Shell-model MD with constraints fails (Not yet implemented)

[Gabriele Tocci]

Dear CP2K Masters,

I am trying to run an MD of a ZnO slab using the shell model with
constraints on the bottom layer ions, and also an MD of liquid water
on a ZnO surface using SPC/E water (which, of course treats water
molecules as rigid).
However, I have seen that the use of constraints with a shell-model MD
has not been implemented yet.

First of all, do you know a quick/easy way around this, that does not
involve implementing constraints on shell-MD?

Also, is there any interest in implementing constraints+Shell-MD in a
future (close) update? I post below the part of the subroutine
setup_velocities in md_vel_utils.F which states what I have mentioned
and where my code fails:


    IF(shell_present .AND. shell_adiabatic) THEN
       ! shell-model with constraint NYI
       CPPostcondition(.NOT.simpar
%constraint,cp_failure_level,routineP,error,failure)
 
CPPostcondition(ASSOCIATED(shell_particles),cp_failure_level,routineP,error,failure)
 
CPPostcondition(ASSOCIATED(core_particles),cp_failure_level,routineP,error,failure)
       shell_particle_set => shell_particles%els
       core_particle_set  =>  core_particles%els
    END IF

I hope that there is interest in implementing it, as the shell model
can describe well oxides systems and if one ever wants to compute
surface properties constraints may have to be used. Also metal oxides-
liquid interfaces
may need the use of constraints.

Cheers,
Gabriele.