The ABINIT Users Mailing List ( CLOSED )

[lan haiping <lanhaiping@gmail.com>]

Dear All,
  I did several tests on GW calculations with plasom-pole approximation.
I have a question about parameters ecutwfn and ecuteps for screening and self-energy calculations.
 In GW tutotrial, it describes that ecutwfn controls over the number of plane waves for $\chi_0^{KS}$, and ecuteps controns the dimension of 
$chi_0^{KS}$.   When we write the $\chi_0$ explicitely, 
\begin{eqnarray}
\chi_{0GG'}(q,\omega)=\frac{2}{N_k\Omega}\sum_{k}\sum_{j\le N_v}\sum_{N_v<i\le N_b}M_{kij}(q+G)M^*_{kij}(q+G') \times[\frac{1}{\omega-(\epsilon_{kj}-\epsilon_{k-qi})-i\eta}-\frac{1}{\omega-(\epsilon_{k-qi}-\epsilon_{kj})+i\eta}]
\end{eqnarray}
where M_{kij} is  the strength of oscillator defined as 
\begin{equation}
M_{kij}(q+G)=<k-qi|e^{-i(q+G)r}|kj>
\end{equation}

Since the state |kj> is the KS eigenstate, 
we can obviously have  that ecuteps  controls the dimension of 
$G\times G$ for  $\chi_0$ in reciprocal space.  
My question is what does ecutwfn  account for $\chi_0$ ?  From the equation for $\chi_0$ i donot find any extra parameters  to reach final result.  This setting really makes me confusion, any hints and comments are appreciated.

Regards,
Hai-Ping

-- 
Hai-Ping Lan 
Department of Electronics ,
Peking University , Bejing, 100871
lanhaiping@gmail.com, hplan@pku.edu.cn