Variables Affecting Gas-Evolution Rates from Cores in Contact with Aluminum

Aug. 21, 2007
A procedure has been developed to measure accurately the volume and rate of gas evolution from cores in contact with molten metal. The core-metal contact area and metal temperature largely control the rate of gas evolution, but the core age and core modul


Figure 1. Schematic of the displacement apparatus used to determine gas-evolution volumes and rates during contact with molten metal.

Figure 2. Effects of core length on gas volume and evolution rate per unit area from cylindrical cores made with PUCB resin in contact with aluminum at 1,350°F.

Figure 3. Effects of core length on gas-evolution rate and volume curves when both are normalized by unit area wet by metal.

Figure 4. Effects of core modulus on gas-evolution rates and gas volume from cylindrical cores in contact with aluminum at 1,250°F.

Figure 5. Effects of core modulus on gas-evolution rates and total gas volume evolved from cylindrical cores in contact with aluminum at 1,500 °F.

Figure 6. Effects of core modulus on gas-evolution rates and total gas volume from rectangular cores in contact with aluminum at 1,250°F.

Figure 7. Effects of PUCB core age on gas-evolution rates and volumes.

Figure 8. Effects of sand GFN on gas-evolution rate and volume.

In the past 15 years, developments in molding and core-making technology have allowed foundries to cast parts with much greater intricacy and better dimensional accuracy than ever before. More intricate castings generally have larger core surface areas in contact with molten metal during casting. Venting the cores becomes increasingly challenging as the complexity of the castings increases. And, the difficulty of venting cores increases the tendency for gas defect formation.

To understand and reduce the incidence of gas defects in cored castings, improvements are needed in three areas: 1) better measurement of sand permeability; 2) reliable data on the rate and type of gas evolved; and, 3) methods for calculating gas pressures inside cores.

The effects of permeability on pressure inside cores has been extensively studied1, 2. Based on data developed by Locke and Ashbrook1, gas pressure in molds poured with steel decreased exponentially with increases in core permeability. Naro et al.2 investigated the tendency for blowhole formation with changes in the core sand-grain fineness number (AFS-GFN), and found that the incidence of blows increased with the grain fineness number. It was also found that the permeability increased with a decrease in sand-grain fineness.

Levelink et al.3 measured the pressure inside cores using a probe in the sand and verified that the internal pressure decreased with an increase in permeability. However, it was difficult to quantify the effect of permeability because of the lack of data on the rate of gas generation and the properties of the gas evolved.

A technique was developed4 for measuring the permeability of sand with these factors in mind. The technique measures the ability of gas to flow through pores in cores, and separates the contributions of the type of gas, gas temperature, core geometry, and porosity on core permeability.

The total gas volume produced during loss-on-ignition (LOI) measurements has been used to predict pressures. However, the gas volumes cannot be converted directly into rates of evolution in molds and cores because LOI measurements are made at relatively low heating rates. Levelink et al.3 noted that the heat transfer rate to the core in the casting process was much higher than the heat transfer rate to core samples in a furnace. Hence, gasevolution rates produced using tube furnace procedures would be expected to be lower than rates in a mold or core.

Previous research5 has shown that the pressure developed inside cores increases with a larger core-metal contact area, distance to core print, and gas-evolution rate. The pressure decreases with an increase in core print area and sand permeability. However, the variables influencing the rate of gas generation during core-metal contact have not been documented.

Our research studied gas-evolution data from cores having various lengths, contact areas, and shapes to illustrate the effects of surface area and modulus. The effects of metal temperature, core permeability, and core age on gas-evolution rate and volume were studied, too. These data were used to develop a model that describes gas evolution and pressures developed in cores.

Experimental procedures
Cores made with Phenolic Urethane Cold Box (PUCB) and Phenolic Urethane No Bake (PUNB) resins, with lengths of 2 in. (5.1 cm) and 3.5 in. (8.9 cm), were immersed into aluminum at temperatures of 1,250°, 1,350°, or 1,500°F. Each core was printed into a steel holder to a depth of 0.5 in. (1.3 cm), resulting in 1.5 in. (3.8 cm), or 3 in. (7.6 cm) of the core in contact with molten metal. Core diameters were 1.125 in. (2.86 cm) and 2.4 in. (6.1 cm). Rectangular cores had edge lengths of 1 in. (2.54 cm) and 1.6 in. (4.1 cm).

