[CPMD-list] TDDFT Forces Calculation
Juerg Hutter
hutter at pci.uzh.ch
Thu May 3 09:09:49 CEST 2007
Hi
you seem to use a special version of CPMD from Lausanne.
I can't comment on this version as I have never seen it
and I don't know what has been changed.
The convergence problems of TDDFT at the point of bond
breaking are well known and there is no remedy for this
implemented in the standard CPMD code.
regards
Juerg Hutter
----------------------------------------------------------
Juerg Hutter Phone : ++41 44 635 4491
Physical Chemistry Institute FAX : ++41 44 635 6838
University of Zurich E-mail: hutter at pci.uzh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
----------------------------------------------------------
On Tue, 1 May 2007, Matteo Guglielmi wrote:
> Yes sure... here come my input file:
>
>
> ### CPMD INPUT FILE ###
>
> &CPMD
> MOLECULAR DYNAMICS BO
> RESTART WAVEFUNCTION COORDINATES VELOCITIES ACCUMULATORS LINRES LATEST
> TDDFT
> TIMESTEP
> 4
> MAXSTEP
> 65
> CONVERGENCE ORBITAL
> 1.D-6
> SUBTRACT COMVEL
> 5
> SUBTRACT ROTVEL
> 5
> LANCZOS PARAMETER
> 3000 8 -1 1.D-8
> MIRROR
> &END
>
> &TDDFT
> TAMM-DANCOFF
> LZ-SHTDDFT
> STATES SINGLET
> 6
> FORCE STATE
> 1
> DAVIDSON PARAMETER
> 500 1.D-8 50
> DAVIDSON RDIIS
> 20 3 1.D-9
> &END
>
> &DFT
> NEWCODE
> FUNCTIONAL PBE
> &END
>
> &LINRES
> CONVERGENCE
> 1.D-5
> MAXSTEP
> 1000
> HTHRS
> 0.5
> OPTIMIZER PCG
> &END
>
> &SYSTEM
> POISSON SOLVER TUCKERMAN
> ANGSTROM
> CHARGE
> 1.0
> SYMMETRY
> 0
> CELL ABSOLUTE
> 23.0 23.0 23.0 0.0 0.0 0.0
> CUTOFF
> 90.0
> &END
> &ATOMS
> *H_MT_PBE KLEINMAN-BYLANDER
> LMAX=S
> 13
> 7.730 12.096 11.316
> 9.462 13.921 11.253
> 9.111 7.294 10.855
> 12.609 9.800 10.521
> 11.568 7.565 10.519
> 7.636 9.249 11.153
> 12.091 14.082 11.128
> 12.686 12.985 9.893
> 15.167 15.084 12.221
> 13.312 11.454 13.749
> 11.804 11.485 12.898
> 12.347 12.858 13.711
> 13.716 11.487 11.433
> *N_MT_PBE KLEINMAN-BYLANDER
> LMAX=P
> 2
> 8.735 11.906 11.208
> 12.655 12.032 13.159
> *C_MT_PBE KLEINMAN-BYLANDER
> LMAX=P
> 11
> 9.703 12.876 11.170
> 10.945 12.281 10.983
> 10.712 10.855 10.908
> 9.302 10.659 11.062
> 11.529 9.722 10.705
> 10.940 8.455 10.687
> 9.553 8.298 10.868
> 8.713 9.387 11.044
> 12.241 13.026 10.905
> 13.312 12.437 11.838
> 14.533 13.294 12.190
> *O_MT_PBE KLEINMAN-BYLANDER
> LMAX=P
> 2
> 14.343 14.611 11.961
> 15.542 12.801 12.662
> &END
>
> #######
>
> The system quickly undergoes dissociation (within the very
> first ~60 steps)... that's exactly when the forces calculation
> convergence gets "unstable".
>
> With "unstable" I mean it always converges (so far) but it
> takes much more time (i.e. the number of steps is almost 4
> times bigger) than a "normal" forces calculation.
> This is also when the "oscillating" convergence behavior
> comes out.
>
>
> Ari P Seitsonen wrote:
>>
>> Ciao Matteo-Dude,
>>
>> Did you get any answer to your query? I guess that more or less the
>> only possible person to answer would be Jürg-sensei, he might have
>> been busy... But what is your system like, which functional, which
>> cut-off? I guess that it's some huge biomolecule, with a large
>> HOMO-LUMO gap, no symmetries what so ever, ...
