Fighting veining defects with sand additives.

Another sand additive option exists for foundrymen seeking to reduce certain sand-related defects in ferrous castings.

Sand additives have proven useful to foundries in reducing a number of casting defects. Additives can often be effective in minimizing metal penetration, veining, erosion, lustrous carbon, and subsurface porosity-type defects.

Materials like black and red iron oxides, clay and sugar blends, as well as wood flour can improve molding sand compositions to help reduce the occurrence of these defects.

Additive Options

Iron oxides, when used as a 1-3% addition to sand, reduce the incidence and severity of veining. Veining is evident primarily as metal fins on the casting surface - the result of metal flowing into cracks in the mold wall during the thermal expansion of sand. It is widely believed that at elevated pouring temperatures iron oxide softens the mold and core walls, making them more flexible [TABULAR DATA FOR TABLE 1 OMITTED] and less prone to cracking.

Iron oxide also helps reduce metal penetration defects, possibly because the iron oxide particles fill the spaces between sand grains and actually provide a physical barrier to metal penetration. In addition, iron oxides help lower the incidence of subsurface porosity in gray and ductile iron castings. Though it is not known exactly why this is, one theory is that the oxygen released by iron oxide reacts with atomic hydrogen and/or atomic nitrogen to prevent gas formation. This oxygen also helps control lustrous carbon defects.

However, iron oxides do not completely eliminate problems like veining, and hinder sand flowability while tending to build up on tooling. They also soften core and mold handling strengths to often undesirable levels.

Powdered starches and sugars, which are known as "cushion additives," are thought to reduce veining by lowering core density and burning out at the mold-metal interface, leaving enough room between sand grains for sand expansion to occur without cracking the wall. These materials tend to bum off and release oxygen when exposed to molten metal.

Because they do not expand like silica, the use of specialty sands such as zircon, chromite and olivine can eliminate expansion-related defects. Denser than silica, they can change the metal's solidification rate and promote higher dimensional accuracy. However, their costs are 20-50 times that of silica and they are usually only used in "value added" core and mold applications.

Unlike these materials, whose benefits became known only after trial and error, a recently developed sand additive has been designed to provide the same benefits, as well as some of those of core coatings. Introduced three years ago, this additive, known as Veinseal, is a blend of quartz, iron oxide, mullite, titanium dioxide and other materials.

When incorporated into core sand, the engineered sand additive (ESA) reacts with heat and pressure, fluxing to form a glazing action at the mold-metal interface, which helps reduce veining. It also reacts with the sand to produce a complex silica compound that slows silica expansion. While clay/sugar additives, starch compounds or iron oxides burn off or release oxygen during metal pouring, ESA displays a low loss on ignition (LOI) and actually contracts or remains in a steady state at metal-pouring temperatures. Its density is approximately 20% higher than that of silica sand.

Process Characteristics

Metalcasters should assess process characteristics when considering the use of any sand additive, looking at its impact on tensile strength, for example. Table 1 details the comparative effects of additives on sand tensile strengths. With a 5% addition of ESA to the sand mix, sand tensile strengths lose approximately 15-25% of the tensile attained with the control sand mix. This same addition, however, demonstrated slightly higher strength than those of a 2% black iron oxide mix. Foundries using it have experienced similar results with tensile strengths. Another consideration is that, as a slightly basic material with a typical AFS grain fineness of 180-200, the additive slows down the curing of some of the acid-catalyzed nobake systems.

The chemical and physical characteristics of the new material are such that, while it is about 20% denser than silica sand, it is much less dense than zircon or chromite specialty sands, as well as iron oxide additives. This comparative lack of density means it will not severely affect sand flowability, nor readily build up on tooling. A large ductile iron foundry using 5-10% blend additions, for example, has experienced no buildup problems in its high-speed coremaking system.

To evaluate the material's effectiveness and compare it to other additives, a series of tests was conducted to observe its performance in a laboratory environment where controls could be introduced.

Veining

Laboratory technicians made ductile iron step cone test castings from molds of phenolic urethane nobake binder and lake sand. The resin percentage of the molds was 1.25% based on sand (BOS) with a 55/45 Part I/Part II resin ratio. The cores were made from a phenolic urethane coldbox resin and silica sand with a resin percentage of 1.25% BOS and a 55/45 resin ratio.

The first set of cores contained no additives and was used as a control. The second contained a 5% mixture of ESA with the same amount of binder. A third set of cores contained no additives but was dipped in a ceramic water-based mullite coating. Two step cone castings with a similar sprue were then poured at temperatures from 2605-2626F (1429-1441C). Inoculation procedures consisted of a covered-ladle magnesium ferrosilicon treatment.

Table 2. Chemical Analysis of Ductile Iron Test Castings

Element (%)    Sample A      Sample B      Sample C

C                3.71          3.74          3.66
Si               2.09          1.81          1.82
Mg               0.035         0.034         0.037
Mn               0.28          0.25          0.25
S                0.009         0.007         0.013
P                0.030         0.023         0.024
[T.sub.i]        0.010         0.012         0.057
[O.sub.2]        0.0016        0.0032        0.0061
[N.sub.2]        0.0070        0.0045        0.0053

KEY
A = Standard Phenolic Urethane Core
B = Phenolic Urethane Core, 5% Addition of ESA
C = Phenolic Urethane Core, Water-Based Mullite Coating

In the standard casting without any additive, veining was evident up to the sixth step, while the casting made from the refractory coating-dipped core showed veins up to the fourth step. The casting containing the newer material exhibited no veining [ILLUSTRATION FOR FIGURE 1 OMITTED].

Veining and Penetration

Gray iron test castings using a 2x2-in. penetration casting were employed to analyze veining and metal penetration [ILLUSTRATION FOR FIGURE 2 OMITTED]. All molds and cores for the tests were prepared as above.

