Optimisation of multi-component hard sphere liquids with respect to dense packing

The generalised Bernal model for liquids is used as an approach to the structure of multi-component liquid and amorphous metallic alloys. It is assumed that alloys that can be described by random close packing models of hard spheres with particular high packing fraction may have a better glass forming ability than those with lower ones. We show that the optimisation of computer simulated random close packing models is a promising way to propose models with enhanced packing fraction and to specify the compositions and radii distributions required. A series of multi-component systems with enhanced packing fraction is presented and proposed as candidates for new bulk metallic glasses.