Core binder systems offer an array of choices.

Core Binder Systems Offer an Array of Choices

Cores are an integral part of the casting process, often serving a number of different functions simultaneously. Placed in a mold, they delineate the internal shape of the casting. Cores can also be used to, in effect, create the mold, i.e. an all-core mold. But in sand casting, they can also add strength, and are essential when a pattern contains back draft or projections which cannot be molded. Cores have also been used either to create the gating system (slab cores) or act as filters (pencil cores). Ram-up cores are so-named because they become part of the mold face when it is "rammed up."

Core Materials

Cores can be made of either metal; ceramic; plaster; compressed green or chemically bonded sand. Metal cores, used in permanent molding and die casting, are sprayed with a release agent and mechanically retracted or "pulled" after the casting has solidified. But metal cores are obviously limited to shapes which can be pulled--limitations not shared by collapsible sand cores.

Ceramic and plaster cores are collapsible, and are used where an exceptional casting finish is desired. Ceramic preformed cores are more commonly used in investment casting; poured ceramic cores are used in sand casting.

Coremaking

Cores can be formed as a single unit in a wood, metal or plastic pattern termed a corebox. They can be made one at a time, or in multiples. A single ceramic core is made from the corebox.

Wherever possible, use of small core pieces are avoided. Instead, a central core is designed so that "loose pieces" can be affixed without shifting when the core is placed in the mold.

High-volume core production requires specialized equipment to mix and coat the sand, compact it in the corebox, dry the cores (as needed) and transport and stack the cores for final curing. Complex cores can also be made in pieces and assembled using hot-melt adhesives.

Today, only cores for prototype or short-run parts are produced manually. For decades, the vast majority of cores have been machine produced using either a jolt table, shell core, sand slinger or core blower. With a jolt machine, the corebox filled with sand is raised and dropped against the machine base, compacting the sand around the pattern. Further compaction may be provided by pneumatic hand ramming. Completed cores are then dried and baked.

Shell cores are made by investing the coated sand on a heated platen. The corebox is clamped together above a sand hopper, rolled over, and a measured amount of sand is blown into the corebox. After the core is invested for 5-10 sec, the corebox is rolled back above the hopper. Unused sand flows back into the hopper, while the core cures for 15-30 sec.

A sand slinger uses the downward force of sand, propelled by rotating vanes, to achieve the required level of sand compaction in the corebox. Since a large volume of sand is delivered quickly, they are generally used for larger coreboxes.

Coreblowers date to the turn of the century, and their advantages for producing certain types of cores were quickly recognized. A metered amount of coated sand is blown in under carefully regulated air pressure to achieve the required core compaction. One of the main advantages of blown cores is that they have no parting line fins if the the parting lines match to within 0.005 in.

But a coreblowing system must also include a catalyst gas generator, a compressed air system (including a dryer) and a gas scrubber to treat and/or neutralize the exhaust vapors. A dryer is of particular importance because the moisture of the air in the system must be carefully controlled. Excess moisture will result in reduced binder strength and shortened shelf life. Various dryers are available to maintain an atmospheric dew point of -40F or less. Core Setting

Cores may be placed or "set" in the mold either manually or mechanically. They are secured in the mold cavity by special recesses called core prints. Chaplets--metal supports in a variety of shapes and sizes may also be used. During pouring, chaplets melt and become part of the casting. Their microstructure must be, if not similar, at least compatible with the metal cast.

Core Binders

The past thirty years have seen major advances in core binder technology. Until the 1960s. the shell process, core oil, various protein binders and (to a lesser extent) sodium silicates were the primary systems. Today, chemically-bonded resins are, as a group, the most widely used binders owing to their excellent repeatability, resistance to defects, and rapid curing. (Green sand cores are also used, however. See "High Pressure Green Sand Coremaking," modern casting, pp. 25-27 [Dec 1985].)

