Hello,
I have a system contains particles that are formed of dpd beads and randomly distributed in a solvent medium. Particles don’t have bond inside them. The particles are rigid using fix rigid/small. My question is that how it would be possible to prevent particles being crossed by the solvent beads in the system. I was wondering if pair_style srp can be useful but there isn’t any bond inside the particles. I would appreciate to know any ideas in this case.
King Regards
Katerina
it is my understanding of the dpd model that it effectively represents a simulation of particles on a time scale beyond what you would get from conventional atomistic particle simulations. thus its trajectories would be equivalent of those from a regular simulation and then looking at time averaged positions instead. now if you would take a trajectory of a simulation of the same kind of system as you have and then write out a trajectory where you use time averaged positions from, say, 1000 steps, you should find that particles appear to be overlapping that otherwise would not be overlapping. consider a solvent atom that would be diffusing around a large rigid object during those steps. in the averaged position it would instead appear to be inside the rigid object. for illustration purposes, i would suggest to try running a similar simulation using a lj/cut pair style (and a correspondingly much shorter timestep) and then visualize the resulting trajectory in VMD, which has an option to apply a sliding window average trajectory smoothing in the Trajectory tab of the Graphical Representations dialog. The larger you set the trajectory smoothing window size, the more particles would appear to overlap.
thus whether it is necessary to modify your interactions to repel solvent particles from overlapping (too much) with your rigid bodies is a matter of the model itself and the time scale you are looking at.
that said, it should be trivial to change the interactions in your simulation by either adding an additional particle to the center of your rigid bodies that will interact with a different, non-crossing, more repulsive potential using pair style hybrid, or by taking a particle at or near the center and adding such an interaction to it with pair style hybrid/overlay. what this means for the correctness of your model is something you will have to decide and verify yourself.
axel.
Dear Axel,
Thank you very much for your explanations and suggestions.
King Regards
Katerina
it is my understanding of the dpd model that it effectively represents a simulation of particles on a time scale beyond what you would get from conventional atomistic particle simulations. thus its trajectories would be equivalent of those from a regular simulation and then looking at time averaged positions instead. now if you would take a trajectory of a simulation of the same kind of system as you have and then write out a trajectory where you use time averaged positions from, say, 1000 steps, you should find that particles appear to be overlapping that otherwise would not be overlapping. consider a solvent atom that would be diffusing around a large rigid object during those steps. in the averaged position it would instead appear to be inside the rigid object. for illustration purposes, i would suggest to try running a similar simulation using a lj/cut pair style (and a correspondingly much shorter timestep) and then visualize the resulting trajectory in VMD, which has an option to apply a sliding window average trajectory smoothing in the Trajectory tab of the Graphical Representations dialog. The larger you set the trajectory smoothing window size, the more particles would appear to overlap.
thus whether it is necessary to modify your interactions to repel solvent particles from overlapping (too much) with your rigid bodies is a matter of the model itself and the time scale you are looking at.
that said, it should be trivial to change the interactions in your simulation by either adding an additional particle to the center of your rigid bodies that will interact with a different, non-crossing, more repulsive potential using pair style hybrid, or by taking a particle at or near the center and adding such an interaction to it with pair style hybrid/overlay. what this means for the correctness of your model is something you will have to decide and verify yourself.
axel.