It was noted that all magnesium products begin as liquid, whether die-cast, wrought, or even advanced composites and powder metallurgical concepts that have been proposed recently. It was further noted that the vast majority of current magnesium alloy products are die cast. The need to understand solidification effects on microstructure and properties, in that case, are obvious.
It was celebrated that the thermodynamics (free energies, phase diagrams, etc.) of many of the alloy systems of interest are in reasonable shape and are continuing to develop. Accordingly, we do not view this as an area of desperate need for increased support, but the present level should continue, since one may expect the continuing emergence of new alloy systems which will need further exploration. On the other hand, there are large gaps in our understanding of fundamental diffusion kinetics in both solids and liquids – more complete diffusion databases are needed and new simulation capabilities needed – first principles, cluster variation, and kinetic Monte Carlo techniques were mentioned. Additionally, we need to develop a better quantitative understanding of the properties of Mg alloy liquids (viscosity, thermal conductivity, diffusivity, and solution type, e.g. regular or other, etc.) With these properties in hand, simulations of liquid metal flow as well as solidification itself, will improve. Details of the nucleation of the solid within the liquid are still viewed as an area requiring further research. Finally, better models of porosity formation are needed for many metal systems and Mg alloys are no exception.
More broadly speaking, higher fidelity simulations of the entire casting process are required. For example, most casting simulations employ Scheil solidification models, which are inadequate to describe solidification during the typically high rates of cooling employed during most casting operations. New simulation models to more accurately simulate non-equilibrium cooling response are required. These models need to provide a means to link solidification paths to the final microstructure formation (e.g. solute distribution is critical for predicting subsequent age hardening response). It is also important to consider that as new casting and solidification processes develop, their needs will put further stress on the fidelity of existing simulation methodologies.
Producing fine grain microstructures is a goal of many metallurgical processes, and the properties of magnesium alloys appear to respond particularly favorably to grain size reduction. For instance, grain size reduction has been shown to improve the creep resistance of some magnesium alloys, which is unusual. (It is wondered if this result is restricted to alloys based upon the Mg-Al system.) The tremendous grain refining potential of Zr in Al-free Mg alloys is well-known. However, research into other inoculation strategies is still viewed as worthy of research. The potential of adding small (nano-sized) particles to improve the properties of cast Mg should be further explored. Questions arise concerning the grain refining possibilities associated with semisolid processing. In fact, there is a need for models for flow of semisolid material over the range conditions encountered in die casting, thixomolding, etc. Additionally, the possibility of hybrid cast-wrought processes (such as nanoMAG and twin roll casting) should not be overlooked. The grain sizes that are produced by these processes can be quite small and the potential of these processes has not been fully explored.
The potential of twin roll casting has been trumpeted for some years in the Mg community, based upon laboratory-scale results, but the process appears to be very technologically challenging during scale-up. Holistic engineering and simulation strategies are needed. The U.S. research community has played essentially no role in this area, which is dominated by Korean, Australian, Chinese, German, Turkish, and Canadian companies and research institutes. Furthermore, it has yet to be demonstrated that the investment costs for such sheet products has a reasonable return on investment.
The final area explored within the context of casting and solidification was the development of Mg-based metal matrix composites. A more rigorous microstructure design framework is needed: micro- and nano-composites do show promise, but less Edisonian approaches are needed. What is needed is modeling of in-situ phenomena (solid/liquid interfacial phenomena) and property prediction before making the composite. Use of new 3D characterization methods, such as synchrotron facilities to uncover a better understanding of liquid-solid phenomena should be encouraged. Dispersion control seriously inhibits the potential of many composite processing strategies, and techniques to prevent agglomeration, ranging from colloid-type approaches to agitation (ultrasonics) should be evaluated. Lower cost approaches aimed at selective reinforcement (e.g. for wear resistance) via local compositing should also be explored.
The workshop participants agreed that new Mg alloys are needed which target specific property combinations, both to address existing limitations of Mg (susceptibility to corrosion), and to further enhance the advantages of Mg. Currently, there are a very limited number of Mg alloys commercially available from which design engineers must choose. Furthermore, there was a consensus that today’s pace of technology is such that non-Edisonian alloy development efforts need to take place within five-year periods, rather than the 80 years it took to develop the current suite of ultra-high strength aluminum alloys. Rapidly solidified Mg-Y-Zn alloys with tensile yield strengths of 600 MPa and elongation ~ 5% already exist. Developing compositions and processing strategies that make these such property achievements cost-competitive is a real goal.
Mg-Zn is viewed as perhaps the most promising binary system for developing high strength casting or wrought alloys because of the strong precipitation hardening response in this system. In addition, this system meets the need for low-cost alloys. To advance this and other promising alloy systems, principles or rules need to be established for selecting appropriate alloying additions and including micro-alloying additions; e.g., to further enhance age hardening response and mechanical properties. Some published work has already illustrated the great potential of micro-alloying strategies in the Mg-Zn system. There is a balancing point of view, however, since there are ranges of Zn content which suffer from poor hot cracking and corrosion resistance. These facts have historically limited the applicability to die casting and they have limited the application of high strength commercial alloys like ZK60.
Despite the fears surrounding the current high price of rare earth (RE) metals and the nearly sole-supplier status of China, it is suggested that Mg-rare earth systems still deserve more research. Rare earth additions to Mg have proven very potent for improving strength (numerous alloys), creep resistance (numerous alloys), resistance to flammability (e.g. WE43), and reduction in the texture strength of wrought products and subsequent improvement in formability. For one thing, developing alloys which retain these advantages at lower RE content could improve the weight and price competitiveness of Mg alloys. For another thing, the scientific knowledge which is developed through the study of RE alloys may be translated into alloy development strategies applicable to non-RE-containing systems. As an example, RE-containing AE44 was developed to enable the production of a lightweight engine cradle for the Corvette. Later, the AX alloys were developed on the basis that Ca was a low cost alternative to RE which forms similar precipitate phases.
At the risk of redundancy, it is again observed that the development of thermodynamic databases has progressed at a good pace in recent years. However, there continues to be a need for more accurate binary and multi-component phase diagrams. For example, the Mg-Nd binary is controversial. Perhaps more urgently, there is a great need for diffusivity data for alloy development. Interfacial energies are also important. This will take a combination of experimental measurements (e.g., using three dimensional EBSD measurements) and some calculations. Interfacial energy of precipitates is also needed for developing precipitation hardenable alloys. More modeling work is needed to support science-based alloy development, including first-principles calculations, molecular dynamics, monte carlo, phase-field, etc, to cover major issues on thermodynamics, kinetics, precipitation, crystal plasticity, strengthening, etc. The lecture by Dallas Trinkle illustrated the potential of first-principles atomistic calculations within a concurrent multiscale framework for predicting bulk alloy properties.
Finally, while many emphasized the need to develop alloys with high strength (in combination with various other properties such as corrosion resistance), some participants expressed the view that not all applications demand high strength. This contingent suggested an equally important focus should be placed on the development of high formability alloys through enhancements in: i) work hardening and ii) resistance to damage initiation. Some suggested that damage initiation, especially at high strain rates, is linked to one of the deformation twinning modes. They suggest that twinning should be suppressed by alloying or grain size reduction. Evidence that such an approach will work is still lacking. In fact, some researchers point out the possibility of enhancing certain types of twinning in ultrafine grained Mg alloys. It would be beneficial if hard answers to these implied questions could be provided by the scientific community.