The device used to immerse cores and measure the volume of gas produced is illustrated in Figure 1. When the core contacted the metal, high-temperature pyrolysis gases flowed through a preheated line connecting the specimen holder to a hot oil tank. The line and oil tank were kept hot in order to minimize moisture condensation. As the gas flowed into the oil tank, oil was displaced, and the oil overflowed the tank through a vent tube into a container located on a precision electronic balance. The weight of displaced oil was measured as a function of time, and the oil weight and density were used to calculate the gas volume that caused the oil to be displaced.

Temperatures were monitored at three locations in the line connecting the sample to the oil chamber to assure that all parts of the system were hot enough to prevent moisture condensation. The precipitation of one gram of water in the system would reduce the measured gas volume by over one liter.

Effects of core length and core metal contact area
Cores made with 2.5% Phenolic Urethane Cold Box (PUCB) binder and with a diameter of 1.125 in. (2.8575 cm) and lengths of 2 in. (5.08 cm) and 3.5 in. (8.89 cm) were immersed in aluminum at 1,350°F. The effects of length on gas-evolution rate and volume are illustrated in Figure 2. The 2 in. (5.08 cm) and 3.5 in. (8.89 cm) cores produced 120 cm3 and 200 cm3 of gas, respectively, in 60 seconds when measured at 1 atm and 250°C. There were two gas-evolution rate peaks, occurring at approximately 15 and 41 seconds after metal contact. The maximum rate peaks from 2 in. cores were 2.75 and 2.6 cm3/sec., and the total volumes after 60 seconds of contact were 120 cm3.

Higher rate peaks and volumes were observed from the 3.5 in. length cores. The maximum rate peaks were 3.75 and 5.2 cm3/sec., and the total volume was 200 cm3.

The increase in the volume of gas evolved was due to the increase in the core metal contact area. If the volume curves are divided by the metal contact area the volume curves fall on top of each other, as illustrated in Figure 3. The maximum rates of gas evolved from 2 in. and 3.5 in. length cores were almost the same at 0.07 cm3/cm2/sec., about 40 seconds after immersion. The volume of gas evolved at 60 seconds was about 2.8 cm3 of gas per cm2 of surface area.

Effects of modulus (specimen diameter) and metal temperature
Cylindrical cores made with 2.4% PUNB and having diameters of 1.125 in. (2.86 cm) and 2.4 in. (6.1 cm) were placed in contact with aluminum at 1,250°F and 1,500°F. The length of the core was held constant at 2 in. with 0.5 in. of the core printed in a holder. The modulus (volume to surface area ratio) of cores having a diameter of 1.125 in. was 0.8 (cm3/cm2) and the value for the 2.4 in. inch diameter core was 1.2 (cm3/cm2).

Cores with a modulus of 0.8 and 1.2 produced gas-evolution rates of 6 and 14.5 cubic centimeters per second, respectively, as illustrated in Figure 4. Three additional rate peaks were observed within 60 seconds from cores having a modulus of 0.8. However, only one additional rate peak was observed from cores with a modulus of 1.2. The volume of gas produced was higher with the larger surface area cores, with a value of about 420 cm3, compared to a value of about 240 cm3 from the lower-modulus cores.

The higher number of peaks in lower-modulus cores was probably due to the temperatures reached in the center of the core after a fixed amount of time. Higher temperatures along the core centerline caused re-vaporization of materials that had been condensed after partial pyrolysis near the surface.

Consequently, more peaks can be expected in lower-modulus cores, and there may be particular vapor species associated with the individual rate peaks.

Cores in contact with higher temperature metal produced higher gas-evolution rates and volumes, as illustrated in Figure 5. Cores with modulus values of 0.8 and 1.2 cm in contact with 1,500°F aluminum produced gas volumes of 400 and 640 cm3, respectively, compared to volumes of 240 and 420 cm3 at 1,250°F.

Three gas-evolution rate peaks were observed for 0.8 modulus cores. The peak rates increased from 5.5 to 10.5cm3/sec with longer metal contact time.

The number and magnitude of the rate peaks from cores with a modulus of 0.8 continued to increase, and the highest rate peak was observed after about 55 seconds. The continuous increase in gas-evolution rate peaks suggests that some volatile materials from the surface had condensed toward the center of the core and were reevaporated as the center of the low-modulus core was driven to higher temperatures. This conclusion was verified by observing temperature plateaus from thermocouples placed on the core centerline.

