Nesting function

[dinhloc]

Dear Samuel and Carla,

I am interested in the nesting function in EPW. From its formula, the calculation is dependent only on the band structure. Is it possible to set up for EPW calculate nesting function without any prior calculation such as electron-phonon coupling? I try to set elph=.false. but epw returns nothing in the output.

In addition, the nesting function defined in EPW is not really representative for the Fermi surface nesting characteristic. The static bare susceptibility is a better parameter to consider the mechanism of the nesting vector. (Equation 1 and 2 in http://journals.aps.org/prb/abstract/10 ... .77.165135 and page 2 in the first paragraph http://journals.aps.org/prl/pdf/10.1103 ... 106.196406).

I saw the similar topic in the forum viewtopic.php?f=3&t=77&p=315&hilit=static+bare+susceptibility#p315. A simple set g=1 in the self_elph.f90 is not bring to the static bare susceptibility. Could you give me some hints to modify the code to calcualate this physical parameter?

With many thanks,
Loc

[dinhloc]

Dear Samuel and Carla,

I have search in the forum about the calculate the bare susceptibility (
2/N(k))*sum[(f(k+qn)-f(km))/(e(k+qn)-e(km)]. The formula is quite similar with the real part of Lamda function.
Is it correct if I change the weight in the selfen_phon.f90: ?

Code: Select all

weight = wkf (ikk) * (wgkk - wgkq) / ( ekk - ekq )

and put g2=1.

Thank you for all your help in advance.

With best wishes,
Loc

[sponce]

Dear Loc,

Sounds good.

The reason you cannot bypass the electron-phonon interpolation (i.e. using elph = .false.) is because the interpolation routines also interpolates the eigenenergies and phonon frequencies.

You can however modify the routine ephwann_shuffle and suppress some of the calls that computes the el-ph matrix elements specifically if you really wish so.

Best,

Samuel