reax/c and quenching

Dear All users

I have a system which contains carbon and hydrogen atoms and I want to quench it to 0 K using reax/c potential, here is the content of my input file for this simulation:

Dear All users

I have a system which contains *carbon *and *hydrogen *atoms and I want
to quench it to 0 K using reax/c potential, here is the content of my input
file for this simulation:

-------------------------------
variable set index relax
variable end index 200000
variable dts index 200

units real
boundary p p p
atom_style charge
atom_modify map array

read_data data.${set}

variable rigidfile atomfile group-rigid.\{set\} variable stochfile atomfile group\-stoch\.{set}

group stoch variable stochfile

pair_style reax/c NULL
pair_coeff * * ffield.reax H C

fix 01 all qeq/reax 1 0.0 10.0 1e-6 reax/c
fix 02 stoch nve
fix 03 stoch langevin 0.00 0.00 0.0001 529304

timestep 0.1
-----------------------------------------------

but while running the simulation I got this error:
********
ERROR: Atom count is inconsistent, cannot write restart file
(../write_restart.cpp:246)
********

I have tried the same simulation with another many-body potential (comb3)
without any error but while I want to quench the same system with reax/c I
got the error.

Does anyone have any idea where this error comes from?

​you have asked this before and received answers and explanations already.
what else do you want?

axel.​

because those answer didn’t help, I was wondering if someone else has better idea!

because those answer didn't help, I was wondering if someone else has
better idea!

​if this is the case, then you should at least say so. yet, i see that you
are *still* doing in your input what you were advised to not do. if you
don't follow advice, you need to explain why the given advice is bad. also
*didn't help* is a very bad description, you should at least explain what
didn't help how. what else you tried etc.

good and helpful answers require good questions and people that follow
advice. http://lammps.sandia.gov/guidelines.html

if you ask the same question again, only because you got an answer you
don't like (and that is the impression you are projecting), then you will
only achieve to irritate people as it is a sign of disrespect. you are
*extremely* unlikely to get anything else, particularly since in your case
you got a response from the maintainer of the USER-REAXC package and the
lead developer of LAMMPS. how much more of an expertise on LAMMPS issues
with reax/c are you looking for?

axel.

thanks for the information and I will take them into consideration for my future posts; I didn’t mean any disrespect for sure, sorry If the way I asked the question lead to any bad impression.