The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear Nuno Galamba,

The fact that the energy does not converge in 50 SCF steps is very  
strange for NaCl, that
should be a good insulator. In principle, there should be no  
influence of one atom being
with xcart > acell, since the simulation is periodic.

I would suggest that you take a better look at the specific geometry  
where this lack of convergence
happens. Could you perform distinct fixed geometry calculations, to  
understand the electronic structure corresponding
to this problematic geometry ? For that geometry, you might try to  
place the atom with xcart > acell
to the corresponding position inside the cell. You should see that  
nothing has changed (or this is a bug in ABINIT).

Also, could you set nband, by hand, to a value larger than the one  
automatically chosen by
ABINIT, in order to see what is the size of the gap (also use nbdbuf  
to avoid other type of convergence problems).
In order to force the convergence, you might have to play a bit with the
parameters that govern the SCF cycle : try iscf 3 instead of iscf 5 ;  
try diemac 2 or 3 , or even more . Note
that you have been using parameters (diemac, diemix) that are  
suggested for a molecule in a big box, while you are working
with an ionic condensed system, that has definitely a larger  
dielectric response.

If it happens that your system is becoming metallic (this would be  
surprising, but ... who knows ...), you would have
to use another value of occopt to do your simulation.

Finally, I would also suggest to chose a tolerance criterion based on  
toldff instead of the one based on toldfe :
you want your forces to be consistently accurate, and this might not  
be guaranteed by the convergence on
the total energy.

Good luck,
Xavier Gonze



On 26 Sep 2005, at 19:58, ngalamba@cii.fc.ul.pt wrote:

> Dear ABINIT users
>
> (Thanks for you answer to my first question Anglade)
>
> In fact the problem I have runned into in my MD simulations of  
> ionic systems (NaCl) is that after a
> number of time-steps (about 150) the total energy stops being  
> conserved and
> the energy does not converge (even within 50 SCF cycles). At first  
> I thought
> this may had something to do with the way I was linking the partial  
> runs, but
> now I am positive it does not as it should. In fact the differences I
> describe are only a matter of my energy convergence tolerance as it is
> normal.
>
> I notice that when convergence is not achieved and the total energy  
> is lost a
> particle comes out the boundary of my supercell, i.e. xcart > acell  
> in one or
> two space dimensions.
>
> I have been trying different time-steps and energy tolerances but  
> without much
> success. Perhaps anyone could suggest me possible causes for this  
> behaviour.
> Below its my my complete input file. I am using the LDA Troullier- 
> Martins
> pseudopotentials.
>
> Maybe I am just doing something terribly wrong and did not notice.
>
> Thanks in advance
>
>
> Nuno Galamba
>
>
>
> INPUT FILE:
>
>
> #Definition of the unit cell
> acell 3*24.76786
>
> ixc 1             # Integer for eXchange-Correlation choice
>                   # 1=> LDA or LSD, Teter Pade parametrization
>                   # 11=> GGA, Perdew-Burke-Ernzerhof GGA functional
>
> #Definition of the atom types
> ntypat 2          # There are two types of atom
> znucl 11 17       # The keyword "znucl" refers to the atomic number  
> of the
>                   # possible type(s) of atom. The pseudopotential(s)
>                   # mentioned in the "files" file must correspond
>                   # to the type(s) of atom. There are two types, Na  
> and Cl.
