[Wien] force optimization with MSR1a

2014-08-28 Thread shamik chakrabarti
Dear wien2k users, We are doing force minimization of a spinel compound using MSR1a. After even 80 iterations, force on some atoms are 200 mRy/a.u. while on some atoms it is ~ 7 mRy/a.u. We are using a mixing grid of default 0.2 in case.inm. My query is that should we

Re: [Wien] force optimization with MSR1a

2014-08-28 Thread Laurence Marks
In 99.9% of cases this indicates something wrong with your model, e.g. bad RMT, RKMAX or other. Without more information that is all anyone can say. On Aug 28, 2014 9:46 AM, shamik chakrabarti shamik...@gmail.com wrote: Dear wien2k users, We are doing force

Re: [Wien] force optimization with MSR1a

2014-08-28 Thread Peter Blaha
80 iterations for a more complicated MSR1a optimization is NOT much. I often need a couple of 100 iterations for a big (metallic and magnetic) system with 3d elements. However, make sure: grep :DIS case.scf is showing a reasonable convergence (eg. it should not stay long above 1.0) grep :ENE

Re: [Wien] force optimization with MSR1a

2014-08-28 Thread shamik chakrabarti
Dera Prof. Blaha Sir, Thank you very much for your reply. It is indeed very helpful. Yes energy convergence has been reached up to ~ 0.001 so far. Also E-tot FR is decreasing. We are using 14 k-points for a 56 atomic cell. However, there is one doubt...the material is found to be

[Wien] (no subject)

2014-08-28 Thread sikandar azam
Dear All please help me I am making the supercell, when i give  x supercell command  then i get this problem Filename of struct file: case.struct  Fatal Error occured:  Unknown lattice type: CYZ  Program terminated. 0.000u 0.008s 0:14.57 0.0%      0+0k 0+0io 0pf+0w with regards

[Wien] (no subject)

2014-08-28 Thread sikandar azam
Dear All please help me I am making the supercell, when i give  x supercell command  then i get this problem Filename of struct file: case.struct  Fatal Error occured:  Unknown lattice type: CYZ  Program terminated. 0.000u 0.008s 0:14.57 0.0%      0+0k 0+0io 0pf+0w with regards sikander

Re: [Wien] (no subject)

2014-08-28 Thread Oleg Rubel
If you would resubmit your request to the mail list WITH the structure file, someone can possibly help. Oleg On Thu, Aug 28, 2014 at 9:19 AM, sikandar azam sikandar...@yahoo.com wrote: Dear All please help me I am making the supercell, when i give x supercell command then i get this

Re: [Wien] (no subject)

2014-08-28 Thread Tomas Kana
Dear Sikander, The program supercell just tells you that it does not want  CYZ lattice.  You should rewrite your original cell in a Primitive lattice  (P instead of CYZ). Your original CYZ lattice is base-centered.  Rewrite the CYZ (or the number of space group) in w2web to P. Then you must

Re: [Wien] (no subject)

2014-08-28 Thread delamora
You could also try rotating the cell to have a different axis, x=y, y=z, z=x or x=y, etc., and maybe supercell will accept it. De: wien-boun...@zeus.theochem.tuwien.ac.at wien-boun...@zeus.theochem.tuwien.ac.at en nombre de Tomas Kana k...@seznam.cz Enviado:

Re: [Wien] (no subject)

2014-08-28 Thread Kevin Jorissen
And if you are a matlab/octave user, I believe there is an octave program that can be used as an alternative, and that is more robust. (on wien2k website) Cheers Kevin On Thu, Aug 28, 2014 at 8:34 AM, delamora delam...@unam.mx wrote: You could also try rotating the cell to have a

[Wien] problem with eece

2014-08-28 Thread Hemza Kouarta
Hi, i use HF function on the 3d metal (Fe) intercalated in NbSe2 ,my calculation is fine and converged,but when i do : x lapw2 -up -so -eece -qtlto calculate dos ,they give me this error : WARNING: EF not accurate, new emin,emax,NE-min,NE-max -2.96097358895238 -2.96097358242661

Re: [Wien] problem with eece

2014-08-28 Thread Kevin Jorissen
mistake in your input? On Thu, Aug 28, 2014 at 5:28 PM, Hemza Kouarta hemzakouarta@gmail.com wrote: Hi, i use HF function on the 3d metal (Fe) intercalated in NbSe2 ,my calculation is fine and converged,but when i do : x lapw2 -up -so -eece -qtlto calculate dos ,they give me this