constraining atoms to be inside region

Hello list:
I am doing some simulations of a nanodrop of copper and would like to prevent the surface atoms from evaporating. To do this, I embed the copper in an inflated spherical region and use fix wall/region to constrain the atoms inside with a harmonic potential and a small cutoff. However, the code crashes with a segmentation fault. I'm not sure if I'm doing this correctly however:

region vSphere sphere 47 47 47 20 side in units box
fix restrainP all wall/region vSphere harmonic 10.0 1.0 20

Alternatively, I'm thinking of implementing a flat bottom potential and using fix spring.
Has anyone attempted such a simulation and have some tips they would like to share?

Thanks,
Mike

Do the thermodynamics of the system blow up right
before the crash?

Steve