Effects of annealing and specimen geometry on dynamic compression of a Zr-based bulk metallic glass

High strain-rate compression experiments were performed with a split-Hopkinson pressure bar (SHPB) at 500–4000/s on cylindrical samples of a Zr-based bulk metallic glass (LM-1) in both the fully amorphous and annealed conditions. The effects of changes to the specimen geometry (i.e., L/D ratio) and the material heat treatment [i.e., annealing versus amorphous (as-received)], on the peak stress, strain-to-failure, and failure behavior were determined with the aid of an in situ video obtained by using a high-speed digital camera in conjunction with the split-Hopkinson pressure bar (SHPB). Examination of the in situ video recordings and light optical microscopy showed that the failed samples revealed preferential failure initiating at the sample ends due to stress concentration at the sample-insert interface. A new insert design was developed using transient, elastic-plastic finite-element simulations to reduce the effects of these stress concentrations. SHPB testing, combined with in situ video, subsequently revealed that this new experimental configuration promoted failure within the gage length and away from the sample ends in the samples tested. Significant effects of specimen geometry, insert design, and annealing on the apparent values of the peak stress, strain-to-failure, and fracture behavior were exhibited.