Hi, I am doing the Full DFPT method to obtain NLO properties. In dataset 4 and 7 it generated DDB files. I created a new folder and doing anaddb. With the mrgddb code I did merge tnlo_6_DS4o_DDB and tnlo_6_DS7o_DDB files. These files do not contain non-linear susceptibilities. But this is printed out in the output of dataset 7 in .abo file not in the DDB file. Therefore, when I apply anaddb it shows an error saying this:
Need much more detail, in/output files etc. At this stage we have no idea what you did in the different dtsets or what 4 and 7 do or should correspond to.
Thanks for the kind response. I am doing the non-linear optical property calculation using full DFPT method (following tutorial tnlo_6.abi). In dataset 4 it calculates the response functions (2nd derivatives of E) and in dataset 7 it calculates 3rd derivatives of E. These calculations were performed properly following the tutorial and worked well. Here is the postprocessing part attached where I did mrgddb to merge the DDB files of 2nd derivatives and 3rd derivatives - raman.tgz (437.7 KB)
Quick look at the ddb file: you have some of the 3rd derivatives, but only for 2 E fields and atoms 1-4 (out of 56). Perhaps rfatpol is limited in that dtset? Send your DFPT inputs as well
the variable takes 2 numbers, the 1st and last atoms. Not 56 values, so it is ignoring anything beyond “2”. The DDB file goes up to ipert 4 possibly due to symmetries.
Use
d3e_pert1_atpol7 1 56
As I mentioned, in the latest versions of abinit this now defaults to 1 natom, which is what you would have wanted anyway.
NB: ecut 60 for PAW potentials is insanely high… but I presume you converged things properly. Maybe try another set of potentials (GBRV is softer)
I was doing the same calculation with d3e_pert1_atpol7 1 56 , it starts to perform 105 no. of perturbations and after 82 of it failed after 3days of calculation with this error message. Could you please put some hints what could be the reason of this type of failure?