The first set of cores was standard and used as a control. The second set incorporated a 5% ESA addition. The third set used a standard core dipped in an alcohol-based zircon coating, while the fourth used a 2% black iron oxide addition. The castings were poured with a charge made from pig iron and ferrosilicon ladle additions, at temperatures ranging from 2650-2675F (1454-1468C).

It was found that the standard phenolic urethane core showed the highest degree of metal penetration and veining, while both the core dipped in coating and the one with the black iron oxide addition displayed some degree of veining. The core with the new material produced no veining or metal penetration.

Erosion in Gray Iron

Technicians used an erosion wedge test casting to measure resistance to erosion in gray iron. All the molds were made with a phenolic urethane nobake resin at 1.25% BOS and a 55/45 ratio in lake sand.

The cores were made with an acrylic-epoxy resin system, using a sodium-dioxide catalyst - a system most prone to an erosion defect. The system used a 1.2% resin addition with silica sand. A standard core was used as the control, and a test set of cores was made with a 5% addition of the new material. Charge materials consisted of pig iron and ferrosilicon, and melting temperatures ranged from 2650-2675F (1454-1468C).

The casting made from the test cores showed an improvement in erosion resistance over the control casting. Due to the severity of the test, a clay/sugar blend additive is needed to completely eliminate erosion.

Lustrous Carbon in Gray Iron

Soot plate castings, which require no cores, were poured in gray iron to judge any effects on the lustrous carbon defect commonly associated with the phenolic urethane binder system. All molds for this evaluation were made with a phenolic urethane nobake binder system and 1.75% resin with lake sand at 55/45 ratio. A second set of molds was made with 5% additive.

Table 3. Image Analysis Results for Ductile Iron Test Casting

                 SAMPLE A            SAMPLE B           SAMPLE C
             Edge        Core    Edge         Core   Edge       Core
% Graphite   12.0        11.2    11.6         10.8   11.5        9.8
% Pearlite    4.2         8.2     0.0          6.1    9.8        6.6
% Ferrite    83.8        80.6    88.4         83.1   78.7       83.6
Ferrite/
Pearlite     20:1        10:1    [infinity]   14:1    8:1       13:1
Ratio

KEY
Sample A = Standard Phenolic Urethane Core
Sample B = Core with 5% Addition of ESA
Sample C = Core with Water-Based Mullite Coating
Edge = Analysis Performed on Core/Metal Interface Edge
Core = Analysis Performed on Area Approximately 1/2 in. from Edge of
Casting

The molds were poured with pig iron and ferrosilicon at temperatures of 2540-2560F (1393-1404C). These lower metal temperatures were used to increase the likelihood of creating the defect.

The control casting exhibited dark, sooty deposits on the back side of the casting, while the test casting had very few deposits on its surface [ILLUSTRATION FOR FIGURE 3 OMITTED]. After cleaning the castings with a wire brush, the test casting showed fewer shiny, airborne particulates than the control casting (castings with a large amount of lustrous carbon on their surfaces usually release a lot of these types of particulates).

Iron oxide has been shown to decrease lustrous carbon defects because it releases oxygen at metal pouring temperatures. The oxygen combines with the carbonaceous decomposition products evolving from the binder system, forming carbon dioxide. However, since the new additive releases little oxygen at metal pouring temperatures, the mechanism for lustrous carbon prevention is not fully understood.

Ductile Core/Metal Interface

Chemical and image analyses were performed on three ductile iron step cone test castings to evaluate the new material.

Chemical Analysis - The chemical analysis results are shown in Table 2. Comparing the three castings, titanium, oxygen and nitrogen show the most disparity at the core/metal interface. The titanium and oxygen levels for the casting with the coated core are much higher than those in the other two samples. The nitrogen levels are appreciably lower for the castings containing the mixture and the coated core. A possible explanation may be that the cores in these cases are forming a barrier to gas evolution, thereby redirecting the gas away from the core/metal interface.

The sample containing the mixture did not show an appreciable difference in oxygen (about 16 parts per million) when compared to the standard phenolic urethane core. The result could indicate that little oxygen evolves from the material.

Image Analysis - The initial image analysis results are listed in Table 3. Image analysis was performed on the core/metal interface at the edge of the casting and at one-half in. from the edge of the casting. The most interesting result from this study showed that the casting with the new additive did not contain any pearlite at the core/metal interface. Some elements that affect pearlite formation in ductile iron include phosphorous, chromium, copper and tin. Because a pearlitic mix is enhanced by rapid solidification due to an increase in cooling through the lower critical ranges, it is expected that the absence of pearlite at the edge of the sample is due to a cooling rate reaction. An explanation of this effect may be ESA's fluxing action, which could promote slower solidification.

Gray Iron Subsurface Porosity

The step cone castings used in this test were made with a phenolic urethane nobake binder at 1.25% and a 55/45 ratio with lake sand. Cores were made with phenolic urethane coldbox resin at 2% BOS and 90 silica sand for low permeability. The resin ratio was altered to 40/60 to exaggerate the effects of the nitrogen contained in the isocyanate Part II resin. The cores were not vented. The second set of cores was made with the same specifications, except that it contained a 5% ESA addition.

The castings were poured in a gray iron foundry with an induction furnace at temperatures from 2690-2710F (1477-1489C). The high pouring temperatures were used because nitrogen and hydrogen gases are more soluble in the higher-temperature metal.

The castings made with the offset resin ratio and the high grain fineness sand with no venting showed subsurface porosity upon sectioning. In contrast, the castings made with the sand mixture showed no subsurface porosity upon sectioning.

The results of all these tests, bolstered by similar findings in field tests at production iron foundries, show the new additive to be a viable tool in reducing the above-mentioned casting defects.