I’m simulating water molecules passing through a membrane, but I’m encountering an issue with the flow rate. While published papers report a constant flow rate, my simulation shows no initial flow in the early stages.
One possible reason I suspect is the initial setup: I’ve placed the water box 1 Å away from the membrane to avoid atomic overlaps, which might be affecting the density and flow dynamics.
Has anyone faced a similar challenge? Any advice on how to:
Maintain realistic water density near the membrane?
Ensure proper initial flow in such simulations?
I’d appreciate any suggestions or references to best practices.
Nobody will be able to help you with such vague description. We don’t even know what is your membrane, or why you think that the water density near the membrane is not realistic.
To provide further clarification, I have attached the data and input files. My inquiry focuses on two main questions:
What is the standard procedure for filling a simulation box with water molecules? I am currently using VMD and maintaining a 1 Å gap from the carbon wall. However, this approach results in an inconsistent flow rate of water molecules passing through the membrane, as reported in published studies. An alternative method is to construct the water box without the 1 Å gap and then use the delete_atoms command to remove overlapping atoms. However, this may affect the results by reducing the number of water molecules. Could you share insights on how other experts achieve a consistent flow rate in such simulations?
For the initial relaxation of the structure, I used the following commands:
However, I observe that water molecules tend to cluster near the walls, leaving fewer molecules in the center of the box (see the attached image captured at 100 ps). This gap in the middle becomes more pronounced in relaxation longer than 100 ps. Is this issue related to the relaxation technique, the nature of the system, or another factor?
Thank you for your insights and suggestions. output.txt (26.5 KB)
The “gap in the middle” is a bubble and it is formed when the condensed phase is forced to occupy an unnaturally large volume that makes it thermodynamically advantageous to form an entirely new small liquid-vacuum interface.
The solution is to put in more water or to reduce the water-available volume.
By the way, I can’t see from your screenshot how this setup is analogous to water flowing through a membrane.
I suppose the middle layer is pierced with a hole, and the pink layer will eventually move in the z direction to force the water in the right reservoir.
As a complement to the suggestion by @srtee: you can allow your pink wall to move freely along the z direction, like a piston, and even apply a force along z to the pink atoms and mimic an external pressure. Its easier than trying to insert the right number of water molecules.
