The strain-hardening of conventional polycrystalline metals underpins their importance as structural materials, imparting ductility and resisting catastrophic failure.  While metallic glasses (MGs) can be tough, in simple tension they fail catastrophically without ductility, because plastic flow is severely localized in shear bands ‒ a direct consequence of strain-softening.  

Prof. Yi Li (Institute of Metal Research, Shenyang, China), and colleagues, showed that tensile loading of notched cylinders of a bulk MG can induce relaxation [1].  He discussed with me his results on compressive loading of the same notched cylinders in compression.  These showed extreme induced softening that was of clear interest:  relaxation (ageing) of MGs leads to degraded mechanical properties (embrittlement), so the opposite process of “rejuvenation” (raising the energy, and softening) should be useful.  Indeed, Yonghao Sun, Amadeu Concustell and I had suggested that rejuvenation might even be a way to achieve strain-hardening in MGs [2], though we were not confident that sufficient rejuvenation would be possible without causing spontaneous crystallization of the MG.  

With lead contributor Dr Jie Pan at the IMR, Yi Li and I submitted a manuscript on ‘extreme rejuvenation’ to Nature Communications.  Ultimately, in revised form, this was accepted for publication [3].  The Peer Review file shows that we are indebted to Reviewer #1, who pointed out that (contrary to our initial view):  “the deformed area must extend substantially above and below the notched region”.  When we mapped the hardness across a sample cross-section, it was immediately clear that the reviewer was correct! ‒ see the coloured map in the poster image.

When Yi Li and I had discussions at the BMG XII conference in Seoul, Korea, it occurred to me that the mapping showed that it should be possible to extract cylindrical specimens (dashed box) for testing of mechanical properties on mm-scale samples that had been highly rejuvenated.  In September 2018, I saw the first results from the IMR on simple uniaxial loading of such extracted specimens.  The stress-strain curves showed strain-hardening, defying the paradigm that MGs cannot show strain-hardening in simple macroscopic tests because their initial high yield stress means that ‘the only way is down!’  If there is strain-hardening, there should be no shear-banding, and indeed there was evidence that shear-banding was suppressed.  

We confirm that the mechanism of the hardening is structural relaxation of the MG, by showing that the enthalpy of the rejuvenated glass decreases when it is deformed.  Furthermore, Dr Yurii Ivanov in Cambridge showed that the MG adopts a lower density upon rejuvenation, and reverts to a higher density on subsequent deformation; he used transmission electron microscopy, which seems to be able to give more reliable results than X-ray diffraction because the subtle structural changes can be obscured by residual elastic strains in the larger samples used in the latter technique.  The strain-hardening seen in our work is associated with relaxation to lower-energy states, in complete contrast to the strain-hardening of polycrystalline metals, which necessarily implies an increase in defect population and the attainment of higher-energy states, as shown by G.I. Taylor some 85 years ago [4].  

Lindsay Greer, 7 March 2020 

Listen to Lindsay talk about this an episode of the Nature Podcast.

[1]    Wang, Z. T., Pan, J., Li, Y. & Schuh, C. A. Densification and strain hardening of a metallic glass under tension at room temperature. Phys. Rev. Lett. 111, 135504 (2013).

[2]    Sun, Y. H., Concustell, A. & Greer, A. L., Thermomechanical processing of metallic glasses:  extending the range of the glassy state. Nat. Rev. Mater. 1, 16039 (2016).

[3]    Pan, J. et al. Extreme rejuvenation and softening in a bulk metallic glass. Nature Comm. 9, 560 (2018).

[4]    Taylor, G. I.  The mechanism of plastic deformation of crystals. Part I.‒ Theoretical. Proc. Royal Soc. A 145, 362‒387 (1934).