[gmx-users] pressure scaling

[Mu Yuguang (Dr)]

Dear all,
May it be some problems not related to the pressure coupling?

Nicolas, did you try without pressure coupling, still crash ?

I found in my simulations, some trajectories crash repetitively without pressure coupling. The phenomena I found is that such crash usually happens when my molecule cross the box border.
I solved it in the following way:
I reorganize the simulation system, by placing the solute (usually big) in the center of the simulation box, using trjconv command, then the simulation can go on.
  
Such crashes do not dependent on force fields used, PME or cut-offs, I use simple cubic box.

I think maybe some bugs in the treatment of periodic conditions ??

Regards,
Yuguang
 
Dr. Yuguang Mu
School of Biological Sciences
Nanyang Technological University
Singapore 637616
Tel: 0065-63162885
 

-----Original Message-----
From: gmx-users-bounces at gromacs.org [mailto:gmx-users-bounces at gromacs.org] On Behalf Of nicolas Dinter
Sent: Tuesday, August 03, 2004 1:51 PM
To: Discussion list for GROMACS users
Subject: RE: [gmx-users] pressure scaling

Dear Berk,

I used a tau_p = 10, then I tried different values for compressibility:
1st:4.5e-5 (initial crash), 14.5e-5, 0.5e-5, 0.5e-6, 0.5e-4 and 4.5e-3. But
there isn't many changes, the cash occured allmost at the same step: 303
(4.5e-5),291 (14.5e-5),301 (0.5e-5),301 (0.5e-6),299(0.5e-4) and 293
(4.5e-3).
Ok so if I undestood well the pressure scaling error occurs when the box
enter in resonance, but normaly with low compressibility it must not occur
or at least at different steps (more difficult to change the box) especialy
with low pressure.

So I would like to know which are the parameter which drive this error:
there is the time step (I took:0.0005ps), the tau_p parameter (I took 10ps)
and the compressibility. And what else?

I still not well undestand what does represent tau_p is my draw is a correct
understanding? (see atached file).

Thank for any help.

Nicolas.


-----Original Message-----
From: gmx-users-bounces at gromacs.org
[mailto:gmx-users-bounces at gromacs.org]On Behalf Of Berk Hess
Sent: lundi 2 août 2004 18:18
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] pressure scaling


A smaller tau_p is not the solution.
If anything, it should be larger.
What matters in Berendsen pressure coupling is the ratio:
compressibility/tau_p
So even if tau_p seems reasonable, the scaling may go wrong
due to a too large compressibility value.
I don't know which value you have used and which value you
should use for your system.

>
>   Try to use lower tau_p (maybe 3ps?).
>On Aug 2, 2004, at 9:31 AM, nicolas Dinter wrote:
>
>>
>>Dear Gromacs users,
>>
>>I am performing simulation on diphenyl carbonate and I get the presure
>>scaling error.
>>I used a small pression about 1 bar, tau_p = 10 ps and my time scale is
>>0.0005 ps.
>>The NVT simulation runs well, to solve this problem for hight pressure is
>>easier but low pressure knowing that this error is related with tau_p and
>>timestep I would like to know how to fix it.
>>Thank you.
>>
>>nicolas

_________________________________________________________________
Talk with your online friends with MSN Messenger http://messenger.msn.nl/

_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.