Published in January 2007, the Implementation of Metal Casting Best Practices is a collaborative effort by Robert Eppich, Eppich Technologies, and Robert Naranjo, BCS Inc. Eppich coordinated the project, including guiding data collection and analysis. Naranjo assisted and prepared the report, with assistance from Lee Schultz, Rajita Majumdar, Bill Choate, Ellen Glover, and Krista Jones.


Metalcasting has a prominent place in the U.S. Dept. of Energy's Office of Energy and Efficiency and Renewable Energy, which funds process-focused research and development through its Industrial Technologies Program (ITP) Metalcasting Portfolio. In a partnership with the Cast Metals Coalition (a consortium of the American Foundry Society, North American Die Casting Assn., and the Steel Founders' Society of America), they examined cases where metalcasters implemented ITP research results. Two of the cases studied involved lost-foam casting operations.

Finding cost savings — The first case involved an aluminum foundry producing engine blocks and cylinder heads, with 100+ employees operating two shifts per day. Foundry No.1 ships approximately 12,000 tons of aluminum each year.

The spacious and highly automated plant was built in 2000, started up in 2002, and reached full production by 2003. It consists of four casting cells, with three reverb furnaces — two in operation and one used as a backup furnace. Aluminum is melted on site and delivered to each cell via a launder system.

Foam patterns are constructed and steamed on site. Robotic systems dip them in a ceramic slurry and hang them to dry. Robots transfer the dried patterns to a vibratory table that packs sand around them. Then, the pattern moves to an automated station where molten aluminum is poured into it.

Once poured, the casting continues to shakeout, and a robotic arm removes the single gate and riser. The sand is reclaimed (approximately 10% of the sand is disposed annually) and the removed gates and risers are collected and returned for remelting.

The cooled casting is inspected for quality, and some machining is done.

In the year prior to joining the ITP, Foundry No.1 averaged an internal scrap rate of 8%: this has dropped to approximately 5%, and customer returns at total 0.35%. Casting quality is monitored closely, and the facility can trace each casting back to its pouring sequence and melt batch.

Active in the Lost Foam Consortium — Metal Casting Consortium research is performed at the University of Alabama - Birmingham's (UAB) Lost Foam Technology Center, where R&D has focused advanced process controls, as well as on a coating system to improve foam-related defects.

Foundry No.1 is part of the Lost Foam Consortium — a group of casting producers, foundry suppliers, and the AFS, formed to develop a better understanding of the lostfoam process.

Foundry No.1 gained a better understanding of foam properties and coating systems used in the lost-foam process; 75-80% of all the scrap produced there was due to the foam. Research performed by the Lost Foam Consortium helped the foundry understand, control, and systematically reduce casting defects from foam problems.

The Consortium's research into high liquid expandable polystyrene (LEPS) saturation rates and high LEPS saturation capacity led to a 6.5% reduction in scrap for the foundries inline four-cylinder block castings. The scrap was caused by gas-porosity-related defects. Foundry No.1 produces 2,100 block and head castings per day, or 462,000 per year, at 220 days of production. The 6.5% scrap reduction represents a savings of 30,030 block castings per year (6.5% X 462,000 = 30,030).

Foundry No.1 engineers demonstrated the impact of this scrap reduction by calculating the effected energy savings. The facility uses 83,747,280 kWh of electricity and 1,158,670 million cubic feet (mcf) of natural gas per year to produce lost-foam block and head castings.

This computes to 2.86 X 1011 Btu of electricity (1 kWh = 3412 Btu) and 1.16 X 1,012 Btu of natural gas (1 mcf = 1 million Btu). This yields consumption of 618,497 Btu of electricity (2.86 X 1011 X 462,000) and 2,507,944 Btu of natural gas per casting, totaling 3,126,441 Btu per casting.

Calculating energy savings from scrap reduction, the total amount of annual scrap was multiplied by the Btu/year consumption for block and head casting (30,030 X 3,126,441 = 9.38 X 1010 Btu/year). This amount was multiplied by two-thirds (0.66) to correct for the metal differentials between a block casting and a head casting, yielding an annual savings of 6.26 X 1,010 Btu for the 6.5% reduction in block casting scrap.

Another instance of energy savings for Foundry No.1 followed its incorporation of work done by the Consortium on Expandable Polystyrene (EPS) beads, which lower viscosity of the liquid EPS decomposition product. This discovery yielded a 0.85% reduction in scrap due to porosity and leak (folds) defects for in-line six-cylinder blocks.

Participating in the Lost Foam Consortium's R&D has enabled Foundry No.1 to incorporate technological advances directly into its operations, bringing it substantial benefits as seen by its energy-savings calculation — benefit that will increase natural gas and electricity prices rise.

