raw, particulated core sand that had not been exposed to heat, the rebonding results were somewhat inconclusive because the residual coatings of heat-exposed core sand can affect rebonding. Also, the tests used to evaluate the effects on green sand were green compression, green shear, compactibility and permeability, leading to some question of the physical effects on rebonding, as these tests are not the most indicative of effects of core dilution. The residual acidity or basicity of the nobake sands after curing had an insignificant effect on bentonite bonding. No correlation between residual pH and green properties was found.

There have been instances where foundries using acid-catalyzed core processes have needed to add either a great amount of new sand to make up for the acidity or add soda ash to the return sand for rebonding. Years ago, a paper by B. Booth showed that acidic sand does not produce the same green sand properties that are produced by basic sand.

In their 1982 research on "Particulated Core Scrap as a Replacement for New Sharp Sand Additions to a Green Molding Sand Line," A.L. Graham and G. Good found 15% particulated sand from phenolic hotbox and phenolic urethane coldbox cores could be blended with new sand to produce phenolic hotbox cores that exhibit only slightly lower core strengths. Particulated phenolic hotbox and coldbox core sands can replace new sand with minimal effects on properties, and these properties could be brought to standard through prolonged mulling. Adding sodium carbonate also can minimize the effects on properties. The study also showed the effects over time and the effect on casting quality.

Lustrous Carbon

Lustrous carbon is another material that has been blamed for the difficulty of rebonding clay sands. The 1975 work (prior to the use of many of today's core binders) of I. Bindernagel, A. Kolorz and K. Orths, "Controlled Additions of Hydrocarbon Components to Molding Sand Mixtures Improve Casting Surface Finish," addressed hydrocarbon additions to molding sand and their effect on casting finish. As the molding sand is heated, properties and the casting surface results are influenced by: amount of volatiles, lustrous carbon deposited from the gas phase and the softening characteristics of the hydrocarbon additive. Wood flour evolved the largest volume of gas and at the fastest rate. The paper also addressed the softening characteristics of the additives.

The gases that escape by the dry distillation of additives in the absence of air are volatiles. Depending on the composition of the gases, varying amounts of lustrous carbon are produced in a reducing atmosphere by distillation. The authors defined lustrous carbon as a shiny, microcrystalline carbon that is separated from gas phase and on to surfaces at temperatures of 1170-1830F (632-999C)

The paper described the lustrous carbon effects from additives such as seacoal, wood flour, peat, brown coal and synthetic resins and discussed the use of poor and rich hydrocarbon materials and their technological effects. The lustrous carbon on the mold surface sand grains can improve the casting surface finish, but it also can negatively impact the rebonding of the sand grains.

In 1977, R.L. Naro and D. Tenaglia wrote "Formation and Control of Lustrous Carbon Surface Defects," which examined defects in iron castings made in phenolic bonded test molds. Variables studied included pouring temperature, pouring time, binder level, binder ratio, sand types, additives, core coatings and baking. Lustrous carbon defects usually appear as areas of wrinkled, shiny surfaces that often resemble cold shuts or seams.

As the metal enters the mold, the thermal degradation of the binder produces hydrocarbon gases. The hydrocarbon gases recondense as pyrolitic graphite deposits as films or sheets on the mold surface.

Lustrous carbon deposits and defects were not present on castings poured at temperatures of 2700F (1482C) or higher, and rapid filling tended to minimize the amount of carbonaceous films. Binder levels - even in extreme ranges of 3% - did not affect casting results as long as they were poured at 2700F or higher. Excess Part 2 binder promoted increased amounts of carbonaceous residue.

At lower pouring temperatures, lake sand molds produced more defects than did high-purity silica sands. At higher temperatures, both sands produced defect-free castings with minimal lustrous carbon films. A 2% addition of hematite or yellow ochre was required to eliminate lustrous carbon films and defects. When core or mold coatings were used, the hydrocarbon gases and residues generated from the binder decomposition during pouring appear to diffuse through the coating and lustrous carbon defects still occur. Test molds baked at 425F (218C) for 2 hr showed no improvement over unbaked molds; however, molds baked at 400F (260C) for 2 hr were free of lustrous carbon wrinkling.

