shape of the supercell

General discussion of the Cambridge quantum Monte Carlo code CASINO; how to install and setup; how to use it; what it does; applications.
Post Reply
Mauro_Causa
Posts: 8
Joined: Sun Jun 23, 2013 1:41 pm

shape of the supercell

Post by Mauro_Causa »

presently Casino allows supercells of the type
n 0 0
0 n 0
0 0 n

I would like to run not diagonal supercells
of the type for example
0 1 1
1 0 1
1 1 0
that could be useful when the primitive cell is very anisotropic.

If reasonable it is easy to modify the code?
Or may be there is already an option in CASINO that i didn't find?

thank you
Mike Towler
Posts: 239
Joined: Thu May 30, 2013 11:03 pm
Location: Florence
Contact:

Re: shape of the supercell

Post by Mike Towler »

Hi Mauro,

The input keyword block scell_matrix does exactly that. This is a generalization of npcell that allows for arbitrary combinations of the primitive-cell lattice vectors.

Cheers,
Mike
Mauro_Causa
Posts: 8
Joined: Sun Jun 23, 2013 1:41 pm

Re: shape of the supercell

Post by Mauro_Causa »

thank you Mike

how can i communicate to crysgen09 the supercell matrix?
have i an analogous of scell_matrix option also for crysgen09?

thanks again
mauro
Mike Towler
Posts: 239
Joined: Thu May 30, 2013 11:03 pm
Location: Florence
Contact:

Re: shape of the supercell

Post by Mike Towler »

Hi Mauro,

So the purpose of CRYSTAL here is to define the 1x1x1 'primitive' cell for CASINO, and as far as CASINO is concerned that is defined just by the lattice vectors that CRYSTAL sends to it via the gwfn.data file. You can construct these lattice vectors as an arbitrary linear combination of the actual primitive lattice vectors using the CRYSTAL 'SUPERCEL' keyword. The matrix defined by SUPERCEL does not need to be sent to CASINO, just the resulting modified lattice vectors.

The purpose of the CASINO scell_matrix block is then to define a QMC simulation cell as arbitrary linear combinations of the (possibly non-primitive) lattice vectors in the gwfn.data file (if they're just nxnxn then you can use the simpler npcell block instead, which just specifies the diagonal elements of scell_matrix). This is clearly a separate thing from the CRYSTAL SUPERCEL stuff.

So the answer to your question is: 'It isn't necessary to communicate to crysgen09 the supercell matrix.'

Mike
Mauro_Causa
Posts: 8
Joined: Sun Jun 23, 2013 1:41 pm

Re: shape of the supercell

Post by Mauro_Causa »

thank you Mike.

I'll try to do in
Post Reply