Hello,
I have done previous MD stress strain tests on bulk (p p p BCs) and have discovered that NPT produces more realistic results than NVT as NPT allows for lateral contractions during deformation – understandable.
My next goal is to deform a nanowire, however in this case the wire typically has periodic at along its length and then shrinkwrapped in the other two dimensions (creating a free surface). However, once I have non-periodic BC I cannot use fix NPT on those dimensions. Is there a way to observe the lateral contractions of the dimensions not in the direction of straining without using NPT?
One method I have seen in literature is when cutting the cylindrical nanowire to leave a vacuum area, thus creating the free surface and allowing the wire to contract during deformation. Does this make sense?
Thanks,
Stephen
Hello,
For your question in your second paragraph, I think you just need to dump the positions for a chosen group with a given frequency. Then you can calculate the size of the cross section by yourself.
Your last paragraph is not very clearly written. You may need to rephrase it.
Best,
Bruce
Stephen Handrigan <stephen.handrigan@…1064…202…> 于 2018年12月5日周三 18:30写道：
I have told you how to observe the lateral deformation: simply dump some coordinates into disk and check them by writing a small piece of code.
Using a vacuum layer is no different from using the s boundary conditions in LAMMPS (please correct me if I am wrong here). From a physical point of view, they both correspond to free (open) boundary conditions.
With free boundary conditions, there is no way ( and no need) to control the pressure in the corresponding directions. In your problem, there is no need to control the pressure in the transvers directions if you are studying a suspended (free standing) nanowire. If your nanowire is embeded into some other materials, that is another story.
Best,
Bruce
stephen.handrigan <stephen.handrigan@…7202…> 于 2018年12月5日周三 22:34写道：
It is known that during tension testing there should be deformation in the transverse dimensions (Poisson’s effect). NPT allows for this to happen, but NVT does not.
This is wrong for nanowire systems, because the volume of a nanowire is not necessarily the volume of the box. Therefore, using NVT does not mean the volume of the “system” is fixed. It only means the volume of the artificial box is fixed. Using NVT and fix deform, you can definitely make the nanowire longer and thinner. Just dump some coordinates out and analyze or simply visualize them!
Best,
Bruce
Bruce Fan <brucenju@…24…> 于 2018年12月5日周三 22:52写道：