Atom missing when the vacuum space was added

Dear users:

I am simulating the surface tension of liquids using LAMMPS. So the interface should exist.

In the previous simulations, I often output data after equilibrating the system. Then, to produce the interfaces, the vacuum space was added by artificially changing lo and hi in the datafile. Finally, the produced datafile was read by the inputfile to calculate surface tension using NVT ensemble. In some situations, the programs could work well. However, in some other situations, atom missing was reported after importing the artificially modified datafile by the read_data command. I also tried to use energy minimization when atom missing happened. But after applying minimization, sometimes the atom missing problem still happened. I don’t know how to solve this problem.

In addition, I also tried the change_box command to produce the vacuum space like this

change_box all z delta -35.5 35.5 units box

Atom missing was also obtained. I found some bonds were extended according to snapshots. So it’s no wonder the atom will lose.

4E1322F4@...6386....jpg

To make my meaning more clear, I conclude my questions here. Firstly, after the system was equilibrated and the vacuum was added, if minimization was then used, did that have an effect on the final simulation results such as the density profile of the system? Secondly, how to avoid atom missing when vacuum spaces should be added and what measures could be taken when energy minimization can’t solve this problem. Could someone give me any suggestions?

Best regards,

Tingting Chen

300_image.0010100.jpg

When you created vacuum by extending the box dimensions along the z direction, what happened to the molecules that spanned across the periodic boundary? They got cut into fragments, right? So you should do something differently in 1) your equilibration stage where you don’t allow molecules to wrap around the PBC along z, or 2) when you create vacuum by not cutting through molecules (to wrap molecules around the PBC when increasing the z dimensions.

Ray

4E1322F4@...6386....jpg

Thank you very much. I will think about your suggestions carefully and have a try.

4E1322F4@...6386....jpg