Modern core binders are expected to satisfy a number of performance requirements. Binders must provide green strength so that the core can be molded. The core sand bond must have enough integrity to withstand fracture from physical forces during assembly, and erosion and thermal stress during pouring. But if the bond is too strong, thermal expansion in the mold may cause the sand surface to crack, creating veining defects. At the same time, cores must be sufficiently porous to permit adequate venting of the core gases evolved during pouring. Porosity requirements are compounded if the core is used to vent the mold. Finally, the cores must be sufficiently friable to ensure relatively easy and complete shakeout.

Core Classification-Cores can be classified in a number of ways. They can be grouped according to their bonding material, i.e., ceramic, green sand, dry sand, chemically bonded. They can also be classified according to their position in the mold, i.e., horizontal, vertical, ram-up, etc.

A further distinction is made based on the organic or inorganic nature of the binder. Organic binders include cereals, resins and core oils. Cereal binders--corn and wheat flour, dextrine, starch--commonly used with core oils, provide both green and baked compression strength. Inorganic binders include the three types of clay (Western and Southern bentonite and fireclay), sodium silicates, cement, phosphates, and a number of patented chemical processes.

Some binders and metals are inherently incompatible. In the case of aluminum, for example, the pouring temperature may not be high enough to break down certain binders, inhibiting shakeout. If sodium silicate binders are used in cast iron production, a secondary bond may develop hindering shakeout.

Perhaps the most convenient classification of chemical binders is one based on curing temperature: nobake, heat-cured and coldbox. Since space does not permit a detailed discussion of all the processes within these three groups, their differences will be contrasted by focusing on one or two processes within each group.

Nobake Processes

"Nobake" is the singular term applied to self-setting, non-gassed processes that cure at room temperature. Included under this category are organic systems: furan-acid, phenolic-acid, ester-cured phenolic, alkyd urethane and phenolic urethane; and inorganic: sodium silicate ester-cured, cement bonded and phosphate. Most nobake systems use either an acid, base, or ester catalyst.

Furans offer an instructive contrast, since they are available as nobake, cold, warm and hotbox systems. (Furan is the generic term applied to furfuryl alcohol-based systems.) Introduced in the late '50s, furan nobakes quickly became (and have remained) one of the most widely used binder systems in jobbing foundries. It should come as no surprise, then, that they deliver excellent dimensional accuracy, resistance to surface defects, and excellent shakeout. The wide-spread use of furan binders is owed in part to the great variety available for gray iron, nonferrous, ductile iron, and steel production.

Coldbox Cores

"Coldbox" produced cores are made from binders which cure at room temperature using a gas catalyst. Phenolic urethane amine, furan-[SO.sub.2], acrylic epoxy-[SO.sub.2] sodium silicate-[CO.sub.2] and phenolic ester are all consider coldbox processes. Coldbox cores are primarily used with metals having low pouring temperatures--typically heavy-sectioned iron castings. Several variables determine the curing time for coldbox cores: * temperature of the sand-binder mixture; * type of catalyst; * percentage of binder in the sand mixture; * volume and velocity of air; * concentration of catalyst in the air-catalyst system; * binder formulation; * pressure in the corebox; * permeability of the sand.

In most coldbox processes, the sand is coated with a liquid resin which is then "cured" when exposed to a gas catalyst. Coreblowing equipment was readily adapted to the gas curing of cores--sodium silicate/[CO.sub.2] and the so-called "coldbox" systems. Cores could then be blown and gas cured on the same machine.

Sodium Silicates-While a gelled-silica process was patented in England in 1898, H. W. Dietert traced the first use of an alcohol hardened sodium silicate binder to tests made at Dow Chemical in 1943. In their earliest application, silicate binders were cured by baking. The use of [CO.sub.2] as a "hardener" was first tried in Europe, and a patent issued in England in 1948. By the mid-50s, the sodium silicate-[CO.sub.2] process had become the first practical coldbox system to be used in foundries. In the 1960s, silicate binders were used as "air set" systems, incorporating various reactive powder hardeners to cure the binder.

