Hello
I am doing tutorials trf1_3.in and I am stuck in the place where they calculate the dynamical matrices for phonons. Can someone help me out with that?
Thank you in advance.
Phonon calculations
Moderators: maryam.azizi, bruneval
Re: Phonon calculations
Dear saikrishna,
What do you mean by "stuck", I mean what is blocking you exactly?
Cheers,
Eric
What do you mean by "stuck", I mean what is blocking you exactly?
Cheers,
Eric
-
saikrishna
- Posts: 15
- Joined: Fri Jun 21, 2019 12:11 pm
Re: Phonon calculations
Hey Eric
I found the answer for my previous question but I have a doubt on what is a q point in brillouin zone?
Thanks in advance
Sai krishna.
I found the answer for my previous question but I have a doubt on what is a q point in brillouin zone?
Thanks in advance
Sai krishna.
-
saikrishna
- Posts: 15
- Joined: Fri Jun 21, 2019 12:11 pm
Re: Phonon calculations
Hey
I have a doubt here in executing the file which I wrote. Please tell me my mistake and the correct me asap.
PFA.
Thanks in advance
Sai krishna
I have a doubt here in executing the file which I wrote. Please tell me my mistake and the correct me asap.
PFA.
Thanks in advance
Sai krishna
- Attachments
-
phonon3.in- (532 Bytes) Downloaded 454 times
-
saikrishna
- Posts: 15
- Joined: Fri Jun 21, 2019 12:11 pm
Re: Phonon calculations
Hey everybody,
I am running phonon calculations for a material for which I obtained negative frequencies. May I know what does it mean or is it something wrong in the code which is responsible?
Thanks in advance
Sai krishna N.
I am running phonon calculations for a material for which I obtained negative frequencies. May I know what does it mean or is it something wrong in the code which is responsible?
Thanks in advance
Sai krishna N.
Re: Phonon calculations
Dear Sai krishna,
Negative frequencies corresponds actually to imaginary frequencies.
1) These values can appears if your system is not stable w.r.t. atomic displacements (unstable phonon), meaning your structure is not the ground state one. You can find the ground state structure by displacing the atoms following the eigenvector given by the unstable mode(s) (also called mode condensation). Then you know your structure is stable once no "negative" frequency is present in your calculation. In this case these negative frequencies are physically relevant.
2) Your system is supposed to be stable locally or supposed to be the ground state and you still find negative values for phonon frequencies. Here two source cases can be discussed:
a) Your calculation gives the so called "acoustic sum rule" (ASR) problem (the translation invariance is broken in your calculation, which is revealed by missing 3 phonons at zero frequency exactly). You can fix that after your Abinit calculation by using the post-processing tool anaddb and re-enforce the ASR by using the flag asr https://docs.abinit.org/variables/anaddb/#asr.
b) if after imposing the asr it is still a problem then there is probably a problem of convergence of your calculation.
Note that the ASR can be violated a bit a by few cm-1, this still gives a good result for the phonons.
ASR can be imposed directly in the abinit DFPT calculation (https://docs.abinit.org/variables/eph/#asr) but if an ASR violation is present you will not notice it, which can be usefull information; I advise to impose it after DFPT calculation with anaddb for example.
Best wishes,
Eric
Negative frequencies corresponds actually to imaginary frequencies.
1) These values can appears if your system is not stable w.r.t. atomic displacements (unstable phonon), meaning your structure is not the ground state one. You can find the ground state structure by displacing the atoms following the eigenvector given by the unstable mode(s) (also called mode condensation). Then you know your structure is stable once no "negative" frequency is present in your calculation. In this case these negative frequencies are physically relevant.
2) Your system is supposed to be stable locally or supposed to be the ground state and you still find negative values for phonon frequencies. Here two source cases can be discussed:
a) Your calculation gives the so called "acoustic sum rule" (ASR) problem (the translation invariance is broken in your calculation, which is revealed by missing 3 phonons at zero frequency exactly). You can fix that after your Abinit calculation by using the post-processing tool anaddb and re-enforce the ASR by using the flag asr https://docs.abinit.org/variables/anaddb/#asr.
b) if after imposing the asr it is still a problem then there is probably a problem of convergence of your calculation.
Note that the ASR can be violated a bit a by few cm-1, this still gives a good result for the phonons.
ASR can be imposed directly in the abinit DFPT calculation (https://docs.abinit.org/variables/eph/#asr) but if an ASR violation is present you will not notice it, which can be usefull information; I advise to impose it after DFPT calculation with anaddb for example.
Best wishes,
Eric
-
saikrishna
- Posts: 15
- Joined: Fri Jun 21, 2019 12:11 pm
Re: Phonon calculations
Hey Eric,
Thank you for your reply and now I have a doubt with the code which I have attached and so please help me in solving the issue.
Thank you for your reply and now I have a doubt with the code which I have attached and so please help me in solving the issue.
Re: Phonon calculations
Dear Sai krishna,
Could you tell us a bit more about what you want to calculate, system, etc?
Thanks,
Eric
PS: note that you posted in the GW forum while it is a problem of phonon/linear response you are asking for. Please be careful to choose the right forum topic that is in line with your problem as much as possible. Thanks!
Could you tell us a bit more about what you want to calculate, system, etc?
Thanks,
Eric
PS: note that you posted in the GW forum while it is a problem of phonon/linear response you are asking for. Please be careful to choose the right forum topic that is in line with your problem as much as possible. Thanks!