Dear Hu Qiu, did you find how to make transiesta to converge? I'm having the same problem for certain bias values and I also tryed several modifications as Kemal did without any improvement. I think that not only me would appreciate someone to share information about what to touch in order to tune the convergency in these kind of systems. I will let you know if I find something.
Juan Manuel On Mon, Jul 11, 2011 at 3:19 AM, Hu Qiu <[email protected]> wrote: > Thanks for response. > > How did you fix this problem? > > Sometimes, I rerun the transiesta calculation without changing any > parameters in the input fdf file that previously stopped, and found the > calculation can complete normally. Certainly, it still cannot converge in > 500 steps. I don’t know what happened. > > > > > > Hu Qiu. > > ________________________________ > > 发件人: [email protected] [mailto:[email protected]] 代表 zahra > akbarinejad > 发送时间: 2011年7月10日 19:59 > 收件人: [email protected] > 主题: Re: [SIESTA-L] About the convergence of transiesta > > > > Hi > > perhapse it depends to your computer. > > namely  in the  middle of run it can not continue. then the program stops > without any error. > > This happened for me. > > > > > > > > > > > > > > > > > > > > On Sun, Jul 10, 2011 at 1:24 PM, Hu Qiu <[email protected]> wrote: > > Dear all, > >  > > We constructed a graphene nanogap and tried to calculate its transport > properties. Pls see the attached file for our model (model.JPG), the red > atoms in graphene are the electrodes and the cyan atoms are the scattering > region. However, after successfully optimizing the system, we found it very > difficult to converge in transiesta calculation and sometimes the run > suddenly stopped without any error message. Below is the lowermost part of > the output file. > >  > > TranSiesta: Qsol,Qtot: 1003.732 1008.000   1.000 > > transiesta: 407  -38476.1444  -38449.0654  -38449.0654 1.1437 > -5.5380 > > TranSiesta: Qsol,Qtot: 982.619 1008.000   1.000 > > transiesta: 408  -38632.6402  -38447.6087  -38447.6087 3.4957 > -5.5380 > > TranSiesta: Qsol,Qtot: 962.912 1008.000   1.000 > > transiesta: 409  -38721.4507  -38445.9697  -38445.9697 8.0143 > -5.5380 > > TranSiesta: Qsol,Qtot: 1017.218 1008.000   1.000 > > transiesta: 410  -38473.8597  -38452.6161  -38452.6161 1.3445 > -5.5380 > > TranSiesta: Qsol,Qtot: 1014.635 1008.000   1.000 > > transiesta: 411  -38445.4879  -38452.8883  -38452.8883 1.4824 > -5.5380 > > TranSiesta: Qsol,Qtot: 1032.013 1008.000   1.000 > > transiesta: 412  -38522.8264  -38452.0826  -38452.0826 3.1563 > -5.5380 > >  > > We listed below the major parameters of transiesta calculation. > > Attachments are all the input fdf files we used in transiesta calculation > (including coordinate of structure), whose parameters settings are taken > from http://www.mail-archive.com/[email protected]/msg01041.html. > >  > > Could anyone help us check the input files to see what mistakes we made? Was > the structure of system optimized improperly? > >  > >  Any help will be greatly appreciated. > >  > > Best regards > >  > > Hu Qiu > >  > >  > > #Major parameter settings in fdf file > > MaxSCFIterations       1000                   > > #DM.Tolerance          1.d-5                  > > DM.NumberPulay         > 6                      > > DM.NumberBroyden       0                      > > DM.MixingWeight        0.1000000000           > > DM.OccupancyTolerance  0.1000000000E-11       > > DM.NumberKick          > 0                      > > DM.KickMixingWeight    0.5000000000           > > DM.Tolerance           0.1000000000E-03 > >  > > # Transiesta information > > SolutionMethod Transiesta > > # GENGF OPTIONS > > TS.ComplexContour.Emin       -28 eV > > TS.ComplexContour.NPoles      16 > > TS.ComplexContour.NCircle     16 > > TS.ComplexContour.NLine       10 > > # BIAS OPTIONS > > TS.biasContour.NumPoints      10 > > TS.UpdateDMCROnly             F > >  > > # TS OPTIONS > > TS.Voltage 0.0 eV > >  > > # TBT OPTIONS > > TS.TBT.Emin -3 eV > > TS.TBT.Emax +3 eV > > TS.TBT.NPoints 500 > > TS.TBT.NEigen 3 > > TS.TBT.Eta       0.000001 Ry > >  > > # Write electrode hamiltonian > > TS.SaveHS  .true. > > TS.SaveLead .true. > >  > >  > >  > > # LEFT ELECTRODE > > TS.HSFileLeft ../4-Electrode-Nanogap/elec.TSHS > > TS.NumUsedAtomsLeft  64 > > TS.BufferAtomsLeft   48 > >  > > # RIGHT ELECTRODE > > TS.HSFileRight ../4-Electrode-Nanogap/elec.TSHS > > TS.NumUsedAtomsRight 64 > > TS.BufferAtomsRight  48 > >  > >
