Genesis of a new process: magnesium lost foam casting.

The casting industry is creative and entrepreneurial. Throughout its history, casting methods and materials have evolved to meet the challenges of changing production, quality and customer requirements. Someone always seems to be developing a new or better way to improve casting properties and to produce more economical components.

The industry is at it again. Preliminary work has begun on the development of a new casting process--magnesium lost foam casting.

Growth in Mg

The last 110 years have seen a dramatic increase in the production and utilization of magnesium castings (Table 1). This demand primarily is driven by the automotive industry's desire to increase vehicle fuel efficiency by using alternative materials in vehicle components to reduce vehicle weight. Because of its low density and high strength-to-weight ratio, magnesium has found a high growth market in automotive components. Component applications also are developing in other transportation markets such as heavy truck, where weight reduction results in increased payload capacity. According to Stratecasts, Inc., magnesium casting shipments are forecast to grow from 78,000 tons in 2001 to 242,000 tons in 2011.

Current and potential magnesium automotive applications include front engine covers, oil pans, instrument panels, seat components, door latches and closures, transmission cases, rocker arm covers and front end structures. Presently, the majority of current cast magnesium components are produced via high-pressure diecasting. According to Stratecasts, cast magnesium content in automobiles is expected to grow from its current status of 11-15 lb/vehicle to 35 lb/vehicle by 2012.

For many years, significant sand cast magnesium production has occurred--a mature market primarily serving military and aerospace. As the automotive market expands, a renewed interest in magnesium castings should spark other markets to develop. However, potential new magnesium components could require geometries and properties that cannot be produced by high-pressure diecasting. While precision sand casting can produce a wide range of part shapes with high levels of soundness, the process is expensive and generally limited to low-volume high-value components.

A genuine need exists for alternative casting processes to fill the gap between diecasting and sand casting. The APS Magnesium Div. was established in May 2000 to address the need for technological development and alternative casting methods, and to new encourage new magnesium component applications.

The Magnesium Div. identified as its primary goal the development gravity and low-pressure permanent mold casting processes, but also explored the possibility of combining the benefits of lost foam casting with the advantages of magnesium. The lost foam process is capable of making complicated part shapes with low manufacturing costs and can consolidate several fabricated components into a single cast part-benefits that could be combined with the weight advantage of magnesium to help create a market for new magnesium casting applications.

It was quickly discovered that little had been done in this area, and that there was little documentation of any previous research to provide casting guidelines. Moreover, many of the 'experts' consulted warned that it probably could not be done because of two primary issues--magnesium's low thermal heat content (heat loss of the metal as it burns out the foam pattern would create misfill castings) and concern that the chemical reactivity of magnesium would be affected by the coated foam pattern. The Magnesium Div. decided to begin an experiment to cast magnesium via lost foam.

Project Development

The first step was to research materials best suited to cast magnesium lost foam using current lost foam casting practices in iron and aluminum as starting points.

Although aluminum is often cast via lost foam, there are a number of differences between aluminum and magnesium that could prohibit this process in magnesium. The two primary differences between aluminum and magnesium are thermal conductivity and chemical reactivity.

Because of magnesium's low thermal heat content, filling a lost foam mold would create the greatest challenge, because the heat required to vaporize the foam pattern could draw enough latent heat from the molten magnesium to prevent mold filling. (The specific heat k/cal value of aluminum is 0.70 compared to 0.57 for magnesium.) While aluminum would require a lost foam molding media that has a high conductivity and low thermal expansion, magnesium would require a low conductivity media (to retain more melt heat) with low thermal expansion properties.

This fluidity difference between molten magnesium and aluminum alloys continues to be the biggest potential barrier to success. The second potential technical barrier, the reactivity of magnesium, could potentially be controlled by the use of inhibitors, common in the magnesium sand casting industry, and aided by the use of a high-porosity pattern coating.

Material Selection

It was decided that casting trials at a foundry with magnesium experience was more important than casting trails at a lost foam foundry. To help minimize the fluidity concerns, foam specialists recommended lower density foam beads. As metal is poured directly onto the foam patterns, decomposition products are formed at the degradation (leading) edge of the advancing metal. Selecting a low-density foam material would result in less decomposition products present during metal filling to help maintain heat in the metal as it advanced through the foam pattern during casting.

A test box pattern and a window test pattern (regularly used to evaluate the aluminum foam casting process) were selected for the casting trials. Foam patterns were produced using low-density silene lubricant beads with a triblend mixture of three different pentane isomers as the blowing agent. The two bead grades selected were T170 and T180C and produced with 1.5 and 1.04 densities (specific gravities). Once the foam patterns and in-gates were glued together, the patterns were coated with a mica-based slurry typically used in lost foam casting of iron because of its high thermal insulation and permeability characteristics.

