QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: arielzn on November 25, 2014, 12:26
-
Hi,
I'm trying the VNL package, it's a really nice tool. I prepared the POSCARs for different surfaces of a given system, in these calculations is usual to relax some topmost layers but to fix the bottom ones.
By using the VASP scripter plugin I didn't find an option to control which atoms will be optimized and which don't. Is there such an option? Or should I edit the T flags by hand after generating the POSCAR?
thanks,
-
We are glad you like it!
You can do what you need by setting the tag name 'Constraints' to the atoms you want to constrain.
From the Builder,
- select the atoms
- Got to Select->Tags tools and specify the tag name above.
The F flags will be added to the exported POSCAR file.
At the moment you cannot really specify the three T/F flags for the three direction independently.
You will have to do edit the generated POSCAR file.
However, the VASP plugin is under constant development adding new features. We may add this one soon.
You can check from the AddOn Manager if you have the latest version installed or subscribe to our news RSS feed (http://feeds2.feedburner.com/QuantumWiseNews) to be always updated.
Currently it is the 1.3 version for VNL 2014.
-
Thanks for the fast answer, that was non trivial to find out but it worked.
I actually need the three coordinates to be fixed or not, so that approach did the trick for me.
Another related question, is it possible somehow to decide the order of the species when exporting the POSCAR? It seems to be ordering them alphabetically.
I'm just using VNL to build different POSCARs for complex surfaces. The INCAR and POTCARs I prefer to keep it editing by hand, since I did that for years, and I like to control the POSCAR species order to match my POTCARs of previous calculations.
Meanwhile I'm just moving the species coordinates blocks by hand after exporting.
-
That's correct, the ordering is alphabetic.
I personally think the "proper" way to solve this would be if the VASP format would just take the chemical species on each line, with the coordinates... It seems silly to rely on an implicit ordering of the elements, even if the VASP 5 format is an improvement over VASP 4 where you really had to be careful.
Do you need complete control of the order or just a few options, like "Order by Z" or "Alphabetic by element"?