Old ReaxFF potential parameter set

Dear Lammps users and Developers
in my knowledge,Dr. Adri van Duin has changed the functional form for ReaxFF around 2005 and the field.reax files befor 2005 don’t work with the current ReaxFF.
but as Aidan said there is some swiches to use for current ReaxFF code provide these parameter set file. I tried to do some test and there is some questions:

1- as I know the field.reax.RDX is for year 2003 “Strachan, A.; et al. Phys. Rev. Lett. 2003, 91, 098301” and I expect to set some option for using old functions, but at the readme.reax said " This version requires hbnewflag = 1" that is for new function. I can’t understand this.
2- as I understand from documentation of reax/c, there is just control file parameters (switches) to make some change. But I couldn’t do any valuable activity with these parameters to use the field.reax that is for before 2005. please advise me to do it correct.

Thanks in Advance

Older force field files of course work with the reax/c implementation
of ReaxFF. Compare the properties to published values to verify.

The "hbnewflag" is for reax, the fortran implementation. For certain
materials, e.g. TATB, hbnewflag has to be set to 1 to match the reax/c
results (which includes the changes by Adri in 2004).


Dear Ray
Thanks for your reply.
I tried to use reax/c and reproduce the properties published but the results
are very strange (even in minimisation of potential energy)
Also I tried to minimise potential energy wit fortran reax but
the simulation stoped by following error. the filed.reax file
is for year 2004 that Dr van Duin has said it is befor of they have changed the functional form of RexFF code. how the reax/c can verify the version of field.reax file and use the related functions.

Setting up minimization …
ia(i1,2) = 21 reax_defs.h::MBONDDEF = 20 after
310 of 528162 pairs completed.
Too many bonds on atom. Increase MBONDDEF


Dear Ray

please apologize me to send this email again.

I guess you did not see my email.

Thanks in advance


Please search the mailing list for "MBONDDEF" to find the solution to
the problem.

It is also hard to comment without more information. You may post a
simple example if you think there is a bug or an inconsistency. But
people here may not have time to test, validate, or verify force