The ABINIT Users Mailing List ( CLOSED )

[Steven Homolya <steven.homolya@spme.monash.edu.au>]

Nuno A. G. Bandeira wrote:
> Hello all,
>
> I've been trying to optimise the unit cell of a polymer and I've set up 
> a supercell model to do it.
>
> The problem is abinis creates a lot of FFT grids because the supercell 
> is large. I tried to change them with the ngfft key to speed up the 
> calculations but the program won't parse this key and will keep running 
> a 216 point fft grid no matter what.

I don't know why your way of doing it did not work, but you can just set 
boxcutmin to something less than the default 2.0 if you want an undersampled 
Fourier mesh. Abinit should take care of the (coarser) grid selection, so no 
need to specify ngfft.

Steve