>
> #Definition of the atoms
> natom 64           # There are 64 atoms
>                    # Na is type 1 Cl is type 2
> typat 1 1 1 1 1 1 1 1
>       1 1 1 1 1 1 1 1
>       1 1 1 1 1 1 1 1
>       1 1 1 1 1 1 1 1
>       2 2 2 2 2 2 2 2
>       2 2 2 2 2 2 2 2
>       2 2 2 2 2 2 2 2
>       2 2 2 2 2 2 2 2
>
> xcart              # This keyword indicate that the location of the  
> atoms
>                    # will follow, one triplet of number for each atom
>   5.65610083276636E+00  6.93007271203097E+00  3.98004166996406E+00
>  -1.58868622079202E+00  2.63595639332169E-01  5.57671342725475E+00
>  -1.49430205342578E+00  8.27373084310587E+00  4.28663901715516E+00
>   6.20932440764681E+00  6.19744320361232E-01  3.81901561805629E+00
>   1.66283625121153E+01  5.40662784441812E+00  6.76384828606851E+00
>   1.19212420178225E+01  1.74664875846193E+00  8.84595707944380E+00
>   1.36798041986300E+01  5.19789720677238E+00  1.18842635240986E+00
>   1.57698075022031E+01 -2.47088396806683E+00  1.71827272598309E+00
>   8.82385488253922E+00  1.93800507464858E+01  8.03045840728254E+00
>  -2.96940590475408E-01  1.43296676103769E+01  7.94873369769508E+00
>   2.75193713439659E+00  1.66809052988040E+01  2.21186590278981E+00
>   5.75064910174398E+00  1.11428516068944E+01 -1.19206927353112E+00
>   1.78970548725394E+01  1.96098850528562E+01  8.36735017031766E+00
>   1.32951507724950E+01  1.30273516304540E+01  8.66302964244192E+00
>   1.09075464378747E+01  1.74430765618708E+01 -7.18628166590436E-01
>   1.79309814303632E+01  1.32659758625039E+01  2.78007618501099E+00
>   2.01757826162747E+00  3.70882481536942E+00  2.13777333539610E+01
>   5.45781986810109E-01 -8.53595129266538E-01  1.58192342348621E+01
>   2.58739506332547E+00  8.57151174405726E+00  1.16833167489726E+01
>   6.31615279367772E+00  3.21631280617979E+00  1.48884408706589E+01
>   1.80911588944757E+01  3.96995354871461E+00  2.15402530489654E+01
>   1.03525180976381E+01 -1.06550027411973E+00  2.04211691806326E+01
>   1.37565960570218E+01  5.51216185937855E+00  1.52867729052235E+01
>   1.74905236377205E+01  2.79139073706610E-01  1.22055002724973E+01
>   6.06615778959029E+00  1.87655731011766E+01  1.71278047467966E+01
>   3.27913887860641E-01  1.41249716583982E+01  1.71225139608150E+01
>   2.32242959524589E+00  2.04172925233845E+01  1.12789799054557E+01
>   8.64607887529689E+00  1.20514391946926E+01  1.71318269534357E+01
>   1.96817656859490E+01  1.93523059315386E+01  1.86473997160438E+01
>   1.64319052040267E+01  1.24034827539144E+01  1.83563535673751E+01
>   1.31683971038350E+01  2.05498272275597E+01  1.41933514996297E+01
>   1.86645236989813E+01  1.04990490908508E+01  1.24197018270058E+01
>  -3.44974273400398E-01  3.39183432946808E+00  1.26158860130092E+00
>   1.21850776103177E+00  5.38621021103299E+00  7.27867587909737E+00
>   6.71668083259319E+00 -3.04893236900015E-01  9.19282809484037E+00
>   5.52626261040444E+00  6.68068005936578E+00 -1.33470871179536E+00
>   1.10729252318763E+01  7.34629036967786E-01  1.38425105592482E+00
>   1.22420458091739E+01  5.42730478887774E+00  5.64406749630738E+00
>   1.64808041637378E+01  1.04623190409022E-01  7.05990714953118E+00
>   1.81489484769993E+01  5.44020577061157E+00  1.48758605236480E+00
>  -1.00712948377719E+00  1.30466527849894E+01  3.24896054062410E+00
>   2.85687073048752E+00  1.87221152967594E+01  6.64479963145403E+00
>   6.18553802180375E+00  1.13751327461171E+01  3.94029445346575E+00
