scf convergence problem |
- Date: 2010/06/19 08:42
- Name: j
<wony65@gmail.com>
- Dear All
I am trying to run bandstructure calculation of Bi2Se3.
The problem I got into is in the convergence of scf.
I used basis orbital
1s1p1d for Se
2s3p2d for Bi
and gga pseudo potential.
tags about scf run are
scf.XcType GGA-PBE
scf.SpinPolarization NC #Non-Collinear calculation
scf.SpinOrbit.Coupling ON #turn on SOC
scf.ElectronicTemperature 300
scf.energycutoff 150
scf.maxIter 200 #max number of electronic steps
scf.EigenvalueSolver Cluster
scf.Kgrid 1 1 1 #if scf.EigenvalueSolver=Band, a set of number should be provided
scf.Mixing.Type rmm-diisk
scf.Init.Mixing.Weight 0.30
scf.Min.Mixing.Weight 0.001
scf.Max.Mixing.Weight 0.40
scf.Mixing.History 20
scf.Mixing.StartPulay 8
scf.Mixing.EveryPulay 8
scf.Kerker.factor 2
scf.criterion 3.67e-8
scf.lapack.dste dstevx #default=dstevx, avoid eigenstate calculation of high energy region
I have modified all mixing tags such as Mixing.History, Mixing.StartPulay, Mixing.EveryPulay, Kerker.factor etc.
But I could not get convergence.
This is the result of run with above setting
SCF= 172 NormRD= 5.038951068186 Uele= -96.262327996637
SCF= 173 NormRD= 5.088961841267 Uele= -96.267626921219
SCF= 174 NormRD= 4.925935362699 Uele= -95.935384438308
SCF= 175 NormRD= 4.394763085380 Uele= -96.421243272637
SCF= 176 NormRD= 10.084563813317 Uele= -96.281483189469
SCF= 177 NormRD= 6.744178259390 Uele= -96.250869052439
SCF= 178 NormRD= 7.969208338267 Uele= -96.330601483736
SCF= 179 NormRD= 8.598475041693 Uele= -96.383496336167
SCF= 180 NormRD= 8.817142174165 Uele= -96.409527808628
SCF= 181 NormRD= 8.915736623649 Uele= -96.422424222339
SCF= 182 NormRD= 4.963821823390 Uele= -95.185661170874
SCF= 183 NormRD= 4.198039852247 Uele= -95.192184312103
SCF= 184 NormRD= 3.403327308878 Uele= -95.278966589133
SCF= 185 NormRD= 4.707285610412 Uele= -95.549843213043
SCF= 186 NormRD= 4.909433424596 Uele= -95.627969484908
SCF= 187 NormRD= 5.639274323026 Uele= -95.684105180358
SCF= 188 NormRD= 6.104124920275 Uele= -95.711168560752
SCF= 189 NormRD= 6.350527575868 Uele= -95.724533057399
SCF= 190 NormRD= 5.957950724543 Uele= -95.101351180719
SCF= 191 NormRD= 6.339959078313 Uele= -95.235343439018
SCF= 192 NormRD= 3.722661879834 Uele= -95.368869344811
SCF= 193 NormRD= 6.317657654993 Uele= -95.638888136259
SCF= 194 NormRD= 7.361285540261 Uele= -95.734719861572
SCF= 195 NormRD= 7.675430899139 Uele= -95.778203522343
SCF= 196 NormRD= 7.805854892847 Uele= -95.799369496322
SCF= 197 NormRD= 7.866555851226 Uele= -95.809817323164
SCF= 198 NormRD= 9.755482668392 Uele= -95.714790254981
SCF= 199 NormRD= 9.203465350333 Uele= -95.714884755332
SCF= 200 NormRD= 7.612337044205 Uele= -95.831988293847
Could some one give me any ideas for better convergence?
Thanks.
 | |
This thread is locked.Only browsing is available.
Re: scf convergence problem