Phenolic Urethane Coldbox Effects on Sand Properties

In 1990, L.J. Pedicini and M.B. Krysiak wrote the definitive paper on coldbox core dilution in green sand - "Coldbox Core Dilution in Green Sand: An Analysis of Effects on Casting Quality." They studied the effects of phenolic urethane coldbox binders on green sand properties and used core sand samples at shakeout - sand exposed to heat.

Additions of 15% each of new sand and core sands in different states of thermal decomposition were added to molding sands. The effect of these additions on green sand properties was studied. The levels of new sand additions had no effect on properties when the rebond additions of clay brought the methylene blue clay back to its original level. Core sand additions then were evaluated at 20%, 40% and 60% at different states of decomposition - yellow, particulated sand with no thermal degradation; brown sand with partial thermal degradation but no lustrous carbon residue; and gray sand with thermal degradation and a layer of lustrous carbon. The molding sand was mulled for 4 min and 8 min. This series of tests showed a dramatic reduction in cone jolt toughness and an increase in friability.

In particulated sand grains, where the grain-to-grain bond is broken and the sand is coated with a clean layer of resin unexposed to thermal degradation (yellowsand), it appears that clay can bond just as well to resin coating as it can to new or calcined sands. Similarly, the brown sand - exposed to heat in an oxidizing atmosphere in which it was allowed to burn clean and left no lustrous carbon coating - also rebonded well. The gray core sand, with the least thermal degradation and a loss on ignition of 0.43%, was coated with a layer of lustrous carbon.

Whether the core sand is a product of a reducing or oxidizing atmosphere determines whether the residual coating is of a lustrous carbon type. They noted that the phenolic urethane coldbox sand that is not completely calcined in a reducing atmosphere features a residual graphitic coating and negatively impacts green properties. They advised that if the returning core sand does not have residual lustrous carbon and does not interfere with the clay bonding, the core sand influx should be treated as new sand. If it does have a layer of lustrous carbon and affects the rebondings, steps must be taken to counteract the effects. They provided suggestions for preventing or reducing the negative effects of core sand dilution - that can apply to any binder system. Among them are:

* don't use bulky cores;

* avoid binders that form lustrous carbon;

* reduce binder content in cores;

* vent cores and molds adequately;

* pour hotter and quicker;

* do not return large, partially calcined chunks of used cores;

* avoid returning core sand that has cooled in molds overnight;

* increase mulling time;

* mull at a higher compactibility;

* increase western bentonite levels.

They also pointed out that improper moisture-to-clay ratios also can produce brittle sand. Low clay, low moisture, hot sand, lumps in return sand and high new material inputs all cause brittle sand. If a great amount of new material is added on each cycle, a great amount of system sand must be discarded. The cumulative mulling that takes place in normal new material additions does not take effect because of the great turnover rate of the system.

Phenolic Ester and Phenolic Urethane Sands' Influence

In their 1989 research, "Influence of Phenolic Ester and Phenolic Urethane Bonded Core Sand on Green Molding Properties and their Effect on Casting Quality," R. Iyer, C.K. Johnson and M.J. Granlund showed the effect of core additions to molding sand. The research, which involved the pouring of castings for a number of heats, showed the effects of the two binder systems. Cores of these two binders were used in two sands and green sand cores were used in the third set of mixes.

The green compression strength and cone jolt toughness increased with the number of heats poured, but the rate of increase in each of the sands containing the cores made from the chemically bonded sand was not as great as the increase in the straight clay-bonded sand. This means degradation was occurring, however, in this case, the degradation appeared to have been overcome with the bond addition and the remulling. The mulling time was 4 min after water addition, and this longer mulling time may account for the excellent results, even after 10 heats. The castings did not decrease in quality with increased core dilution.

Pouring 10 heats using a sand system of 320-320 lb is not the same as a foundry sand system. The system in this research did hit a higher percentage of core dilution than can be possible in an operating foundry sand system, but not at a large sand mass.

Other Causes of Brittleness

High core sand dilution is not the only cause of brittleness in molding sand. Other factors include: excessive new sand additions, low clay content, low compactibility, hot sand, low sand level in the system or sand system capacity too small for hourly mulling rate, which would require large additions of new material, and removal of fines from the return sand.