Re: [gmx-users] Pressure coupling constants

2013-12-13 Thread kpsanto
got you. Thanks again! Santo - K. P. Santo Post doctoral fellow Department of Chemistry University of North Carolina at Chapel Hill Chapel Hill, NC, USA -- View this message in context: http://gromacs.5086.x6.nabble.com/Pressure-coupling-constants-tp5013292p5013308.html Sent from the GROMAC

Re: [gmx-users] Pressure coupling constants

2013-12-12 Thread Michael Shirts
You're right, NVE would probably be better! Otherwise, you would not get expansionary cooling. On Thu, Dec 12, 2013 at 7:08 PM, kpsanto wrote: > guess that right! but why do you think NVT scaling of velocities is physical > which should also affect the non-equilibrium expansion process? > > than

Re: [gmx-users] Pressure coupling constants

2013-12-12 Thread kpsanto
guess that right! but why do you think NVT scaling of velocities is physical which should also affect the non-equilibrium expansion process? thanks in advance Santo - K. P. Santo Post doctoral fellow Department of Chemistry University of North Carolina at Chapel Hill Chapel Hill, NC, USA --

Re: [gmx-users] Pressure coupling constants

2013-12-12 Thread Michael Shirts
The most physical thing would be to prepare a system in a small NVT box, then make the box bigger along one dimension, then do an NVT simulation of that system and see what happens along that dimension over time. Barostats are simply not designed to do nonequilibrium right. They magically create v

Re: [gmx-users] Pressure coupling constants

2013-12-12 Thread kpsanto
Hi Thanks a lot. But if one wants to simulate the sudden expansion (in a crude way ,of course) smaller coupling constant would be more appropriate. Isn't it? santo - K. P. Santo Post doctoral fellow Department of Chemistry University of North Carolina at Chapel Hill Chapel Hill, NC, USA -- V

Re: [gmx-users] Pressure coupling constants

2013-12-12 Thread Michael Shirts
If the goal is to get the system to relax from a compressed state as part of preparation, then berendsen pressure coupling is the best way. However, it is not a physical method -- it overly constraints the volume distribution (which is one of the reasons it's good to force a system into a reasonab

[gmx-users] Pressure coupling constants

2013-12-12 Thread kpsanto
Hi all I am using NPT ensemble to relax a highly compressed system. I find that the way it relaxes depends on the pressure coupling constant ( berendsen barostat). My question is if the choice of the coupling constant would depend on the system conditions? The compressed system would tend to at ex