The ABINIT Users Mailing List ( CLOSED )

[paul.boulanger@umontreal.ca]

Hi,

I never did amorphous systems with Abinit, but I think their are a few stange 
things in your input file. First, you seem to increase your acell by 5 bohrs 
in every direction at each dataset. Couple this with the fact that you use 
reduced coordinates for your atoms and you get that you increase the bond 
length between your atoms at every dataset. Are you sure this is what you 
want?? By the fourth dataset, you have tripled the bond length of your atoms: 
the Si-O bond length is 9.456 bohrs is your last dataset.

Secondly, you are using getwk -1 which states that you use the wavefunctions 
of the previous datasets as input. I am not sure this is desirable, because 
the atomic positions vary greatly. Hence, your input wavefunctions are not a 
good approximation for the new wavefunctions.

Thirdly, you define ecut? 14 without any increments. I think the code will 
only ignore the extra character (?) and define ecut 14 for every dataset. But 
is this want you wanted? Furthermore, your ecut seems a little low for 
oxygen, maybe increasing it will yield a better relaxation.

Well, that all I can think of. I nver simulated amorphous systems and hence I 
don't really know the tricks used to generate such systems in a periodic 
code. But it seems to me that your are currently doing a crystalline 
calculation with  different cell parameters.

Hope this helps a little,

Paul