I just want to quickly add to this discussion, as it may be useful to others. The acceleration here really appears to be in the sense of p = rho*g*h, which is hydrostatic pressure at the bottom of e.g. water slab of height h under "gravity" described by g.

In other words, knowing the box length in the direction of the applied pressure allows you to precalculate the value of g, which is the acceleration in question. It is an entirely fake parameter in this case (and it must be set judiciously). That said, one should not be thrown off by values like 10^13 m/s^2 for ~tens of MPa across nanoscale systems. Entertainingly, the units (nm/ps^2 = 10^15 m/s^2) make the input values appear much less insane.


On 6/12/2018 11:45 AM, Alex wrote:
Okay, that's a reassuring start. I guess I will just have to try and track the cumulative flux to see if it's constant. I wouldn't want the flux to be increasing throughout simulation. Never used this option, obviously, so thanks for the tip!


On 6/12/2018 11:38 AM, Justin Lemkul wrote:

On 6/12/18 1:37 PM, Alex wrote:
Hi Justin,

Looking at acc-grps text in the manual, it's a bit unclear how a constant flow rate would be achieved in this case. Is a constant force in the amount of molecule mass, times the acceleration vector applied to each molecule in the group, or is there actual acceleration magically maintained?

I'm not familiar with the code, but it's what I know people have used to impose a constant force on a bulk solvent without having to use the pull code (which will throw errors).




On 6/12/2018 8:24 AM, Justin Lemkul wrote:

On 6/11/18 7:38 PM, Alex wrote:
Hi all,

I'm trying to explore pressure-induced processes, i.e. externally induced water flow across a porous membrane that spans the entire box in XY. There
are obvious choices:

1. Introduce the same type of a membrane without pores and make it a piston with pull code, e.g. constant force or tiny constant velocity along Z.
2. Apply constant force pull directly to the entire solvent group.

Any other ideas on osmosis-like processes? Has anyone tried any of the
options above?

I would think the pull code would be more headache than it's worth in this case. I'd simply try inducing the flow with acc-grps.


Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Reply via email to