[Pw_forum] help: Pt phonon calculation

[Stefano Baroni]

On Sep 17, 2006, at 3:13 PM, hailin yu wrote:

> Dear all:
>       I have some questions about the phonon calculation of Pt.
>       1). Is it reasonable to calculate the phonon dispersion of Pt  
> by using the Pt.pbe-rrkjus.UPF pseudopotentials?

why not?

>       2). For the value of degauss, how it is reasonable? And for  
> the kpoint?

Only you can tell. This is not so difficult to assess. Take a few  
representative points in the BZ and calculate the phonon frequencies  
as functions of degauss. For each value of degauss, keep increasing  
the number of k points until you reach convergence. Note: (1) the  
resulting number of k-pts should increase with decreasing degauss;  
(2) this is the order to follow (i.e. degauss first, convergence wrt  
to k-points for each given value of degauss: not the other way  
around). Finally choose the value of degauss (and the corresponding  
number of k-pts) which gives you the accuracy you desire.

In normal circumstances, such as Pt, I would say, the optimal value  
of degauss should not depend on the phonon wavevector. At a Kohn  
anomaly, instead, a much smaller degauss (and a correspondingly  
larger number of k-pts) than at "ordinary" phonon wave-vectors is  
usually needed. Can you see why? (if you understand this you  
understand a lot of the fine details of phonon calculations in metals).

Take care
Stefano

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

Please, if possible, don't  send me MS Word or PowerPoint attachments
Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html



-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060917/e3be13db/attachment.htm