[gmx-users] Non-periodicity in only one direction

[David van der Spoel]

On Mon, 31 Dec 2001, Rui Qiao wrote:

>Hi, everyone:
>	I am doing simulation of flow in a nano-channel right now. Becuase
>the flow is confined in channel, the periodicity in one direction (let it
>to be z). Because of the nature of simulaiton, electrostatic interaction
>is needed.
>	Someone has suggested using proper cut-off as a workaround, but
>after I tried, it seems to me that the cut-off for the electrostatic is
>not enough to recover the physics, i.e. I must use full electrostatics to
>get reasonable results.
>	Right now, I use a huge simulation box with 6 times channel width
>in z direction to remove the periodicity. However, it is very inefficient,
>and it is unlikely that I can afford it later. So I think I need to do
>something about the code to work this out. Thus two questions:
>	1. What algorithm should I use. For the PPPM, there exists some
>algorithm but since PPPM is not recommended in pressure coupling, this is
>not an option. For the PME, I am not sure if there exists some algorithm.
>	2. Which code should I revise. 
Check the literature first, I am not sure that the Ewald sum applies to
two dimensional systems without modification. You may have to modify some
equations. Otherwise all code is in pme.c (except for the fourier
transforms). It would be quite interesting to simulate interfaces with two
dimensional periodicity as well.


Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
Husargatan 3, Box 576,  	75123 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://zorn.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++