Metallic glasses can resist higher stresses than crystalline metals of similar composition.  Unfortunately the benefit of this high strength is offset by the fact that when this limit is exceeded the material tends to exhibit a catastrophic failure mode known as “shear banding”.  During shear banding rather than deform evenly throughout the material concentrates all its deformation in a very narrow region.  This region forms a slip plane, like an earthquake fault, which becomes much weaker than the surrounding material.  Typically after the shear band forms the plane becomes so weak that it is susceptible to fracture.

The nature of the atomic rearrangements that cause shear banding are not well understood.  Furthermore these kinds of instabilities are difficult to study directly in the laboratory because they occur on very small length and short time scales.  Instead, by performing careful atomic scale simulations using a technique known as “molecular dynamics simulation” the Falk research group simulated the motion of all the atoms in the metallic glass as it was subjected to large enough strains to observe the shear banding phenomenon and the subsequent evolution of the shear band.  Interestingly, shear bands in the simulated model get wider and wider as more strain is imposed on the material.  This appears to indicate that the shear bands may not be stable objects, but are rather constantly evolving as they continue to be subjected to further strain.

One interesting aspect of the analysis is the indication that a stable backbone of relatively ordered structure in the glass may simultaneously provide its strength, and also result in its destruction. This backbone of atoms with ordered local environments is relatively rigid.  In its absence the material tends to deform throughout, but if it exists then once the material breaks down in some region all further deformation is likely to happen in that same location.  This seems to play an important role in the localization phenomenon.

Most importantly these simulations have permitted the testing of a quantitative hypothesis regarding the relationship between the rate of slip on the band and the degree of bonding in the metallic glass.   Such “constitutive relationships” are critical to be able to model shear banding instabilities and to determine the aspects of metallic glass response which are responsible for determining their strength and ductility.  This work aims to provide the basis for mathematical analysis of such problems in metallic glasses and other materials that experience similar failure modes.