>   6.81250206482103E+00  2.00454488296633E+01  3.00772892811612E-02
>   1.54649514651704E+01  1.38577538331537E+01 -1.98271195022125E+00
>   1.27794240773162E+01  1.74400126309805E+01  7.94295780064528E+00
>   1.73868429705585E+01  1.05386873127956E+01  6.32376714120894E+00
>   1.97748603591924E+01  1.85203159628838E+01  3.55269786882534E+00
>  -1.77300902536039E+00 -1.05286467903429E+00  9.76102740263670E+00
>  -2.87794597891296E+00  1.16086992092086E+00  1.83754896486290E+01
>   5.60009863356821E+00 -3.55360184993267E-01  1.82504568294007E+01
>   5.58914364795695E+00  8.10228710234038E+00  1.51389875002830E+01
>   1.15794247479560E+01  1.52694144559577E+00  1.37711277243081E+01
>   1.28383634052612E+01  5.11571607980042E+00  2.01284534540082E+01
>   1.69508684302737E+01 -1.04812430369407E+00  2.11689116152223E+01
>   1.77409434748636E+01  5.30549495482421E+00  1.15541098433276E+01
>  -1.12188435476261E+00  1.08626174281289E+01  1.11423767086881E+01
>   2.89157679340877E-01  1.87787040490874E+01  1.61738341528294E+01
>   4.98236213443105E+00  1.46260682731989E+01  1.88044193480543E+01
>   7.47991890414865E+00  1.63109639380093E+01  1.22188263902333E+01
>   1.42863296997519E+01  9.66031303328476E+00  1.28482069102145E+01
>   1.05373826482952E+01  1.75813280969912E+01  1.85699286439348E+01
>   2.02139995948367E+01  1.42926205456365E+01  1.58161454757025E+01
>   1.78909497953517E+01  1.99708303037689E+01  1.28418647876767E+01
>
> vel
>   3.61345755190633E-04  1.59222580487926E-04 -2.77175074102560E-05
>  -7.94124752426413E-04  2.62143756552861E-05  2.46005846575847E-04
>   6.06414817868265E-05  4.86080442874640E-04 -5.69213988070817E-04
>   1.94696258009353E-04  7.56388019803739E-05  1.55924145855197E-04
>  -6.81305347480819E-04 -6.88615541563244E-05  7.60960234425053E-04
>   4.26915966036115E-04  1.66635414244097E-04 -2.32659334756115E-04
>  -1.99991078188162E-04  1.91948865478680E-04 -1.49301991428125E-04
>   4.84155499711047E-05 -8.57300783824437E-05  2.18255753050848E-05
>  -3.03532143194674E-04 -1.43398741615671E-04  1.16603320920703E-05
>  -2.58727287730702E-04  2.17402859081001E-04  4.51481556108830E-04
>  -3.43055308460619E-04  1.90754552449349E-04  1.78812746828536E-04
>  -4.31953399259863E-04  3.62309251085063E-05 -4.55754759246047E-04
>   6.40081806837475E-04  6.51547571151856E-04  1.00072373719001E-04
>   5.44381826373121E-04 -2.70740536621591E-04 -1.91607989883120E-04
>  -2.48358368769158E-04 -8.74190122605132E-04 -1.52884001727244E-05
>  -3.60419313307425E-04  1.07639507430958E-03  1.61553411068105E-04
>   2.13141742631626E-04  1.28701984080752E-05  2.58948971472996E-04
>   2.57577902732932E-04  7.09749198743148E-04 -9.15557293102000E-05
>   2.36831507122643E-04 -5.51371441063780E-04 -1.82064948476860E-04
>  -8.47405521228355E-06  5.49285683128093E-04 -3.81845184558150E-04
>  -4.11036573594706E-04 -2.12672547690139E-05  6.78055500494184E-04
>  -2.43250732701722E-04 -1.86492505154946E-04  3.87732680303958E-04
>   4.59852228359158E-04 -2.28067503063192E-04  2.87374293466821E-04
>   3.79436388130732E-04 -1.21622520562676E-04  4.28857628326593E-04
>  -5.66732944218849E-05  3.82666891466614E-05  3.55185691916205E-04
>   1.62940953891326E-04  4.10183379721428E-04 -7.30677721712143E-04
>  -8.55537849604790E-05  2.48522310391418E-04 -6.39272002565119E-05
>   6.91922912594730E-04  3.79786181579471E-04  2.85738195646260E-04
>   3.48876992080368E-04  7.96703170255665E-05  5.82640845293898E-04