The Nishayama process, a Japanese development that uses a powdered ferro-silicon hardener, has proven popular in that country. However, problems associated with the inherent exothermic reaction and generation of hydrogen gas during curing have limited its widespread use in the rest of the world.

All inorganic, silicate-[CO.sub.2] binders cure in a two-step process. Introduction of [CO.sub.2] gas to th silicate-sand mixture results in the formation of a mild carbonic acid (from the interaction of [CO.sub.2] with water in the silicate binder). The pH falls, reducing the percentage of soda ([Na.sub.2O]), and effectively increasing the percentage of silica ([SiO.sub.2]) in the binder. Rapid gelation of the silicate binder occurs, further accelerated by the physical movement of the [CO.sub.2] gas through the binder-sand mixture, removing additional water, and increasing binder solids. During this gelation, approximately 30-40% of the ultimate strength is developed.

The silicate-[CO.sub.2] process can be controlled in a number of ways. Not only are different ratios of silica to sodium oxide (soda) available, but by varying the amount of water added, it is possible to significantly change the chemical and physical properties of the binder.

Initial tensile strength of the cores gassed for 5 sec varies from 37 to 45 psi (depending on the resin used), and increases to 100-200 psi after 24 hrs. Since part of this increase is due to the dehydration of unreacted silicates and continued gelling of the mixture, it is influenced by relative humidity.

Ultimate hardening occurs within 24 hr through a dehydration process known as "syneresis." The "glassy bond" which results accounts for the remaining 60-70% of the ultimate binder strength. In principal, any sodium silicate binder can be cured entirely by dehydration, but [CO.sub.2] gassing substantially decreases the cure time.

One of the most attractive aspects of these inorganic binders is their low toxicity. They are typically odorless resins, nonflammable and suitable for all types of production. Coreboxes can be made of any material. They produce no noxious gases during the mixing or coremaking operation, and only minimum volatile emissions during pouring, cooling and shakeout.

Improved as-gassed and ultimate strengths, as well as shakeout, have been achieved using the newest proprietary additives. These liquid co-binders have the added advantage of improved resistance to moisture pickup, lengthening storage times. Figure 7 compares the compressive strength of these new ester-silicate systems to conventional acetate ester-silicate binders. (For a more detailed discussion of these two-part binder systems, see "Environment Right for Silicates Rediscovery," modern casting, March 1989, pp. 23-26.)

Heat-Cured Systems

Hotbox, warmbox and oven cure systems include the furan hotbox, furan warmbox, shell, core oil (oven bake) and phenolic hotbox processes.

Hotbox Process-In this process sand is coated with a mixture of thermosetting binder and an acid salt catalyst before being blown into a vertically- or horizontally-parted corebox heated to 450-550F. The mixture starts to cure almost immediately upon contact with the hot pattern.

The hotbox process made its debut in nonferrous foundries in 1959. It won widespread acceptance in high-volume automotive foundries because of its rapid cure and high tensile strength. Cores with high tensile strength can be ejected within 20-40 sec. Unlike Shell core sand, hotbox sand is damp, and can be blown using conventional coreshooters. But because the core produced is solid, more mixed sand is used. Both the hot and warm processes use a moist binder-sand mixture.

Hotbox processes, introduced to nonferrous foundries in 1959, are classified either as furan or phenolic. Furfuryl alcohol (furan) used in conjuction with urea formaldehyde resins, supplanted the original unmodified urea formaldehyde resin. In the 1960s, phenol formaldehyde polymers were developed when a worldwide shortage of furfuryl alcohol gripped the industry. Today, owing to their higher hot strength and lower cost, fully 90% of the hotbox resin used is phenol-based. Formaldehyde, which has an extremely irritating odor, makes up 4-10% of conventional hotbox resins. Recently introduced low-nitrogen hotbox processes offer significant reductions in ammoniacal nitrogen and free formaldehyde but are more costly.

Reclamation of hotbox cores until very recently, has not met with much success, and most systems are typically run with 100% new sand. Thermally reclamation is an area of ongoing research and development, but little rebonding at this time is actually taking place.

