[DFTB-Plus-User] DFTB+ 1.1 release candidate
Ran Friedman, Biochemisches Inst.
r.friedman at bioc.uzh.ch
Tue Apr 20 08:05:58 CEST 2010
Hi,
10fs is probably too large a timestep. I would be cautious even with 1fs.
Did you check energy/temperature conservation? Maybe you should study
smaller systems or use approximate methods instead of increasing the
timestep.
Good luck,
Ran
------------------------------------------------------
Ran Friedman
Postdoctoral Fellow
Computational Structural Biology Group (A. Caflisch)
Institute of Biochemistry
University of Zurich
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
Tel. +41-44-6355559
Skype: ran.friedman
------------------------------------------------------
On Tue, 20 Apr 2010 10:49:43 +0800
Jiatao Sun <sunjiatao at gmail.com> wrote:
> Dear members,
>
> I am doing Berendsen thermostat MD simulation in NVT ensemble on nano
> systems by Snapshot 081217 version.
> The Steps is 10000.
> The Timestep keyword is initially 1fs. The program terminated normally.
>The
> timestep 5fs is ok as well.
> When the Timestep is changed to 10fs or larger femtosecond, it exits
> abnormally with an error message "Failure in diagonalisation routine
>zhegvd,
> non-positive definite overlap! Minor 97 responsible." The number " 97
>"
> will change depending on the timestep. Also the program exist at different
> geometry steps depending on the different system and different Step and
> Timestep.
> I have also tried other similar nano-systems. The error messages still
>occur
> for other systems when set larger Steps (e.g. 10000) and slightly larger
> Timestep.
> It seems that 1fs timestep is the most stable without any error message.
> However 1fs is not suitable for very large system due to the very
> time-consuming.
> I have googled this error message. But have no clear solution. Any
> suggestion will be much appreciated.
>
>
> Here are the input files.
>
> Geometry = GenFormat {
> <<< "POSCAR"
> }
> Hamiltonian = DFTB {
> SCC = No
> # SCCTolerance = 1e-5
> # MaxSCCIterations = 100
> MaxAngularMomentum = {
> # Ga = "d"; As = "d"
> C = "p"
> }
> # PolynomialRepulsive = {}
> OrbitalResolvedSCC = No
> OldSKInterpolation = No
> SpinPolarisation = {}
> Eigensolver = DivideAndConquer {}
> OrbitalPotential = {}
> Dispersion = {}
> # Parametrization files
> SlaterKosterFiles = Type2FileNames {
> Prefix = "./mio-0-1/"
> Separator = "-"
> Suffix = ".skf"
> LowerCaseTypeName = Yes
> }
> # K-points sampling in the BZ
> # KPointsAndWeights = SupercellFolding {
> KPointsAndWeights = {
> 0.00000000000000 0.00000000000000 0.00000000000000 1
> 0.00000000000000 0.04166666666667 0.00000000000000 2
> 0.00000000000000 0.08333333333333 0.00000000000000 2
> 0.00000000000000 0.12500000000000 0.00000000000000 2
> 0.00000000000000 0.16666666666667 0.00000000000000 2
> 0.00000000000000 0.20833333333333 0.00000000000000 2
> 0.00000000000000 0.25000000000000 0.00000000000000 2
> 0.00000000000000 0.29166666666667 0.00000000000000 2
> 0.00000000000000 0.33333333333333 0.00000000000000 2
> 0.00000000000000 0.37500000000000 0.00000000000000 2
> 0.00000000000000 0.41666666666667 0.00000000000000 2
> 0.00000000000000 0.45833333333333 0.00000000000000 2
> 0.00000000000000 0.50000000000000 0.00000000000000 1
> }
> }
> #Additional optional definitions
> ParserOptions = {
> ParserVersion = 3
> WriteHSDInput = Yes
> WriteXMLInput = No
> StopAfterParsing = No
> IgnoreUnprocessedNodes = No
> }
> Driver = VelocityVerlet {
> TimeStep [Femtosecond] = 20.0
> MovedAtoms = Range { 1 -1 }
> Steps = 10000
> Thermostat = Berendsen {
> CouplingStrength = 0.1
> AdaptFillingTemp = Yes
> Temperature [Kelvin] = 300
> }
> OutputPrefix = "md"
> MDRestartFrequency = 100
> }
>
> Options = {
> # MullikenAnalysis = No
> # CalculateForces = Yes
> WriteEigenvectors = No
> WriteAutotestTag = No
> WriteDetailedXML = No
> WriteResultsTag = No
> WriteDetailedOut = Yes
> WriteBandOut = Yes
> AtomResolvedEnergies = No
> # RestartFrequency = 20
> RandomSeed = 0
> WriteHS = No
> WriteRealHS = No
> MinimiseMemoryUsage = No
> }
>
> POSCAR
>
> 16 F
> C
> 1 1 0.33333 0.00000 0.25000
> 2 1 0.41667 0.00000 0.25000
> 3 1 0.50000 0.00000 0.25000
> 4 1 0.58333 0.00000 0.25000
> 5 1 0.34722 0.50000 0.25000
> 6 1 0.43056 0.50000 0.25000
> 7 1 0.51389 0.50000 0.25000
> 8 1 0.59722 0.50000 0.25000
> 9 1 0.37500 0.50000 0.25000
> 10 1 0.45833 0.50000 0.25000
> 11 1 0.54167 0.50000 0.25000
> 12 1 0.62500 0.50000 0.25000
> 13 1 0.38889 0.00000 0.25000
> 14 1 0.47222 0.00000 0.25000
> 15 1 0.55555 0.00000 0.25000
> 16 1 0.63889 0.00000 0.25000
> 0.00000000 0.00000000 0.00000000
> 51.3681859436258 0 0
> 0 2.47138050151 0
> 0 0 15
>
>
>
>
>
>
> On Thu, Apr 8, 2010 at 2:53 PM, Bálint Aradi <
> balint.aradi at bccms.uni-bremen.de> wrote:
>
>> Dear List members,
>>
>> thank you very much for taking part in the activity of the
>> DFTB+ community. Your contributions help inexperienced users to make
>> their first steps with DFTB+, experienced users to find answers to their
>> specific questions, and of course your feedback helps us developers
>> to understand which parts of the code and tools need improvement and
>> which features are needed by the community.
>>
>> As sign of our appreciation, and in order to encourage you to continue
>> your active participation in the future, we would like to invite you to
>> try the release candidate for the next DFTB+ release (1.1) before its
>> official announcment. (The official 1.1 version should be released in
>> ca. 5 weeks.)
>>
>> You can download the soucre, precompiled binaries, the autotest and
>> the manual from the website as usual. You will also find a list of the
>> major changes under
>>
>> http://www.dftb-plus.info/about_dftb/release_11/
>>
>> If you have any comments or suggestions regarding the new version, or
>> if you think you having found a bug, please do not hesitate to contact us.
>>
>> We hope, that you'll find our new release useful.
>>
>>
>> Best regards
>>
>> The DFTB+ developer team
>>
>> --
>> Dr. Bálint Aradi
>> Bremen Center for Computational Materials Science, University of Bremen
>> http://www.bccms.uni-bremen.de/en/~b_aradi<http://www.bccms.uni-bremen.de/en/%7Eb_aradi>
>>
>>
>>
>> _______________________________________________
>> DFTB-Plus-User mailing list
>> DFTB-Plus-User at dftb-plus.info
>> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>>
>>
More information about the DFTB-Plus-User
mailing list