[Smeagol-discuss] Why no transmission data generated?

[徐卓]

Hi, I am a  beginner of Smeagol. I could run the example of Au atom wires with no problem at all.
After calculating the Au leads, I try to calculate zero bias conductance and I-V characteristics of a one dimensional nanostructure of two periods, which contains about 100-150 atoms, between Au leads.
The iteration of 0 V (the initial bias) of Smeagol converges to a positive energy value, while the scf calculation of the same nanostructure by Siesta will converge to a negative energy value, is there anything wrong of Smeagol or not?
Moreover, after the convergence of the initial bias, the calculation stops automatically. There are files of  .alloc, .CHR, .DM, .fdf (input), .KP, and .out (output) only, no transmission data (.CUR, .TRC, .LGF etc.) at all. What's wrong with it?
The parameters of my input file are as follows:
%block PAO.BasisSizes
(...)  SZP
%endblock PAO.BasisSizes
%block kgrid_Monkhorst_Pack
  2   0   0    0.0
  0   2   0    0.0
  0   0  15    0.0
%endblock kgrid_Monkhorst_Pack
xc.functional   LDA  
xc.authors     CA  
SpinPolarized   F    
FixSpin                         F
TotalSpin                       0.0    
NonCollinearSpin  F  
MeshCutoff     180. Ry  
MaxSCFIterations  500  
DM.MixingWeight   0.6
DM.NumberPulay   10  
DM.PulayOnFile   F  
DM.Tolerance    5.0E-4
NeglNonOverlapInt  T  
SolutionMethod   diagon
ElectronicTemperature  300 K  
DM.UseSaveDM    T 
SaveElectrostaticPotential T
WriteCoorXmol   T 
SaveHS    F
SaveRho   T  
SaveDeltaRho   F
WriteDenchar   F  
WriteEigenvalues  F
WriteDMT                   T
EMTransport              T 
NEnergReal               0 
NEnergImCircle           15 
NEnergImLine             20
NPoles                   4
VInitial                 0.d0 eV
VFinal                   0.d0 eV
NIVPoints                0 
Delta                    1.d-4
EnergLowestBound        -5.0d0 Ry
SpinConfLeads            0
NSlices                  1 
TrCoefficients   T
AtomLeftVCte             (...)
AtomRightVCte            (...)
NTransmPoints            100
InitTransmRange          -10.0d0 eV 
FinalTransmRange         0.d0 eV 
%block SaveBiasSteps
 0
%endblock SaveBiasSteps
SaveMemTranspK   F
PeriodicTransp   T
TransmissionOverk         F
#ParallelOverK   T 
#Diag.ParallelOverK         T
HartreeLeadsBottom      (...) eV
HartreeLeadsLeft          0. Ang
HartreeLeadsRight       (...) Ang
The output file ends as follows:
siesta:                 ===============================
SMEAGOL                 Bias step =    0, V =  0.000 Ry
                            Begin CG move =      0
                        ===============================
iodm: Reading Density Matrix from files
InitMesh: MESH =   192 x   192 x   144 =     5308416
InitMesh: Mesh cutoff (required, used) =   180.000   191.452 Ry
 
* Maximum dynamic memory allocated =   416 MB
WARNING: There are k-points along z
......
siesta: iscf   Eharris(eV)      E_KS(eV)   FreeEng(eV)   dDmax  Ef(eV)
siesta:    1   (...)
timer: Routine,Calls,Time,% = IterSCF        1   14868.199  99.99
......
siesta:    (XX)   (...) [reaches the convergence criteria]
[the output file just stops here]
If I set kgrid_Monkhorst_Pack to 1 1 1, it says Segmentation Fault and stops at once. I have ever read the Dr. Rocha's research of a DNA molecule between Au leads, which even large than my one dimensional nanostructure. What's wrong of my case? Could anybody suggest any hint for me please?
Thank you very much.
XU Zhuo
PhD. student from Graduate School of Chinese Academy of Science
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.tchpc.tcd.ie/pipermail/smeagol-discuss/attachments/20100224/b60ac99c/attachment.html