Oil Sand Binders-Heat setting oil binders have their antecedents in the raw linseed oil cores first used in the late 1600s. Core oil is a generic term applied to a class of binders which use either natural oils or synthetic resins mixed with sand, water and cereal. The "green" cores are then cured in a hot, forced air oven at 400-500F, up to 1 1/2 hr for a typical core. Catalysts are commonly used to speed up the drying process.

Core oil cores can be blown or hand rammed in relatively simple, vented tooling. But oil-sand cores must be handled carefully while in the "green" state.

Shell Process-This unique process (also known as the "C" or Croning process after the developer, Johannes Croning) is a type of investment casting. Core "shells" are created by blowing or dumping heated sand coated with novolak resin into a corebox surrounding a heated (400-600F) pattern. Curing of the core sand begins closest to the pattern, continuing outward forming the shell. The cure cycle is ended by simply inverting the corebox, as described earlier.

The curing process is a complex transformation from a thermoplastic to a thermosetting state, and requires careful monitoring of the pattern temperature. A shell of from 1/8 to 3/16 in. (3-5 mm) normally develops during the approximately 1 min cure cycle. Both hollow and solid cores can be created using this process, and have excellent storage life.

While the basic shell process has remained unchanged, new additives and binder formulations have led to a safer process with improved sand flowability, reduced strip times, and nearly doubled shell strength.

Because of the problems with formaldehyde and phenol emissions inherent to the coating process, most foundries now purchase commercially coated sand. The benclife of shell sand is almost indefinite, and because it is heat setting, mullers and coreblower need not be cleaned daily.

Environmental problems, specifically formaldehyde and phenol emissions, have proven more difficult to solve. A number of foundries with low-volume production of the higher resin solid cores have switched to no-bake processes. But for those foundries who have made the investment in scrubbing and collecting equipment, Shell will probably remain a viable production process for some time.

Warmbox-The warmbox process was first introduced over a decade ago, and its popularity continues to grow. Curing results from the reaction of the binder resin with a heat-activated catalyst.

As the name implies, the process operates at a lower temperature--100F lower--than the hotbox system. But the two processes differ in many other characteristics. Unlike most hotbox processes which use phenol-based resins, warmbox binders use furfuryl alcohol resins with only 2.5% nitrogen content. Due to this lower nitrogen content, warmbox castings have significantly reduced or eliminated gas-related defects. Furfuryl alcohol resins have the advantage of faster initial cure, permitting quicker ejection. They also provide better shakeout and fewer disposal problems.

Warmbox catalysts are also unique--copper salts, primarily aromatic sulfonic acids in an aqueous methanol solution. Used in concentrations of 20-35% based on resin weight, they are unreactive at room temperature. Optimum sand temperature is 70-80F. To retain good benchlife at high sand temperatures (+90F), catalyst adjustments are required.

Warmbox sands should be used with sands having low acid demand value. Cores can be adversely affected by high relative humidity (+90%). Silanized release agents are available which lessen the effect of high humidity.

Venting

Thermal decomposition of the binder produces a reducing atmosphere in the mold cavity, inhibiting the formation of metallic oxides. These oxides act as a flux with the sand grains and permit metal penetration. The greater the reducing atmosphere in the mold cavity, the less likelihood of metal penetration and casting defects caused by adhering sand.

The binder's rate of thermal decomposition also determines the rate of gas evolution. In green sand molding, the gases generated from the moisture and sand binders must be vented away from the casting in order to prevent casting gas defects--termed core blows. This can be done by increasing the permeability of the mold sand, by scoring the core surfaces during assembly, or by actually drilling vents in the mold over the core prints. Cores may also be vented along the parting line of core halves. Oil-sand cores are often made with embedded wax wires, which, when the core is baked, form vent passages. Strands of wzx, nylon tubing or perforated metal tubing can also be used to create vents.

