[QE-users] Regarding initialization of elastic constants for NEB code

[Tone Kokalj]

On Thu, 2019-10-10 at 00:21 +0000, Jana, Soumyajit wrote:
> Hi,
>  
> I am using NEB code to calculate vacancy jump activation energy in
> pure copper. I am not sure how to initialize elastic constants  (
> k_max and k_min)

Here is a simple empirical recommendation based on my experience: do
not initialize them, but rely on default values - they are OK for most
purposes (K_max and k_min are not really that crucial). But if you want
to initialize them nevertheless, then your k_max is OK, but k_min
appears too tiny, use something like 0.1 instead.

Regards, Tone
-- 
Jožef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia 
(tel: +386-1-477-3523 // fax: +386-1-251-9385)

 
>  
> BEGIN
> BEGIN_PATH_INPUT
> &PATH
>    restart_mode  = 'from_scratch'
>    string_method  = 'neb'
>    nstep_path     = 55
>    num_of_images  = 9
>    opt_scheme     = 'broyden'
>    CI_scheme      = 'auto'
>    path_thr       = 0.05
>    k_max          = 0.3
>    k_min          = 0.001
>    ds             = 0.1
>    first_last_opt = .false.
>    minimum_image  = .true.
>    use_freezing   = .true.
>    use_masses     = .true.
> /
> END_PATH_INPUT
>