Effects of sample geometry on deformation modes of bulk metallic glasses at the nano/micrometer scale

Abstract

Intense debates have been prompted concerning whether homogeneous deformation can be achieved in bulk metallic glasses at room temperature through the suppression of shear bands at the submicron scale. In this short communication, we demonstrate that multiple shear banding can be successfully attained via a proper modification of the microsample geometry, resulting in the appearance of a homogeneous deformation mode at the submicron scale. However, the apparent deformation homogeneity in our microcompression experiment is a manifestation of the sample geometry effect on the propagation rather than nucleation of shear bands.Understanding the behavior of shear bands in bulk metallic glasses (BMGs) is critical in the design of a new class of BMGs with superior mechanical properties. In BMG bulk samples, shear bands tend to propagate in a catastrophic way, resulting in the subsequent brittlelike fracture, which renders the BMGs with limited ductility at room temperature.1 However, at the micrometer scale, shear bands were found to behave in a ductile manner, propagating without causing the sudden fracture of the microsample throughout a large plastic deformation process.2 Moreover, it was reported that shear bands could even be completely suppressed by reducing the sample size into the submicron scale, which led to a homogeneous deformation mode at room temperature. 3,4 However, the idea of the shear-band suppression is controversial and was later challenged by a few researchers based on their results of the microcompression experiments, [5][6][7] in which no homogeneous deformation mode was observed on the micropillars even though their diameters were reduced to a level less than that observed in the homogeneous deformation mode. It can be argued that the difference of the experimental results originates from the chemical composition of the microsamples, i.e., different BMGs have different critical length scales below which the shear bands cannot be nucleated. Thus, the homogeneous deformation is still an intrinsic effect, reflecting the shear-banding behavior at the small size scale. However, there is another extrinsic effect that could also cause the apparent deformation homogeneity, but was somehow ignored in previous studies, [2][3][4]6 which is derived from the sample geometry and the associated boundary conditions.In mechanical testing of single crystals, it is known that the grips play an important role in determining the deformation morphology of a tensile specimen. If the lateral translations of both grips are constrained, a dislocation single slip is not favored, and, thus, the resulting deformation morphology is a reflection of the combined effect of the sample geometry and dislocation dynamics. Likewise, although shear banding is essentially heterogeneous in BMGs, an apparent homogeneous deformation mode is still possible as long as the shear offset resulting from the propagation of the small shear band is insignificant in changing the structural integrity of the microsample. In such a case, the deforma...

The thermal history effect on shear band initiation in metallic glass

“…Before proceeding, we would like to further discuss the physical meaning of shear band initiation and the length scale we extracted. According to the prior experimental 10,11,40,[42][43][44] and computational 9,47 results, shear band initiation was usually associated with the transition from a homogeneous deformation to inhomogeneous (shear-banding) deformation in a uniaxial test, either tension or compression, with the transition size ranging from $100 nm to $400 nm. However, there were also experimental reports that shear banding still prevailed at the length scale of $100 nm; 41,48-51 the recent computational simulation 9,47 even showed that shear banding can be observed down to the length scale of 10 nm.…”