>   7.03013539061508E-05  1.66492540632092E-04 -2.64740054178196E-04
>  -4.81057971763449E-04 -4.02538405413973E-04  5.91208606679974E-04
>   6.84660363085143E-05 -6.31044301666006E-04  2.77728904985394E-04
>  -7.86667076780607E-05 -9.13710047240675E-05 -3.05484958554519E-04
>  -3.14497134487516E-04 -1.44788385895228E-04 -2.53397869895614E-04
>   1.91776118159379E-04  6.20323794597650E-05  8.96482897085823E-06
>  -4.89657996883443E-04 -8.03974639952289E-05  1.44722036361834E-04
>   3.19707866799983E-05 -2.70889502728457E-04  5.33791678975208E-05
>  -2.78349089043350E-06 -2.85186811447324E-04  1.61981842983486E-04
>  -1.04618748793896E-04 -2.28202988141319E-04  7.90379285890031E-06
>  -2.02693812792706E-04  8.69208240839734E-05 -1.84601881427213E-04
>  -8.50218656013146E-05 -8.39331545859563E-05  6.42536289917873E-04
>   2.92505776294561E-04  3.63154915768707E-04  6.03155586847407E-05
>  -3.40104745704510E-05 -2.30454886952254E-04 -2.04088755425137E-04
>  -3.77072359684074E-04  3.97765835567884E-04  1.12993505119051E-04
>   7.55670407586609E-05  9.70859721352099E-05  3.14167124563309E-04
>  -1.44434284361358E-04 -1.32708233980981E-04  5.47298003152002E-05
>  -6.08024450694477E-05 -4.97003786961105E-05 -3.11797153424798E-05
>  -8.56178777930613E-05  1.06254292787205E-04 -1.33999242621271E-04
>  -2.80479286292257E-05  9.74725746784787E-05 -4.75313418554370E-04
>   2.44355000913113E-04 -8.43838982661567E-05 -1.07491644359097E-04
>  -4.81862237841437E-05  2.10948350585710E-04 -2.35769013621597E-05
>   1.40524746720755E-04 -3.22263735677215E-04  2.88687187596357E-04
>  -4.11478798117340E-04 -2.69315384499740E-04 -2.36382259346532E-04
>  -1.07055033660231E-04 -9.75083786074954E-05  5.21628093244381E-05
>   3.31081873016545E-04  2.04484191581289E-04 -2.88181851894121E-04
>   3.42738488329567E-04 -7.15579211811260E-05  2.89191747320400E-04
>  -6.28194429251086E-05 -1.73710346766907E-04 -8.12427997233024E-04
>   8.16713430019538E-05 -2.54381809707164E-04  2.91038657272049E-04
>  -1.55432610422003E-04  4.95419304109589E-06 -2.94084032674493E-05
>   2.28510959359023E-04 -2.15854603544687E-04  2.24991198746045E-04
>  -4.41074379551528E-04  1.67034560302883E-04 -1.53824319378859E-04
>   3.12748574781470E-04  2.67345432333349E-04 -1.77133432911651E-04
>  -3.24409756064884E-04 -4.25684700707660E-04 -1.97272277617721E-04
>   3.78425379320122E-04  1.07138462246563E-05  2.29306378688200E-06
>
> #Definition of the planewave basis set
> ecut 35.0       # Maximal plane-wave kinetic energy cut-off, in  
> Hartree
>
> #Definition of the k-point grid
>
> nkpt 1          # Only one k point is needed for isolated system,
>                 # taken by default to be 0.0 0.0 0.0
>
> irdwfk 1        # Restart simulation with the wavefunction from a  
> previous run
>
> #Definition of the SCF procedure
> iscf 5            #Control of the Self-Consistent Field method to  
> be used
>                   #The default is 5
> nstep 50          # Maximal number of SCF cycles
> toldfe 1.0d-6     # Will stop when, twice in a row, the difference
>                   # between two consecutive evaluations of total  
> energy
>                   # differ by less than toldfe (in Hartree)
> diemac 1.0        # Although this is not mandatory, it is worth to
>                   # precondition the SCF cycle. The model dielectric
> diemix 0.5        # function used as the standard preconditioner
>                   # is described in the "dielng" input variable  
> section.