Gas from rectangular cores
Rectangular cores made with 2.4% PUNB, having a cross-section of 1X1 in. (2.6X2.6 cm) and 1.6X1.6 in. (4.1X4.1 cm), were placed in contact with aluminum at 1,250°F. The results are illustrated in Figure 6. The modulus values for the 1 in. and 1.6 in. square cores were 0.6 cm and 0.8 cm, respectively.

Again, more gas was evolved from the higher-modulus cores. The peak rate of gas evolution in the larger cores was 9 cm3/sec. after about five seconds, and the peak rate in the lower-modulus cores was approximately 6.5 cm3/sec. at the same time. One additional small gas-evolution peak was observed in the higher-modulus cores, and two more pronounced peaks were observed in the lower-modulus cores about 24 and 45 seconds after immersion. The total volumes of gas evolved in 60 seconds was about 200 and 280 cm3 from the cores with modulus values of 0.6 and 0.8, respectively.

Effects of core age
Core age can significantly affect the amount of gas that evolves, as illustrated in Figure 7. PUCB cores placed in contact with aluminum immediately after blowing produced higher rates and volumes. The peak rates obtained after one day and after 10 days were 35 and 6 cm3/sec., respectively. The gas volumes were about 300 and 200 cm3 after the same aging. The lower rates of gas evolution associated with aged cores was probably associated some solvent evaporation and with greater cross-linking of the resin. Although each core was aged in a sealed bag, still some volatiles might be lost with time. Binders with a higher degree of cross linking are also thought to need more energy to pyrolyze, and this might reduce the volume of gas produced. More information is needed to minimize speculation, and such an investigation should involve determining the gases evolved as a function of core age.

Effects of sand permeability
Core permeability did not affect the volume or rate of gas evolution, as illustrated in Figure 8. The gas-evolution rate and total volume evolved from PUCB cores made with 50 and 75 GFN were similar at 34 cm3/sec. and 290 cm3, respectively. The permeability generally decreases as sand-grain size number (GFN) increases. Although permeability does not affect the amount of gas evolved, it is still an important parameter because it influences the pressure developed in the core as the binder decomposes. Lower permeability cores develop higher internal pressures if other casting parameters are held constant.

Measurements, effects
The gas-evolution rates from cores were measured and found to be affected by metal contact area, core age, modulus, and metal temperature. The core-metal contact area was the most significant variable affecting heat transfer from the metal into the core. Cores in contact with higher-temperature metal also produced more gas. The gas-evolution rates were not affected by the core permeability.

These data are being used to develop a model to describe gas evolution from cores as a function of geometry and contact area with metal. The goal is to predict the gas pressure inside cores. If the gas pressure exceeds the metal head pressure, gas will be blown into the metal, resulting in hydrogen absorption, blow holes, pinholes, and oxide trails in aluminum castings. A model is being developed to take into account the chemical reactions, evaporation, condensation, and re-evaporation of volatile compounds in cores.


  1. Locke, C., Ashbrook, R.L., "Nature of Mold Cavity Gases", AFS Transactions, vol. 58, pp. 584-594 (1950)
  2. Naro, R.L., Pelfrey, R.L., "Gas Evolution of Synthetic Core Binders: Relationship to Casting Blowhole Defects", AFS Transactions, vol. 91, pp 365-376 (1983)
  3. Levelink, H.G., Julien, F.P.M.A, and De Man, H.C.J., "Gas Evolution in Molds and Cores as the Cause of Casting Defects", AFS International Cast Metals Journal, March, pp. 56-63 (1981)
  4. Winardi, L., Littleton, H.E., Bates, C.E., "New Technique for Measuring Permeability of Cores Made from Various Sands, Binders, Additives, and Coatings," AFS Transactions, vol. 113, pp 393-406 (2005)
  5. Winardi. L, Littleton, H.E, Bates, C.E., "Pressures in Sand Cores", to be published in AFS Transactions
  • Leonard Winardi is a graduate student, and Harry Littleton is a research supervisor, both with the Dept. of Materials Science and Engineering at the University of Alabama at Birmingham. Charles Bates is the founder and president of AlchemCast L.L.C.
  • The authors extend thanks to the members of the core gas consortium for guidance and financial support throughout this study. The authors also thank Dr. Preston Scarber Jr. for his invaluable insights.