Dear Rajan,as you speak about a ribbon, it looks probable that you'll have
dispersionin one dimension only, so the k-mesh (to test the convergenceof
results) will be linear and hardly prohibitively large.I don't understand what
you mean by "relax my system at each k-point".If you want to test the
convergence of results as the number of k-pointsincreases, you don't really
need to relax the structure... Anyway, the result of the convergence test would
depend on which properties exactlyare you after. To plot energy vs number of
k-points would hardly be usefulby itself, because it is a function which goes
down and down and down...You must know where to stop. You'll not be able to
"evaluate optimized k-point"from the plot.Good luckAndrei Postnikov
----- RAJAN SINGH <singhrajan.i...@gmail.com> a écrit :
>Dear Siesta Users> I have a fairly large system of about 50 atoms of graphene
>ribbon. I want to optimize the same in terms of k-points and mesh cutoff.> But
>if I relax my system at each k-point, It will cost too much time.> My question
>is> Can I skip the relaxation fr each k point??> And plot Energies vs kpoint n
>evaluate optimized k-point??> Please help me in this.With Regarda>Rajan