>                   # Here, we follow the prescriptions for molecules
>                   # in a big box
>
> # Molecular Dynamics Variables Setup
> #
> amu(2) 22.98977 35.453
> ionmov 6                        # N,V,E MD using Verlet algorithm
>
> dtion 206.71                     # time-step in atu 5fs
>
> ntime 100
>
>
> On Friday 23 September 2005 11:11, Anglade Pierre-Matthieu wrote:
>
> > Hi,
> >
> > This won't answer your question but please consider the following:
> > Molecular dynamics is chaotic and you'll never ever get exact  
> > results.
> > Whatever precision you have for your computation you will always get
> > results very differents (after even a not so big number of steps)  
> > from what
> > a non arbitrary precision computation would gives.
> > Then I would say that you can behaves in two different ways:  
> > either you
> > choose to consider that molecular dynamics is not a suitable  
> > technic for
> > your problem; or you decide not to mind that much about the  
> > differences
> > resulting from a small error on your atoms phase space coordinates  
> > at the
> > first restart step.
> >
> > regards
> >
> > PMA
> >
> > On 9/23/05, ngalamba@cii.fc.ul.pt <ngalamba@cii.fc.ul.pt> wrote:
> >
> > > Dear ABINIT users
> > >
> > > I am trying to use ABINIT to carry molecular dynamics simulations  
> > > (N,V,E)
> > > (IONMVOV 6) of simple ionic melts (e.g. NaCl).
> > > Because it is hard to get a complete simulation on a single run  
> > > without
> > > computers crashing for some reason I have been trying to divide my
> > > production runs in several stages that I therefore need to link.
> > >
> > > To link the partial simulations (say a 100 time-steps each with a
> > > time-step of 5 fento seconds) I use the variable IRDWFK 1 to use the
> > > wavefunction from the previous partial run and update the  
> > > positions and
> > > velocities in the new input file. ABINIT tells me that the file  
> > > is good
> > > for restarting the simulation and takes it from there.
> > >
> > > The "problem" is that the first time (MD step 1 not 0) ABINIT  
> > > updates the
> > > positions and velocities using the Verlet algorithm it only gives  
> > > me the
> > > first 6 digits of positions and velocities equal to those I would  
> > > obtain
> > > if I had continued the simulation on a single run. Then as more  
> > > and more
> > > integrations are performed the positions and velocities start to  
> > > get more
> > > and more different from those I would obtain in a single run.
> > >
> > > Is there any way of restarting a MD calculation in a such a way  
> > > that the
> > > results are identical whether we do it in several stages using  
> > > the _WFK
> > > files as input or in just a single run?
> > >
> > > I would be highly appreciated if anyone could give some hint on  
> > > this type
> > > of calculation as I am new to ABINIT and also to Hellman-Feynamnn  
> > > MD.
> > >
> > > Thanks in advance.
> > >
> > > Nuno Galamba
> >
> > --
> > Pierre-Matthieu Anglade
>
> --
> Nuno Galamba
> Grupo de Física-Matemática
> Complexo Interdisciplinar
> Av. Prof Gama Pinto 2,
> 1649-003 LISBOA
> PORTUGAL
>
> Voice: 21 790 48 59 Ext: 4259
> email: ngalamba@cii.fc.ul.pt