Avoiding gas defects through mold and core package design: results from a study on the gas evolution of certain mold and binder formulations can aid metalcasters in predicting the potential for gas defects during the pouring of aluminum castings.

Voids associated with gas from cores, molds, binders and additives tan be difficult to detect and often are not found until the castings are cleaned, machined and inspected. Computer programs have been developed to minimize turbulence by predicting metal flow, but bubbles, folds, oxides and voids produced by mold and core gases during pouring can affect porosity as well.

Metalcasters historically have used their experience to design mold and core packages in a "pour and pray" method because little data existed to help select a binder, coating or core vent geometry to predict or eliminate gas porosity.

The central issues associated with mold/core gas evolution and defect formation are:

1. What is the volume of gas produced when molten metal hits a mold or core surface prepared with a particular binder?

2. How easily does the gas escape through core prints and mold?

3. Does the gas blow through the metal?

4. If gas is blowing through the metal, how/where should the core or mold be vented or what binder/sand/additive/coating changes should be made to eliminate the defect?

Techniques are being developed at the Univ. of Alabama at Birmingham to measure properties of sand that directly affect defect formation in aluminum castings, including:

* gas flow coefficients through molded sand;

* gas volume and rate of evolution when metal contacts molds and cores;

* sand properties, including effects of mold and core density.

In this article, two modern binder systems, phenolic urethane coldbox (PUCB) and epoxy acrylic coldbox, are compared to one green sand formulation. In addition, two common sand additives and one coating are evaluated to help metalcasters understand how sand, sand binder and coatings affect gas porostiy in castings.

Room for Improvement

Improvements in predicting and eliminating gas defect formation were needed in two areas. First, most investigators have measured permeability coefficients at significantly lower pressures than those existing in metalcasting molds. The low pressure measurements overestimate the ability of gas to flow through the sand.

Second, most investigators have used the total volume of gas evolved to predict pressures at the mold-metal interface. The rate of gas evolution and the time at which a solid skin forms have been ignored.

In order to predict and prevent gas defects, the following four items should be considered:

* gas evolution rates under conditions experienced in molds;

* accurate expressions for gas flow through sand to predict pressure at the metal interface;

* solidification models to determine when a solid skin has formed;

* the true effects of coatings.

Dozens of coating formulations on the market use water, alcohol and hydrocarbon carriers. The effects of the carrier and binder pyrolysis, as well as incompletely removed solvents, must be known.

Setting the Stage

Sand cores were prepared as solid cylinders with a diameter of 1.125 x 8 in. (2.86 x 20.32 cm). Two inch-long specimens were cut from these cores and immersed in molten A356 aluminum at 1,350F (730C), and the rates of gas evolution were determined.

In the core immersion apparatus, hot gases formed from binder pyrolysis flowed through a preheated line and displaced oil in a preheated chamber. The displaced oil flowed from the chamber into a container placed on a precision electronic balance, and the weight of the displaced oil was measured as a function of time. From the oil weight and density, the volume of gas was calculated.

Temperatures also were measured at several points to be sure the piping and oil chamber were hot enough to prevent water condensation. Thermocouples were put in the line connecting the specimen to the oil chamber, and the temperatures of the gas did not change during immersion. This was a result of the fact that it takes 50 seconds for the center of a dry core to reach 330F (165C) as a consequence of the low sand conductivity. When binders pyrolyze and mass transfer accelerates the heat transfer, the gases are heated somewhat faster but still require 40 seconds to reach the wand temperature of 330F.

The oil chamber was opened from time to time, and no condensate was found. The core and mold systems included in the study are shown in Table 1.

Battle of Binders

The gas evolution rates for two uncoated cores bonded with phenolic urethane and epoxy acrylic were highest in the range of 30-40 seconds after contact with molten metal. In some cases, two humps appeared that could be associated with solvent evaporation and condensation and re-evaporation of binder pyrolysis products. Not much difference was found between the total volume of gas evolved by the baseline phenolic urethane and epoxy acrylic binders. With the increased difficulty associated with obtaining components for manufacturing phenolic urethane binders, the data suggests that a metalcasting facility might switch to acrylic binders without affecting the total volume of gases produced at the mold interface and the tendency of gas defects to form.

The effect of an 8% AV-VS additive on gas evolution from a PUCB core is shown in Fig. 3. The use of the additive doubled the total gas produced and tripled the peak rate of gas evolution. The rate of gas evolution was about 6 cu. cm/second in 10 seconds and up to 10 cu. cm/second after 30 seconds of metal contact. This is a critical period when many nonferrous molds have just been filled. During this time, gas would be evolved at the maximum rate to produce blows.

[FIGURE 3 OMITTED]

Coating Conundrum

Coatings are used to improve surface quality and hopefully prevent mold and core gas from entering the casting. However, the porous structure of cores allows the carrier in the coating to enter pores, and substantial drying time may be required for solvent evaporation. This problem is worse in areas of relatively low core density, where the volatiles can penetrate much more deeply into the core. This effect of lower core density of gas evolution is shown in Figs. 4 and 5, where coated cores having a lower density always produced higher gas volumes at higher rates. It was clear that good sand compaction is an important parameter affecting the amount of gas produced during pouring.

[FIGURE 4 OMITTED]

Green Sand's Bubbles

The gas volume and evolution rate from green sand containing 8% clay and 4% water in contact with aluminum is shown in Fig. 6. The green sand produced almost 10 times as much as the base organic binders and four times the gas at four times the rate of the coated, organically-bonded samples. Furthermore, bubbling of the molten aluminum around the samples during immersion of several of the green sand cores almost certainly would produce gas defects in green sand Castings. Green sand must have relatively high permeability to prevent this bubbling during the casting. Bubbling was not observed in any of the other sand core systems examined in this study.

[FIGURE 6 OMITTED]

This article was adapted from the 2006 AFS Transactions paper (06-130) presented at the 110th Metalcasting Congress.

For More Information

"It's time to Play Name That Green Sand Casting Defect," I. Kay, M. Nagel and A.T. Spada, MODERN CASTING, September 2001, p. 29.

Preston Scarber and John Griffin are research engineers for the Univ. of Alabama at Birmingham, Ala. (UAB). Charles Bates is a retired research professor from UAB.

Preston Scarber Jr., Charles Bates and John Griffin, University of Alabama, Birmingham, Alabama 

Table 1. Core and Mold Systems Examined

Uncoated                             Al-SiO-Based Coating

Lake sand (GFN 45-55); 1.63% PUCB;   Lake sand (GFN 45-44); 1.63% PUCB;
no additives                         no additives

Lake sand (GFN 45-55); 1.6%          Lake sand (GFN 45-55); 1.6%
epoxy acrylic; 0.05% AV-M additive   epoxy acrylic; 0.05% AV-M additive

Lake sand (GFN 45-55); 1.6% PUCB;
8% AV-VS additive

Green sand; 8% clay and 4% water