Welding bulk metallic glass using nanostructured reactive multilayer foils

We have used Al/Ni reactive foils to weld Zr57Ti 5Cu20Ni8Al10 metallic glasses. The welds are a composite morphology comprised of glass ligaments and intermetallic AlNi (the product of the reactive foil). The presence of the presumably brittle intermetallic (in lieu of the glass) is expected to limit the mechanical properties of the welds. Based on fracture toughness measurements and the crack propagation paths, we conclude that virtually all of the toughness can be ascribed to the presence of the metallic glass ligaments. Increasing the pressure applied during welding increases the fraction of the joint made of these ligaments and so increases the fracture toughness as well.;To eliminate the intermetallic from the weld altogether, we attempted to fabricate reactive mulitlayer foils that form an amorphous product by melting and cooling rapidly during a self-propagating reaction. We began with reactive foils with overall composition Zr2Ni but quickly determined that the foils did not fully melt. We then attempted to lower the melting temperature and increase the glass forming ability and the heat of mixing by adding Al and Cu. These foils again did not fully melt. Finally we systematically determined that foils of overall compositions Hf37Ni63, Ni 80P20, and Ni60P40, which are all known binary metallic glasses, will potentially melt during a self-propagating reaction.;Knowledge of the phase transformations during a self-propagating reaction is necessary to engineer reactive foils for future applications. Furthermore, reactive foils provide an opportunity to study phase transformations under high heating rates not easily achievable. Characterizing the processes in the reaction zone however is challenging, requiring both temporal resolution better than ∼ 100 mus (the time required for the reaction front to pass a fixed location) and spatial resolution of < 100 mum (the approximate width of the reaction zone). Using synchrotron x-ray radiation, we have studied these phase transformations in situ in Al/Ni multilayers. Unlike previous annealing and quenching studies in these multilayers, we observed no metastable or intermediate phases.