[Wien] Relaxation of already known structure

[Parker, David S.]

Jihoon, it depends on the sensitivity of the calculations. I generally always use the experimental lattice constants, and then relax the internal coordinates if forces are significant - greater than 5 or so mRyd/bohr (the units WIEN uses). But this takes a lot longer than a simple scf run and often has little, if any effect on the results.  For a first pass at a calculation, unless the forces are huge, I don't do a relaxation.  If the result is interesting enough that a publication may result, then I do a relaxation. Best, David Parker

From: Jihoon Park <maximumenergyproduct at gmail.com<mailto:maximumenergyproduct at gmail.com>>
Reply-To: A users <wien at zeus.theochem.tuwien.ac.at<mailto:wien at zeus.theochem.tuwien.ac.at>>
Date: Monday, July 21, 2014 3:16 PM
To: A users <wien at zeus.theochem.tuwien.ac.at<mailto:wien at zeus.theochem.tuwien.ac.at>>
Subject: [Wien] Relaxation of already known structure

Dear Users,


I am wondering if we must do the relaxation for all calculations.
I have found some first principles studies with experimental lattice constants, including Dr. Novak's work "PRB 71, 1844422 (2005)."
Therefore, I need to know if the first principles calculations with experimental lattice constants are reliable or in what case, it is good enough or somethings.
Could anybody please give me some guidance?


All my best
Jihoon Park