Finding force/drag coefficient on a molecule in constant flow

Dear lammps users and developers:

I am trying to find the force/drag coefficient on a static molecule in a system composed entirely of solvent and am struggling to do so. It seems this should be a simple simulation but so far I have not succeeded. I have tried:

  1. Setting the molecule gravity force constant and ‘pulling’ it through the solvent until it reaches terminal velocity. Because of finite size effects, the solvent begins to flow and terminal velocity is never reached

  2. Freezing the molecule using ‘fix setforce 0.0 0.0 0.0’, ‘fix freeze’, or not including it in the integration; giving the solvent a constant velocity; and using the ‘compute group/group’ command to find the force on the molecule. The average force always approaches 0. I do not understand why.

  3. Tethering the center of mass of the molecule to a point using ‘fix tether’, giving the solvent a constant velocity, and then calculating the average separation between the molecule and tether point to find the force on the molecule. This may have worked but fluctuations are quite large and it looks like the force will approach zero for very long simulation times

  4. Giving the solvent a constant velocity and the molecule a constant acceleration and trying to find the point where the molecule does not move. I could not find a balancing value.

I set the velocity of the solvent using ‘velocity NULL NULL 0.1’ , and constant force on the molecule using ‘fix gravity’. I am using lammps version 16Mar18. Simulations end without errors.

Any help on how to properly set up this system would be greatly appreciated.

Thank you,

Sean B.

i think you could take advantage of the einstein relation here, relating diffusion coefficient to mobility. in other words, you should apply an artificial force to neither molecule nor solvent

I cannot use the Einstein relation here because I am running in DPD. Since the beads are soft and the molecule under consideration is composed of about 700 beads, the relation would not be accurate. For the purposes of this work, the relation is not enough.