>>
>> Quasi-Paris-always-sunny-leaving-for-a-bi-cycle-tour,
>>
>> apsi
>>
>> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
>>
>> Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
>> IMPMC, CNRS & Université Pierre et Marie Curie
>> Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820
>>
>> On Mon, 30 Apr 2007, Matteo Guglielmi wrote:
>>
>>> Hello World,
>>> Is there any keyword to play with in order to have a better
>>> convergence of the forces calculation from the excited state
>>> dynamics?
>>>
>>> Thanks,
>>> MG.
>>>
>>> ##### OUTPUT FILE ####
>>>
>>> CALCULATION OF TDDFT FORCES
>>>
>>> CPU TIME FOR INITIALIZATION OF LINEAR RESPONSE: .01 SECONDS
>>> NFI GEMAX CNORM ETOT(2) DETOT TCPU
>>> 1 .004355 .000100 -.18123358 .0000E+00 5.204
>>> 2 .001658 .000056 -.24605370 -.6482E-01 5.202
>>> 3 .001384 .000051 -.26147100 -.1542E-01 5.203
>>> 4 .001229 .000056 -.27143725 -.9966E-02 5.201
>>> 5 .001356 .000080 -.28057485 -.9138E-02 5.201
>>> 6 .001229 .000056 161.96293805 162.2 8.649
>>> 7 .000976 .000030 -.27801627 -162.2 5.203
>>> 8 .001825 .000076 -.28073816 -.2722E-02 5.201
>>> 9 .001645 .000065 -.27063144 .1011E-01 5.201
>>> 10 .001825 .000076 -61.26101933 -60.99 8.651
>>> 11 .001278 .000036 -.29556406 60.97 5.201
>>> 12 .000784 .000030 -.30045453 -.4890E-02 5.202
>>> 13 .000994 .000044 -.30416917 -.3715E-02 5.202
>>> 14 .000994 .000044 -8.26612609 -7.962 8.652
>>> 15 .000819 .000024 -.30947952 7.957 5.203
>>> 16 .000919 .000028 -.31216353 -.2684E-02 5.201
>>> 17 .003243 .000121 -.27781729 .3435E-01 5.203
>>> 18 .000919 .000028 33.55626466 33.83 8.649
>>> 19 .000647 .000019 -.31469832 -33.87 5.202
>>> 20 .000935 .000036 -.31638621 -.1688E-02 5.204
>>> 21 .001730 .000066 -.28531446 .3107E-01 5.200
>>> 22 .000935 .000036 -42.63214755 -42.35 8.649
>>> 23 .000609 .000020 -.32001373 42.31 5.203
>>> 24 .000647 .000021 -.32173429 -.1721E-02 5.201
>>> 25 .002774 .000115 -.29408352 .2765E-01 5.201
>>> 26 .000647 .000021 30.72424320 31.02 8.651
>>> 27 .000525 .000016 -.32325831 -31.05 5.202
>>> 28 .000884 .000029 -.32433396 -.1076E-02 5.201
>>> 29 .001227 .000040 -.31171156 .1262E-01 5.200
>>> 30 .000884 .000029 -58.89241894 -58.58 8.650
>>> 31 .000461 .000017 -.32686139 58.57 5.201
>>> 32 .000532 .000016 -.32799657 -.1135E-02 5.201
>>> 33 .001666 .000081 -.31711572 .1088E-01 5.204
>>> _______________________________________________
>>> CPMD-list mailing list
>>> CPMD-list at cpmd.org
>>> http://cpmd.org/mailman/listinfo/cpmd-list
>>>
>
>
> --
> +----------------------------------------------------------+
> | Address: Matteo Guglielmi |
> | EPFL-SB-ISIC-LCBC Ecole Polytechnique |
> | BCH 4121 Federale de Lausanne |
> | CH-1015 Lausanne |
> +----------------------------------------------------------+
> | Contacts: |
> | |
> | Phone --> +41-(0)21-6930323 _____ |
> | Fax --> +41-(0)21-6930320 | | |
> | Mobile1 --> +41-(0)76-5245216 ------- |
> | Mobile2 --> +39-340-2751607 0-0 |
> | Web --> http://lcbcpc21.epfl.ch c |
> | EMail1 --> matteo.guglielmi at epfl.ch +--o00o-----o00o--+
> | EMail2 --> matteo.guglielmi at gmail.com |ICanSeeClearlyNow|
> | ICQ --> #166904425 |...TheBrainIsGone|
> +----------------------------------------+-----------------+
>
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://cpmd.org/mailman/listinfo/cpmd-list
>
More information about the CPMD-list
mailing list