The question of how many smaller spheres can be fit into a larger sphere is fascinating and has been examined extensively. I was curious, though, about the scenario of packing spheres of different radii into a larger sphere. For example: What is the maximum number of rigid spheres of radius 1 and radius 2 which can be packed into a rigid sphere of radius 10, given that there must be an equal number of radius 1 and radius 2 spheres? I'm not sure what packing strategy would produce an optimal result, so any ideas would be appreciated. Thanks in advance.
A recent paper is Marshall and Hudson, Dense binary sphere packings, Beiträge Algebra Geom. 51 (2010), no. 2, 337–344, MR2682460 (2011g:52030). It deals particularly with the case where the two radii are 1 and $\gamma$, where $.444\lt\gamma\lt.482$, whereas you want $\gamma=1/2$, but it also contains an overview of the problem and an extensive list of references.