Materials science

Alloys go with the grain

How do metallic alloys solidify from their original liquid state? A study of the deformation of cooling alloys confirms what had been suspected for some time: solidifying alloys bear exciting similarities to granular materials.

Metallic alloys are in constant, ubiquitous use. Generally, we prefer them in their solid state, and, in most cases, producing them requires cooling down a high-temperature liquid. This change from liquid to solid does not usually occur spontaneously at a well-defined temperature, as it does in pure metals. Instead, a continuous transition from a fully liquid to a fully solid state takes place gradually as the alloy cools.

On page 70 of this issue1, Gourlay and Dahle provide experimental evidence that, during this process, an alloy deforms rather like a granular material. Tools developed to model such materials should therefore allow new insights into the old problem of solidification. As well as being a confident step forward into the twilight 'mushy zone' between alloy liquid and solid phases, these findings could help to elucidate the formation of defects in economically important industrial casting processes.

In the early stages of alloy solidification, small, solid grains nucleate and move freely in the liquid phase. The result is rather like a suspension, with a characteristic fluid-like behaviour. Only towards the very end of solidification — when the solid grains have been bridged together — does the material acquire mechanical coherence and behave with the extreme viscosity expected of a solid at very high temperature. Between these two extremes lies the mushy zone (Fig. 1). As this zone comprises an assembly of individual grains interacting with each other through their contacts, it is natural to conjecture that its behaviour is similar to that of a granular material. But proving this has been difficult.

Figure 1: Through the mushy zone.

An aluminium–copper alloy gradually solidifies from the fully liquid state (above a temperature of 644 °C)4. The growing solid grains appear darker than the liquid phase. Gourlay and Dahle1 show experimentally that processes of deformation in this mushy mixture are similar to those in granular materials.

Gourlay and Dahle1 conduct their work on aluminium and magnesium alloys. Gathering experimental data on the deformation of granular materials is not easy even at room temperature. Working with materials that solidify at high temperatures (500–700 °C), and that have an extremely reactive liquid phase, is even less trivial.

The authors focus on the shear behaviour of alloys at solid fractions above 30%. Shearing occurs when two adjacent regions of a material slide past each other. In a granular material, it induces an effect called dilatancy that was first described in 1885 by Osborne Reynolds2. This phenomenon explains why the area around freshly laid footprints in wet sand becomes dry: deformation of the sand underfoot forces grains to rearrange, opening up spaces into which the surrounding water can flow.

As stepping on solidifying aluminium at 600 °C is less than pleasant, Gourlay and Dahle measured shear stresses in their alloys using an instrument known as a rheometer. The authors modified their rheometer to measure the rise, as shearing proceeds, of a solid component that floats on the initially liquid alloy — equivalent to measuring the inrush of water into the shearing area under a footprint. The authors observed that, after an initial near-linear increase in resistance to shear, the solid component rises. A decrease in the stress caused by further applied shear (an effect known as strain softening) follows, with a concomitant, more gradual volume expansion. These are characteristic behaviours of granular materials.

Another trait typical of granular materials is that they develop shear bands when deformed. These are well-defined regions of intensely sheared material in which the solid grains are significantly less densely packed. A nice feature of solidifying metallic alloys, compared with more standard granular materials, is that they can be quenched — rapidly cooled — to 'freeze' the microstructure at a given point. The observation of this microstructure can reveal important information about what happened when the alloy was a mixture of solid and liquid.

The quenched microstructures observed by Gourlay and Dahle show evidence of shear bands that are typically ten grains wide. Grain size can be controlled by, for example, changing the imposed cooling rate. The authors were thus able to confirm this result for different grain sizes from 10 to 500 μm.

Shear bands that form early on in the solidification process spell bad news for the solidifying alloy, and can have significant repercussions in industrial applications. The extra liquid that drains into the band remains as solidification proceeds. This leads to segregation of the solid and liquid phases, and thus to a heterogeneous composition in the material. Worse still, later on in the solidification process this can trigger cracks — 'hot tears' — that are among the most severe defects encountered in casting and welding processes.

The scenarios sketched by Gourlay and Dahle1 require confirmation by further experiments, as well as numerical simulations that can throw similarities with phenomena in granular materials into sharper relief. The authors' analysis is post-mortem, in the sense that it is carried out when the alloy has fully solidified. It needs verification and consolidation through experiments to track the evolution of microstructures during the actual formation of shear bands. X-ray microtomography, a technique that reveals the three-dimensional structure of materials, is an excellent candidate that has already shown the wealth of information it can provide both on standard granular materials3 and on solidifying alloys4,5.

Although modelling the behaviour of solidifying materials under deformation has already borrowed extensively from the field of granular mechanics6,7, this pooling of resources can and should be pushed further. In particular, new approaches should be developed that recognize explicitly the discrete character of solidifying alloys8. Thus armed, we shall penetrate further into the mushy zones.


  1. 1

    Gourlay, C. M. & Dahle, A. K. Nature 445, 70–73 (2007).

  2. 2

    Reynolds, O. Phil. Mag. 20, 469–481 (1885).

  3. 3

    Aste, T., Saadatfar, M. &. Senden, T. J. Phys. Rev. E71, 061302 (2005).

  4. 4

    Ludwig, O., Dimichiel, M., Salvo, L., Suéry, M. & Falus, P. Metall. Mater. Trans. A 36, 1515–1523 (2005).

  5. 5

    Li, B., Brody, H. D. & Kazimirov, A. Phys. Rev. E70, 062602 (2004).

  6. 6

    Flemings, M. C. Metall. Trans. B 22, 269–293 (1991).

  7. 7

    Ludwig, O., Drezet, J.-M., Martin, C. L. & Suéry, M. Metall. Mater. Trans. A 36, 1525–1535 (2005).

  8. 8

    Vernède, S., Jarry, P. & Rappaz, M. Acta Mater. 54, 4023–4034 (2006).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Martin, C. Alloys go with the grain. Nature 445, 34–35 (2007).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.