Release Agents

Coated sand which adheres to the tooling can increase core release or strip time. Rough coreboxes (with more irregularities where the sand can buildup) and dense cores, which may not cure completely, compound the problem. Any binder that builds up in a corebox must be removed using either blast abrasion or solvents.

A variety of release agents have been formulated to reduce the adhesion of the core to the tooling. They have been specifically formulated for use with different binders. Common agents include: oleic acid diluted with kerosene for use with core oils, and emulsified silicon for heat-cured binders (including Shell). Nobake and coldbox processes use a variety of formulations made from different chlorinated solvents, alcohol, silicons, powdered aluminum, graphite and vegetable oil.

Coatings

Core coatings or washes are used to seal the core surface, resulting in improved casting surface finish. Coatings are suspended particles which have a higher refractory value than the core sand and are smaller than the voids between the sand grains. Coatings reduce metal penetration and prevent veining, reducing cleaning and finishing operations. Less obvious savings result from reduced tooling wear. And because they are refractory, coatings promote chill. Certain coatings--selenium or tellurium paste--can also improve the surface hardness of castings.

The formulation and application of core coatings involves the design and control of five components: a refractory, carrier, dispersant, binder, and chemical modifier(s). There are over a dozen refractory materials available, and selection is based on the metal pouring temperature and casting seciton thickness. The coating must be compatible with the process: water-based washed are generally used with heat-cured cores, self-setting cores with an alcohol-based wash.

Coatings are essential in steel casting because the high pouring temperature is close to the fusion point of silica. Highly refractive coatings--zircon, magnesite, chromite or alumina--are commonly used to prevent surface defects. Iron foundries use a modified silica refractory because of its lower cost. Carbon is frequently added, producing a reducing atmosphere and increasing the fusion point, thereby lessening sand adhesion (burn-on).

Coatings are not intended to compensate for poor core surface. In fact, improperly applied or poorly mixed coatings can result in defects such as pinholing, cratering, orange peel and shrinkage cracks.

PHOTO : Photo of a cast iron corebox used to produce cores for a valve body. The corebox was

PHOTO : machined using CNC, and the close tolerances greatly reduce flash.

PHOTO : Pictured is an open axial corebox used to produce a plaster core. The metal vanes are

PHOTO : pulled from the rear.

PHOTO : Shown is a cheek core with loose pieces used to produce an aircraft gearbox.

PHOTO : Hot melt adhesive being applied to join smaller pieces to a large core.

PHOTO : The shell molding process involves attaching the pattern plate (a) to the dump box (b)

PHOTO : prior to investment (c)&(d) and ejection (d). The completed shell assembly isshown in (f).

PHOTO : A sand slinger shown discharging sand into a large corebox.

PHOTO : As shown, the faster curing alkylene-carbonate esters provide for longer worktime and

PHOTO : increased compressive strength than conventional acetate-ester sodium silicate binders. References Archibald, J. J. and R. L. Smith, "Resin Binder Processes," Metals Handbook, 9th ed, pp. 214-21, ASM International (1988). Carey, P. R., et. al., "Updating the Resin Binder Processes--Parts 1-IX," Foundry Management and Technology (Feb 1986). Chemically Bonded Cores & Molds, American Foundrymen's Society, Inc (1987). Dietert, H. W., Foundry Core Practice, 3rd ed, American Foundrymen's Society (1986). Heine, R. W., C. R. Loper, Jr. and P. C. Rosenthal, Principles of Metal Casting, McGraw-Hill Book Co (1967). LaRue, James P., Basic Metalcasting, American Foundrymen's Society, Inc (1989). Penko, Tom, "Rediscovering Sodium Silicates," modern casting, vol 79, No. 3, pp. 23-26 (March 1989). Sylvia, J. G., Cast Metals Technology, Addison-Wesley (1972). Wile, L. E., K. Strausbaugh, J. J. Archibald and R. L. Smith, "Coremaking," Metals Handbook, 9th ed, pp. 238-41, ASM International (1988).