[Pw_forum] Error in routine sym_rho_init_shell "lone vector" with fixed FFT dimension

[Paolo Giannozzi]

Very strange. The error signals a failure in the algorithm
that finds shells of G-vectors that transform into each other
under symmetry operations. It might however be due to too small
FFT dimensions, maybe. It is hard to figure out what happens 
without an input file.

P.

On Wed, 2014-02-05 at 15:49 +0100, Thomas Gruber wrote:
> Dear all,
> 
> I try to determine the heat capacity at constant pressure and calculate 
> the phonon DOS at different volumes. Since I get jumps in free energy 
> when the FFT dimension change, I fix it to the one from the starting 
> volume. But for on of my systems everything works fine up to 4% 
> isotropic increase of the cell axis and at 5%  I get the following error 
> in the starting scf calculation:
>   %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>       Error in routine sym_rho_init_shell (2):
>       lone vector
>   %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> This error does not show up, when I do not fix the FFT dimension. 
> Without fixing the FFT dimension the FFT dimension is different to the 
> one I choose.
> Is there a way to keep my fixed FFT dimension and get rid of this error?
> 
> Thanks in advise.
> 

-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment, 
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222