Hello,
First make sure to take the enclosed corrected subroutines for Smagorinski. You will see that CS is hardcoded in them, you may change it if you want. Then this turbulent viscosity is added to the laminar viscosity which is considered in this case to be the coefficient for horizontal diffusion of velocities that you give, so prefer 1.D-6 rather than 0.01. Same thing for Nezu and Nakagawa.
With best regards,
Jean-Michel Hervouet