[CPMD-list] How to set the NOSE ELECTRONS parameters.

[HW Sheng]

Dear all,  attached please find the input file.

  For an isobaric-isothermal MD simulation, I need to set the target electron kinetic energy. According to the manual, it's suggested to perform a short run without thermostats, and then take the "natural" electron kinetic energy as the target value (see, e.g, the CPMD manual, version 3.5.2 pp 62).  Still, I found this "target" value very capricious. After all, I arbitrarily set the kinetic energy to be 0.0001 a.u . During simulation I found that the output EKINC (the fictitious electron kinetic energy, ~ 0.007 a.u. after 1000 time steps) differs significantly from this desired kinetic energy 0.0001 a.u. I think this is abnormal, because EKINC should fluctuate around the desired (also predefined) kinetic energy. Can anyone tell me why?

  The ion temperature is also a pain in the neck. I have tried a number of NOSE parameters, but the temperature fluctuation is still too much. For example, the target ion temperature is 5000 Kelvin, and the fluctuation reaches as much as 1000 K. A possible reason is that the target electron kinetic energy is set too small and causes the damping of the ions. Do I make any sense?

  Besides, any rule-of-thumb to set the CELL kinetic energy? 
  In a nutshell, I wish to have a guidance in choosing the Nose-Hover chain parameters. I appreciate your help.

&CPMD
  MOLECULAR DYNAMICS
  PARRINELLO-RAHMAN NPT
  RESTART LATEST ALL 
  NOSE IONS
   5000 2000
  NOSE ELECTRONS
   0.0003 6000
  NOSE CELL
   0.5 1000
  NOSE PARAMETERS
   8 8 4 6 25 2
  CMASS
   4000
  MAXSTEP
   500
  TIMESTEP
   3.0
  SPLINE RANGE
   4.0
  SPLINE POINTS
   5000
  MOVIE SAMPLE
   10
&END

&SYSTEM
  ANGSTROM
  SYMMETRY
  1
  ISOTROPIC CELL
  CELL
  9.9 1.0  1.0  0  0  0
  CUTOFF
  70.0
  TESR
   4
&END

&ATOMS
ISOTOPE
92.91

*Nb-q5 GOEDECKER  
   LMAX=D 
    54
       -3.3000 -3.3000 -3.3000
       -3.3000 -3.3000  0.0000
       -3.3000 -3.3000  3.3000
       -3.3000 -0.0000 -3.3000
       -3.3000  0.0000  0.0000
       -3.3000  0.0000  3.3000
       -3.3000  3.3000 -3.3000
       -3.3000  3.3000  0.0000
       -3.3000  3.3000  3.3000
       -0.0000 -3.3000 -3.3000
       -0.0000 -3.3000  0.0000
        0.0000 -3.3000  3.3000
       -0.0000 -0.0000 -3.3000
        0.0000  0.0000  0.0000
        0.0000  0.0000  3.3000
        0.0000  3.3000 -3.3000
        0.0000  3.3000  0.0000
        0.0000  3.3000  3.3000
        3.3000 -3.3000 -3.3000
        3.3000 -3.3000  0.0000
        3.3000 -3.3000  3.3000
        3.3000 -0.0000 -3.3000
        3.3000  0.0000  0.0000
        3.3000  0.0000  3.3000
        3.3000  3.3000 -3.3000
        3.3000  3.3000  0.0000
        3.3000  3.3000  3.3000
       -4.9500 -4.9500 -4.9500
       -4.9500 -4.9500 -1.6500
       -4.9500 -4.9500  1.6500
       -4.9500 -1.6500 -4.9500
       -4.9500 -1.6500 -1.6500
       -4.9500 -1.6500  1.6500
       -4.9500  1.6500 -4.9500
       -4.9500  1.6500 -1.6500
       -4.9500  1.6500  1.6500
       -1.6500 -4.9500 -4.9500
       -1.6500 -4.9500 -1.6500
       -1.6500 -4.9500  1.6500
       -1.6500 -1.6500 -4.9500
       -1.6500 -1.6500 -1.6500
       -1.6500 -1.6500  1.6500
       -1.6500  1.6500 -4.9500
       -1.6500  1.6500 -1.6500
       -1.6500  1.6500  1.6500
        1.6500 -4.9500 -4.9500
        1.6500 -4.9500 -1.6500
        1.6500 -4.9500  1.6500
        1.6500 -1.6500 -4.9500
        1.6500 -1.6500 -1.6500
        1.6500 -1.6500  1.6500
        1.6500  1.6500 -4.9500
        1.6500  1.6500 -1.6500
        1.6500  1.6500  1.6500

&END

&DFT
  NEWCODE
  FUNCTIONAL PADE
&END


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cpmd.org/pipermail/cpmd-list/attachments/20021115/c566d6ca/attachment.html