To further inhibit potential alloy heat loss, standard foam downsprues commonly used for aluminum lost foam casting were replaced with insulated hollow fiber downsprues. Prior to use, the fiber downsprues were coated with an inhibitor to reduce the risk of reaction with molten magnesium. The fiber downsprues were then glued to the foam ingate just prior to molding the patterns in the flask.

High-permeability 40 GFN silica sand was selected as the backing sand and a mixture of sulfur and potassium fluoroborate inhibitor (used in magiesium sand casting) was mixed in with the sand as an added safety factor to decrease the possibility of reaction. Standard commercial alloy AZ91E magnesium alloy was chosen for the trials because of its high purity level, good corrosion resistance and its reputation as the "workhorse" alloy for magnesium sand casting.

Casting Trials

Using standard melting practices, AZ91E magnesium alloy was melted using [SF.sub.6] cover gas for melt protection. After bringing the melt to temperature and drossing, hexachloroethane tablets were added to the metal as a degassing agent and grain refiner, Gating systems for all initial pours were side-gated with bottom in-gates.

Patterns were placed in a flask half full with sand (Fig 1). The flask was vibrated as more sand was added to pack the sand around the pattern. On the first day of the castings trials, both patterns were invested in the sand flask and poured at 1500F (816C) with the AZ91E alloy. Although these pouring temperatures are considered high for magnesium, the goal was to completely fill these patterns without cold lap or misruns.

Pour times for both molds were approximately 6-8 sec. After the first castings were poured complete without misfills (Fig. 2), it was evident that magnesium and lost foam were compatible.

Since the emphasis of this trial was casting process development and not casting quality, no genuine attempt was made to produce usable, defect-free castings. X-ray examination of the first day's castings showed the presence of oxide films in some of the castings. Because the first casting was successful, the research team decided to attempt to produce an X-ray quality casting on the final scheduled pour.

The gating system on the last pattern was modified to include a stainless steel wire mesh to the bottom of the fiber downsprue (Fig. 3), acting as a filter (common foundry practice in magnesium sand casting). The window pattern was made with a shortened downsprue glued to a foam 'sprue' on top of the window pattern--the casting would be top gated and poured directly into the top of the test window (Fig. 4).

While the previous castings had pouring times of 6-8 sec, the pouring time for this final casting was slightly longer. The pouring cup was initially filled as with the other castings, but the pour was interrupted for a few seconds to allow the metal level to drop. Once the level in the cup dropped, the cup was topped off.

After visual inspection, castings were sent to X-ray for quality inspection and all castings were confirmed grade C minimum. The X-ray results on casting #7, the test window pattern with the modified gating system and the metal mesh filter, confirmed a grade B or higher (X-ray grades per Mil-Std-2175).

After X-ray inspection was completed, the castings went through a T6 heat treatment cycle to obtain the mechanical properties. Test results show values are comparable to those obtained by precision sand castings (Fig. 5).

Potential Applications

The combination of magnesium and lost foam casting will require further technological development, but the success of these initial tests demonstrates the significant potential for magnesium casting. Perhaps the most immediate application will be the use of magnesium lost foam as a rapid prototyping technology. The production of magnesium prototype castings is already a major market, and the potential to cut foam into final component shapes and cast prototypes and limited production runs in lost foam should improve this market niche.

Using cut foam technology, early development castings can be produced without tooling, significantly reducing the cost and risk associated with the development of new magnesium components. In addition, the mature magnesium sand casting market currently produces structural castings via green sand or chemically bonded sand molding. Many of these components may be candidates for conversion to lost foam castings, where improved yield, simplified molding and cleaning, and the potential for part consolidation will be a cost advantage for existing magnesium parts such as high performance manifolds and other specialty automotive components, various military and aerospace castings, and electronic enclosures and components.

The primary long-term potential of this new technology is in the development of new magnesium applications. Current government-funded magnesium development programs include the design and construction of automotive magnesium engine blocks. Many aluminum engine blocks are now produced using lost foam casting, and magnesium lost foam casting offers a cost-effective way to produce these parts in volume.

For More Information:

"Developments in Lost Foam Casting of Magnesium," Michael Marlatt, David Weiss and John Hryn, 2003 AFS Transactions (03-112), AES, Des Plaines, IL (2003).

"Effect of Processing Parameters on Mold Filling in Magnesium Alloy EPC Process," A.Z. Liu., J. Hu, W. Ding, Q. Wang, Y.G. Zhu, Y. Lu and W. Chen, 2001 AFS Transactions (01-014), AFS, Des Plaines, IL (2001).

[GRAPH OMITTED]

[FIGURE 5 OMITTED]

About the Author

Established in 2000, the AFS Magnesium Div. has set as a primary mission the development of alternative casting processes, Its second objective is to better understand the metallurgical issues of gravity-cast magnesium. The Division has 28 members.