How to set the values of DmRngeXd

[mailoliver]

Dear all:

I have tried to calculate the absorption spectrums of semiconductors using yambo. I find that the lines become smooth, when I set a large range of DMRngeXd.
I think it need a convergence test for DmRngeXd. It's right?
But I confuse how to set the firt value of DmRngeXd.
For LiF tutorial, in 02_RPA_no_LF:
% DmRngeXd
0.10000 | 0.30000 | eV # [Xd] Damping range
in 06_BSE:
% BDmRange
0.15000 | 0.30000 | eV # [BSS] Damping range
For 2D_Si_surface tutorial, in 02_RPA_no_LF_par:
% DmRngeXd
0.030000 | 0.10000| eV # [Xd] Damping range

I need your help. Thanks.

Tieyu Lu

[Daniele Varsano]

Dear Teyu Lu,
this is not a parameter that have to be carried to convergence.
It is an artificial damping that it is useful to add in order to compare with experiments.
The results of your calculations are excitations energies and oscillator strengths, and you are
not calculating line widths while in experiments the widths are determined by lifetimes and
finite resolutions. So, the range of the damping is a parameter that you can adjust for
your needs.

Cheers,

Daniele

[mailoliver]

Dear Daniele:
Thank you for your reply.
I find that when I use a large range damping, it needs less kpoints to get a smooth line of absorption spectrum.
So I can use less kpoints with a large range damping to simulate absorption spectrum. It is OK ?

Best wish

Tieyu Lu

[Daniele Varsano]

Dear Tieyu Lu,

of course with a large damping you are smoothing the spectrum,
but remember that your smoothing is artificial, and with a small
k-point sampling you could miss some physics.

bests,

Daniele

[mailoliver]

Dear Daniele:

Thank you for your reply.

Tieyu Lu