I am trying to make a polymer melt of Polyethylene at 150+ deg. C. (No solvent in the system)
FF: OPLS-AA (Topology created using pdb2gmx)
Important Parameters for EM:
integrator = steep
emtol = 50.0
emstep = 0.01
nsteps = 50000
cutoff-scheme = Verlet
compressibility = 0.12095
gen-temp = 423 ; 150 Deg. C
There is no problem in EM. I get the energy minimised structure within 100 steps. NVT simulation works well too.
However, for NPT simulation, the system blows up. The angles rotate more than 30 degrees, and the table limit error shows up.
-------------------------------------------------------------------------------------------------
Error:
WARNING: Listed nonbonded interaction between particles 8022 and 8029
at distance 3f which is larger than the table limit 3f nm.
This is likely either a 1,4 interaction, or a listed interaction inside
a smaller molecule you are decoupling during a free energy calculation.
Since interactions at distances beyond the table cannot be computed,
they are skipped until they are inside the table limit again. You will
only see this message once, even if it occurs for several interactions.
IMPORTANT: This should not happen in a stable simulation, so there is
probably something wrong with your system. Only change the table-extension-
distance in the mdp file if you are really sure that is the reason.
---------------------------------------------------------------------------------------------
Is it due to the high compressibility value of PE (0.12095 bar^-1) [Water compressibility is 4.5e-05] ? How can I rectify this issue?
Thanks.
What is the pressure calculated during NVT?
Always blame the parameters, instead of the molecules. What have you set for 1-4 interactions? What about bond-angle-dihedral parameters? Did you make an unit check?
The other possible reason is too fast volume change (yes, due to the high compressibility). The value of compressibility affects the result only in conjunction with characteristic time constant. Try changing compressiblity to 4.5e-5 (the physical meaning does not really matter) and set tau_p to 1 ot even greater. Does the error occur again?
Thank you for the reply @Marcel Balcik, and Alexey Sir.
Alexey sir, the simulation works well at 4.5e-5 and for tau_p = 2. I guess you are correct. I am planning to do simulations for 1 ns and increasing the compressibility with each simulation to reach to the final value of compressibility in 10-12 simulations.
Hope it works.
Through gmx energy, one can find the isothermal compressibility by using -fluct_props flag. That compressibility has always been very different compared to the one given in mdp file. Is it because of no solvent? Or is that the value one gives for compressibility in the mdp file doesn't account for much? Also, if one gives the value of compressibility then why does the calculated value differs greatly?
@Marcel Balcik: 1-4 interactions and bond-angle-dihedral are the ones that pdb2gmx generates. I will look more in to what you've suggested. Thanks.
What is the origin of the compressibility value you used? Is it an experimental data? For polymers compressibility at macroscopic time and length scales may be not the same, as for the small simulation box and short times of MD simulation. The compressibility extracted from calculation may suits better for you purposes. But the numerical estimation must be performed for dense system, so finish your relaxation first. At any rate, the most straightforward way is not to calculate any compressibility, but use some model value (4.5e-05 for example), since it does not change average volume.
Yes, the compressibility value was taken from an experimental data. I get your point.
I will start with the model value and then calculate the com'bility value through gmx energy -fluct_props. Will then run parallel sims on the model and the obtained value to see if there is any difference .
Thank you for all the help.
- Badges
- Science topic