[Wien] SCAN meta-GGA+ DFT-D

[Gavin Abo]

In the dftd3.f of dftd3.tgz for V3.2 Rev 0 at


https://www.chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/get-the-current-version-of-dft-d3


the lines 1531 - 1538 are:


! read parameter file from home directory
       call system('hostname > .tmpx')
       open(unit=43,file='.tmpx')
       read(43,'(a)')ftmp
       close(43,status='delete')
       call get_environment_variable("HOME", homedir)
       write (*,*) trim(homedir)
       write(dtmp,'(a,''/.dftd3par.'',a)')trim(homedir),trim(ftmp)

In line 1538, it looks like it reads .dftd3par.localhost and not 
8.dftd3par.localhost.

The man.pdf in at the "Short documentation of the program" link on the 
same page has:

s6 a1 s8 a2 alpha6 version
for BJ-damping (version = 4).

Should your .dftd3par.localhost be something like:

1 0.538 5.42 4.4752 0 4

where

s6 = 1
a1 = 0.538
s8 = 5.42
a2 = 4.4752 (? [ 
https://www.chemie.uni-bonn.de/pctc/mulliken-center/software/dft-d3/functionalsbj 
])
alpha = 0 (not used)
version = 4

Perhaps, check the case.scfdftd3 file.  The lines 672 - 678 of dftd3.f are:

       if((version.eq.4).or.(version.eq.6))then
          write(*,'('' s6       :'',f10.4)') s6
          write(*,'('' s8       :'',f10.4)') s18
          write(*,'('' a1       :'',f10.4)') rs6
          write(*,'('' a2       :'',f10.4)') rs18
          write(*,'('' k1-k3    :'',3f10.4)') k1,k2,k3
       endif

It looks like those lines might be outputting the values to the standard 
output (or the case.scfdftd3 file with x dftd3) that were read in 
.dftd3par.localhost.  If so, that could be used to verify that the 
values inputted were read correctly by the dftd3 program.

On 12/26/2017 10:36 PM, Osama Yassin wrote:
>
> Dear Prof Tran
>
>
> I created the file .dftd3par.localhost (localhost is my hostname) and 
> placed it in my home directory. However, dftd3 stops with error after 
> lapw0. You mentioned in your email that the name name has to be 
> "8.dftd3par.hostname". Is the number "8" is a typo or it has to be 
> there?!.
>
>
> I understand that the file is a free format and the parameters are to 
> be in the following sequence (for BJ)
>
>
> 1 0.538 0 5.42 4
>
> where:
>
> s6=1
>
> a1= 0.538
>
> s8=05.42
>
> a2=version
>
> alpha is not used
>
> What is wrong with me?
>
> Osama
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20171227/49b96263/attachment.html>