problem with 'fix orient/fcc'

Dear all users,

I am using the ‘fix orient/fcc’ to calculate GB mobility of fcc material.
The fix adds an orientation-dependent potential energy to atoms
(pe.png),

where (order.png)is the order parameter of each atom i. (For more details,

please read the manual and the publicattion(Nature Mater 2006,5:124-7)).

Then by evaluating the derivative of the artificially added potential energy, we can

get the additional force on each atom
(deriva.png),

While I have some doubts about the above derivative, in my opinion, the derivative should be
(deriva_me.png),

i.e. without (deriva_part.png). Is it right?

I think it should be my wrong understanding of the derivative that leds to my doubts.
However, I still want to know how to specifically get the derivative as given in the publication.

PS:If figures of function inserted in this email cann’t be displayed, please enter this email.
user:yangliang_lammps@…543…
password:yangliang123

Any comments from you will be sincerely appreciated.

Sincerely,
Liang Yang

pe.png

order.png

deriva.png

deriva_me.png

deriva_part.png

PS:If figures of function inserted in this email cann't be displayed,

please enter this email.
user:[email protected]...
password:yangliang123

congratulations. you just announced to the entire online world how your
e-mail account can be hacked.
as if there aren't enough compromised e-mail accounts spewing out spam
already.

axel.

If you think the paper is wrong, or the derivative is
incorrect, then send an email to one of the lead authors
and ask them. I doubt that is the case.

Steve

Dear Steve,

I have sent emails to more than one lead authors about
a month ago, but I have not received any reply. So I have
to turn to lammps-users for help.

I know my problem is silly for knowledgeable people. I
have read some data about derivative of discrete function,
but I still can not figure out how to get the derivative
given in that publication.

Sincerely,
L.Yang

It’s fine if you want to send to the mail list.
I just think it unlikely that anyone here will dig into
the techincal issues you ask about, deep enough
to give a good answer.

Steve