Planning for cost savings — Foundry No.2 is a new facility that produces A356 and A319 aluminum components for automotive, marine, power generation, and OEMs. It has expertise producing complex-geometry castings for niches like snowmobiles and lawn mowers. Low-cost, lost-foam castings for parts that traditionally are produced by machining or diecasting, were critical to helping it gain entry to these markets.

Foundry No.2 produces 2-3 million pounds of aluminum castings a year. Its scrap rate is 1-2%, primarily due to defects in the foam pattern (e.g., dents). The plant operates one daily 8hour shift for pouring, and runs the pattern shop 24 hours per day, 4-5 days per week, to keep up with the demand.

The plant uses bead-blowing and moldmaking machinery to fabricate patterns on site. The process begins with pre-expansion of polystyrene beads to a controlled density, following which the beads stabilize and reach a particular pentane level. Next, the beads are blown into an aluminum mold to create a foam pattern. The blown beads are heated and steamed to expand and fuse them, and the mold is cooled to stabilize the expanded beads that set the pattern. The pattern is hung in a controlled atmosphere for up to six hours to shrink it to a stable controlled size. Pattern segments are glued together using a special adhesive, and multiple patterns are assembled into a tree or cluster.

The assembled patterns are dipped into refractory slurry to ensure a uniform evaporation of the pattern during casting, and to prevent penetration of the foam by the molten aluminum. Once dipped, the patterns are oven-dried, and placed in a flask and packed with mullite sand (using vibration.) The flask is moved onto the automatic pouring line, and molten aluminum is poured directly into the foam cluster displacing the evaporated foam. Once cooled, the casting is shaken out, inspected, cleaned, and finished.

All melting at Foundry No.2 is done in two natural-gas dry-hearth reverb furnaces. All aluminum alloys are purchased as large billets, not ingots, and melted on-site. The plant has two air compressors (75 hp and 100 hp), both of which operate at 123 psi. It reclaims its mullite sand, of which 10% is run through a fluidized bed.

Because Foundry No.2 is a recent greenfield facility that produces only lost-foam castings, the planners were able to implement technologies and practices developed by the DOE's Industrial Technologies Program. The plant had the latest manufacturing technologies, so evaluating the impact of R&D implemented there was difficult.

Foundry No.2 uses only mullite sand in its operation, because of its low coefficient of thermal expansion. This results in castings with better dimensional conditions and less thermal degradation than does silica sand — a fact demonstrated by the LFC/ UAB research. Also incorporated from the outset at No.2 were the results of UAB's coating tests, which help contain the plant's scrap rate at a mere 1-2% a year, whereas the industry average is 5%.

The plant's planners considered which essential technologies would be needed to have a successful lost-foam operation. Their largest investment involved automated casting lines, which selected in order to increase throughput. The planners also installed a state-of-the-art multiple pattern-making machine, which reduces the plant's lead times. This gives the operators a way to to work closely with customers' casting designs.

Foundry No.2 managers use the Lost Foam Consortium as a knowledge base, and have invited the UAB scientists to perform research at their site, in part to offset the fact that they have no R&D staff.

Foundry No.2 developed a state-of-the-art lost-foam facility with a scrap rate well below the industry average, but the assessment team still identified areas for potential energy and financial savings there:

  • Ingot Size — While the foundry purchases aluminum billets (about three times as large as ingots), the ITP assessment noted that the savings on these billets — approximately $50,000 for every 1 million pounds — are impacted significantly by the extra energy purchased to melt the larger mass.
  • Compressed Air System —The 75-hp air compressor at Foundry No.2 is isolated from the 100-hp compressor and operates without a refrigerated drier. The whole system lacks an air receiver system. The ITP team recommended reevaluating the operation of this system, as well as to determine actual compressed-air needs. The current system operates at an unusually high pressure, and reducing that may save 1% of the energy cost for each 3 psi reduction in air pressure. The management might also consider installing a refrigerated drier for the system, as well as an air receiver in the system.
  • Lighting — The assesment team indicates that lighting in Foundry No.2 needs to be re-evaluated. Keeping unoccupied rooms illuminated wastes electricity and raises energy costs. Occupancy sensors will allow lights to turn on only when someone is present in those rooms.
  • Install Cogged Belts — One of the assessment teams top recommendations is to install energy-efficient cogged belts on motor systems. The average implementation cost is $1,263 with an annual pay back of $2,169. Cogged belts are more efficient because they slip less than smooth belts, and may lead to energy savings of 3-5% over what is normally used to operate a motor. They have a longer service, too.
  • Stack Melters — Like many aluminum melting plants, Foundry No.2 was adviced to consider replacing its reverberatory furnaces with a stack melter. With the increasing price of natural gas and the low efficiency achieved using a reverb furnace, this replacement may make economic sense.

Download a copy of Implementation of Metal Casting Best Practices at www.eere.energy.gov/industry/metalcasting/pdfs/implementation_final.pdf