[gmx-users] Orthorhombic structure and Lincs warnings

[Alex]

Your answers to questions 3 and 4 contradict each other in several ways. 
Do not use restraints or constraints (hint: LINCS is a constraint 
algorithm) and make sure your forcefield works. Right now it doesn't 
work, because a somewhat working forcefield does not lead to any 
crumpling, i.e. fix it and use appropriately selected box dimensions, 
especially if you are using periodic boundaries. This is a searchable 
list and people asked the very same questions before.

If your box is rectangular, there is no need to call it orthorhombic. In 
your QM calculations, if you have a rectangular sheet of graphene with 
boundaries parallel & perpendicular to the vector connecting the atoms 
in the unit cell, the in-plane vectors of the cell must be 90 degrees, 
so there goes the accuracy of your quantum espresso setup.

And please, please respond in the appropriate thread. There is 
absolutely no need to flood the entire mailing list, now including the 
digest.

Alex

On 9/6/2018 12:48 AM, sagar bathla wrote:
> Hi Alex ,
>
> sorry to post my problem too many times.
>
> Coming back to your questions related to my problem
> 1. angles are 90,90,120 in x,y and z respectively. actually in quantum
> espresso this structure is little bit easy to handle.
> 2. without LINCS i haven't tried the simulation, i will do it without LINCS
> and get back to you.
> 3. I am basically interested in dynamics of water through porous graphene.
> if i am not using constraint and all it is getting crumbled and shape of
> graphene changes.
> 4. Yes, it is stable.
>
> Regards