Re: [gmx-users] Box size increases in NPT

2013-10-23 Thread Mark Abraham
By "crash" I meant "explode" not "DD is impossible." Explosions don't happen because of parallelism, they happen because the steps are too large for the size of the forces. The forces required to stably expand a box from 20A to 70A seem likely to be so large that I am very skeptical that you could

Re: [gmx-users] Box size increases in NPT

2013-10-23 Thread Dr. Vitaly Chaban
If the job is not "very parallel", it will not crash. It is better to preequilibrate in NVT beforehand. Cyclopropylchloride is probably a liquid at 290K, if the model is parametrized reasonably. So it should not phase-separate. Vitaly On Wed, Oct 23, 2013 at 11:29 AM, Mark Abraham wrote: > On

Re: [gmx-users] Box size increases in NPT

2013-10-23 Thread Mark Abraham
On Oct 23, 2013 5:34 AM, "Nilesh Dhumal" wrote: > > Hello, > > I am running a NPT simulation for cyclopropylchloride(1) in > 50%water(100)+50%ethanol(100) using opls force field parameter . > > After equilibration box size increases from 20 A to 70 A. Really? Seems wildly unlikely to have occurre

[gmx-users] Box size increases in NPT

2013-10-22 Thread Nilesh Dhumal
Hello, I am running a NPT simulation for cyclopropylchloride(1) in 50%water(100)+50%ethanol(100) using opls force field parameter . After equilibration box size increases from 20 A to 70 A. I used the following mdp file. ; RUN CONTROL PARAMETERS = integrator = sd ; start time and