[DFTB-Plus-User] iterated process of MD jobs

전영인 yijhon at uos.ac.kr
Tue Nov 19 13:22:10 CET 2019

Dear Bálint,

Thank you for your answers, Bálint.

I have two questions as follows.

1. The xyz format of DFTB+ has seven columns, that is, seven attributes per atom.

I suppose that the first three are for postion ans the last three are for velocity. Then, what does fourth column represent? 

2. For NPT simulation, three lattice constants should be stored at every MDRestartFrequency. Then, are these values additionally stored in the xyz file, compared to that of NVT ensemble?

3. I'd like to know xx,yy,zz component (or -Pxx, -Pyy, -Pzz) of stress tensor. How can I know these values using DFTB+?

Best regards,


보낸사람: Bálint Aradi
받는사람: dftb-plus-user at mailman.zfn.uni-bremen.de
보낸날짜: 2019-11-18 20:06:15 GMT +0900 (Asia/Seoul)
제목: Re: [DFTB-Plus-User] iterated process of MD jobs

Dear Young, 

 > 1. Can I generate a file of last atomic velocities (similarly to 
 > geo_end.gen for the position) after completion of a MD work, so that I 
 > can insert it in the input of new work? 

 Yes, currently the velocities are stored in the xyz geometry file in 
 columns 6 - 8. Those you can use to restart your calculation. They are, 
 however, in AA/ps, so make sure to specify the units of the velocities 
 in the DFTB+ input accordingly! 

 > 2. For parallel running (4 prcoesses) of "mpirun -np 4 -machinefile 
 > $PBS_NODEFILE dftb+", do I have to explicitly describe group = 4 (not 1) 
 > in Parallel section? 

 Groups only make sense, if you wan't to parallelise over k-points. (One 
 group deals with one k-point at a time). If you do not specify groups, 
 all processes will work on the same Hamiltonian. 

 Rule of thumb: If you have a small Hamiltonian but many k-points, it may 
 make sense. If you don't have k-points, it does not. 

 > What output file will be created for periodic saving (for 
 > MDRestartFrequecy) of cooridnates, velocties, charges of atoms? 

 The xyz geometry file. 

 Best regards, 


 Dr. Bálint Aradi 
 Bremen Center for Computational Materials Science, University of Bremen 

 _______________________________________________ DFTB-Plus-User mailing list DFTB-Plus-User at mailman.zfn.uni-bremen.de https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20191119/e4799e9e/attachment.html>

More information about the DFTB-Plus-User mailing list