selective dynamics problem

Message

forwordom · #1 Post by **forwordom** » Thu Dec 07, 2006 11:58 am

The grephene belt have periodic condiction in z (axial)direction ,but it have limited width in x and y direction ,after bended ,i want to relax this bended graphene belt (about 36 atom in cell),if i have to fix edged atoms use selective dynamics ?and what values of ISIF,IBRION should be use when relaxing?
by the way if it will greatly increase caculation time and decrease caculation speed using selctive dynamics ?

tjf · #2 Post by **tjf** » Thu Dec 07, 2006 5:16 pm

[quote="'smallblacktext'>[ Edited Thu Dec 07 2006, 06:18PM "]</span>

forwordom · #3 Post by **forwordom** » Sat Dec 09, 2006 5:01 am

thank you reply and sorry for my poor english !
In fact it (graphene belt)is very like a half of carbon nanotube cut along axial (z)direction),or like a bridge in some degree.
and keep periodic boundry condition in z direction ,if i should fix edged atom ?

tjf · #4 Post by **tjf** » Sat Dec 09, 2006 7:42 pm

Ah, I see. Well, how you model it depends entirely on what you're trying to model. Obviously if you just let everything relax your graphere will (most likely) flatten out. But if you fix any coordinates and relax you're changing the physics, so you need to have a clear idea of what you want to achieve. You could, for example, fix the x coordinate of the atoms along the edge of your belt (assuming that the sheet is oriented so that the cut of the original nanotube lies in the x-z plane) using selective dynamics, then similarly fix the x coordinates of a row of atoms down the centre to maintain the curvature/v-shapedness. But what does that achieve? You're then modelling a highly stressed system, and you will be modelling a highly stressed system whenever the belt is curved, no matter what you do. Depending on what properties you want to study, I'd be inclined to not relax the system at all, but do calculations on the system with the atoms fixed in the positions they would be in were the nanotube complete.