Coarse-grained polymer in solvent

Hello all,

I have a 100 chain by 100 bead polymer system in a solution of 320 000 single beads (LJ sigma = 0.5, m = 0.125). The system is relaxed with DPD, then NVE, NPT (first at 0.6 then 1.0), and NVT to implement the Lennard-Jones parameters and equilibrate the box (all done at zero pressure). Then before going through a controlled cooling, the system goes through an annealing cycle (instantaneously NPT 0.6 then NPT 1.0).

Since my simulations are all at zero pressure, they don’t end up completing because memory is exceeded - the volume of the box blows up to 10^9 (this happens during the annealing process). What can be causing the system to blow up so much, is it just poor equilibration, every step is run at 1.0e6 timesteps and there aren’t any weird deviations, everything converges.

Any help would be appreciated. Thanks
– Kush

Hello all,

I have a 100 chain by 100 bead polymer system in a solution of 320 000
single beads (LJ sigma = 0.5, m = 0.125). The system is relaxed with DPD,
then NVE, NPT (first at 0.6 then 1.0), and NVT to implement the
Lennard-Jones parameters and equilibrate the box (all done at zero
pressure). Then before going through a controlled cooling, the system goes
through an annealing cycle (instantaneously NPT 0.6 then NPT 1.0).

Since my simulations are all at zero pressure, they don't end up
completing because memory is exceeded - the volume of the box blows up to
10^9 (this happens during the annealing process). What can be causing the
system to blow up so much, is it just poor equilibration, every step is run
at 1.0e6 timesteps and there aren't any weird deviations, everything
converges.

​i assume, this is with reduced units?​
​what is your time step?
how is your energy conservation with fix nve?

axel.

Are your particles purely repulsive? If so, maybe there is nothing to counteract the expansion and hence the system blows up?

Weird artifacts and explosions often happen at the beginning of NPT
simulations (the configuration of the system has not reached
equilibrium yet). Have you tried

1) running the simulation initially at high pressure and then lowering
it slowly?
(Perhaps begin the simulation a few hundred atmospheres, -just enough
to keep the volume from exploding and lower it over the course of
~10^6 timesteps, perhaps lowering it more slowly as you get closer to
the target pressure.)

2) If that fails, try minimizing and running the simulation initially
at constant volume beforehand to randomize and equilibrate the system,
Then switch to NPT conditions (perhaps initially at high pressure).

No idea if these suggestions were helpful.

-andrew