Nesting Function problems

[LHDphy]

First, thanks for your nice software. It is useful for us to plot the fermi surface of some materials.

However, when we use the nesting function of the lastest version of 2.4.0 and after clicking 'delta*delta', then we put 'doubledelta.frmsf' into 'fermifer.exe', nothing is describled. Screenshots are supported.

We hope to obtain your help. Thanks again.

Best regards,
HD Liu.

[mitsuaki1987]

If you want to know the k-points that have high nesting function, you need to set the threshold ("Fermi energy" box in your second image) to some appropriate value.
We can see the maximum value of the nesting function in the text area below the display area. And please try, for example, 95 % of the max. of the nesting function.
Then, you can see small pieces in that the nesting function is large.

You should set the "color scale mode" to "Band index" to represent clearly.

[kraamesh57]

Dear HD Liu and Professor Kawamura,
Thanks for initiating the discussion about computing the fermisurface nesting function. Liu, I got stuck with the first step itself and could not create the double delta—fmrsf file which you shared here. I am attaching the picture. Let me know whether we must choose certain values which i encirled in the attached picture to make it work. If I click double delta or lindhard, it just hangs and does not respond.
Thank you

[LHDphy]

Dear Professor Kawamura,
First, thanks for your reply! It works!
Second, '95 % of the max. of the nesting function' (Fig.1) is necessary? I think other values (Fgi.2) are more suitable. May I can it?
Third, I also use the 'lindhard' (Fig/.3), but 'the max. of the nesting function" is 0. What should I do?

[LHDphy]

Dear kraamesh57,
It is true that the function will work long time on the PC. Moreover, it is not necessary to set other values what you indicated.

Best regards,
HD Liu.

[kraamesh57]

Dear Liu
Thank you for the message and suggestion. Yes. This time, I patiently waited for a couple of hours for the program to complete the calculation, and I succeeded in generating the double delta function. I also observed that when I used 30 to 40 per cent of the maximum value as Fermi Energy, complex surface filling than when I chose 95 per cent value as Fermi level. Hope Prof Kawamura will give us some guidance regarding this. I have two more questions.

1. The maximum eigenvalue of the double delta operator is more than 20000, and I hope this value is not in energy units. (Fig1)
2. Fermi surface nesting will have two parallel lines along a particular direction in the projection Fermi surface diagram. How do we read the Fermi surface nesting function in the reciprocal space? Is it a peak structure along a specific direction? (Fig. 2 and 3)
3. It would be great if Prof. Kamamura showed an example system to help experimentalists like me.
   Thank you so much for giving us this free software.

Regards
K Ramesh Kumar

[LHDphy]

Dear K Ramesh Kumar,
The problem what you present may be caused by the target 'band ubdex', you should set it according the first answer.
Moreover, a extremely dense grid results to a long time 'a couple of hours', may be set to the sparse?
Last, what is displayed is relative strength of nesting, we can adjust it.

Regards
HD Liu

[kraamesh57]

Dear Liu
Thank you for your answer. I, too, have the same question which you asked Prof Kawamura a couple of weeks ago. Setting the Fermi Level 95 % per cent of the eigenvalue necessary? Suppose setting the Fermi Level 95 % of the eigenvalue and observing tiny packets confirms the nesting in the system ? I hope Prof Kawamura will help us and give some direction regarding this question.
Regards
Ramesh

[mitsuaki1987]

Dear @kraamesh57 and @LHDphy
This function needs a long computational time, and we can use OpenMP parallelization to accelerate that if you have PC with multi CPU-cores.

The setting of the threshold is arbitrary. For example, you should vary it as 100%→99%→98%..., and check how the isoserface varies. The unit of the double-delta function is $\textrm{energy}^{-2}$.

[mitsuaki1987]

"95 % of the max." means 16.128×0.95 for @LHDphy 's case and 21690×0.95 for @kraamesh57 's case.

[kraamesh57]

Dear Prof. Kawamura,

Thank you for your suggestions. I followed your instructions and set the maximum eigenvalue to 95 percent. As a result, I noticed a small surface anomaly. During my phonon calculation, a negative frequency appeared at the M point, coinciding with the presence of a small surface anomaly in that direction.

I'm sorry for asking what may be basic questions, but I'm trying to make sense of these observations. Your guidance is greatly appreciated.
Thank you