The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <mmikami@rc.m-kagaku.co.jp>]

Dear Henry,

Before talking about ABINIT, we should talk about Physics...
Your problem is really depends on which frequency range
you are talking about.

In general, before running calculations,
we might want to consult with good database
(experimental facts !) such as "Landolt-Boernstein" series.
In the volume III-15b, you will find "Optical constants of metals".

If you are talking about IR region,
you might want to learn about Drude model, e.g., by Googling,
http://www.mse.ncsu.edu/WideBandgaps/classes/MSE%20704/Handouts/ 
OpticalProperties.pdf
The dielectric properties in IR range is beyond the present ABINIT...
(Please consult with the papers about the implementation in ABINIT,
 before you use the ABINIT functions ...)

As I talked in the last Summer School at UCSB:
ABINIT is really versatile, like a swiss army knife.
But the effectiveness will depends on users !
(We do not blame a swiss army knife
 when it cannot cut beef cattle　into pieces of beef, do we ?)

Good luck,
Masayoshi

On 2006/03/16, at 19:18, lfe11@yahoo.com.tw wrote:

> Dear all:
>
> I had the same problem to calculate the frequency-dependent dielectric  
> tensor of silver bulk.We know that the silver bulk has 11 electrons,I  
> must set nband at least to 6.But I got the similar ERROR showed that :
>
>  In non-linear calculation, nband must be equal
>  to the number of valence bands.
>
> So how to set nband? Or what should I do?
>
> Thanks for your help.
>                                                  Henry Chen