Reax problem

Hi,

With your Feb '13 distribution and input files, including in.FC, field.reax.FC and data.FC (we have tried 2 other reax input files from your distribution), I got the following error message:
ERROR: Invalid pair style (force.cpp:168)

I joined your mailing list and got a response saying I didn't compile the reax module. I have since been bumped from your list and cannot rejoin the list.

LAMMPS has since been recompiled. We get the same error.

Here is some information:

The version you are using:
61 harold-l4> grep -i reax lmp_openmpi
Binary file lmp_openmpi matches
62 harold-l4> pwd
/mnt/lustre/usrcta/unsupported/lammps/13feb13reax/lammps-14Feb13/src

The "regular" version of LAMMPS:

78 harold-l4> grep -i reax lmp_openmpi
79 harold-l4> pwd
/mnt/lustre/usrcta/unsupported/lammps/13feb13/lammps-14Feb13/src

Unfortunately this was done for me as I gave up compiling LAMMPS myself.

Brett

Brett I. Dunlap
Code 6189
Chemistry Division
US Naval Research Laboratory
Washington, DC 20375-5320
Phone: (202) 767-3250
Fax: (202) 767-1716
Email: [email protected]

If you type lmp_openmpi -h
do you see your pair style listed (reax or reax/c, you
don’t say which).

If not, you didn’t compile LAMMPS with it.

Steve

Classification: UNCLASSIFIED
Caveats: NONE

Steve,

  I see reax and reax/bonds

-James

100 harold-l4> lmp_openmpi -h | grep --color -i reax
Pair styles: adp airebo beck bop born/coul/long born/coul/msm
born/coul/wolf born buck/coul/cut buck/coul/long buck/coul/msm buck
buck/long/coul/long comb coul/cut coul/debye coul/dsf coul/long coul/msm
coul/wolf dpd dpd/tstat eam/alloy eam/fs eam eim gauss hbond/dreiding/lj
hbond/dreiding/morse hybrid hybrid/overlay lcbop lj96/cut
lj/charmm/coul/charmm lj/charmm/coul/charmm/implicit lj/charmm/coul/long
lj/charmm/coul/msm lj/cubic lj/cut/coul/cut lj/cut/coul/debye
lj/cut/coul/dsf lj/cut/coul/long lj/cut/coul/msm lj/cut lj/cut/tip4p/long
lj/expand lj/gromacs/coul/gromacs lj/gromacs lj/long/coul/long
lj/long/tip4p/long lj/smooth lj/smooth/linear mie/cut morse reax rebo soft
sw table tersoff tersoff/zbl yukawa
Fix styles (upper case are only for internal use): adapt addforce ave/atom
ave/correlate aveforce ave/histo ave/spatial ave/time balance box/relax
deform deposit drag dt/reset efield enforce2d evaporate external gravity
heat indent langevin lineforce MINIMIZE momentum move nph nph/sphere npt
npt/sphere nve nve/limit nve/noforce nve/sphere nvt nvt/sllod nvt/sphere
orient/fcc planeforce press/berendsen print qeq/comb READ_RESTART
reax/bonds recenter RESPA restrain setforce shake SHEAR_HISTORY spring
spring/rg spring/self store/force store/state temp/berendsen temp/rescale
thermal/conductivity tmd ttm viscosity viscous wall/harmonic wall/lj126
wall/lj93 wall/reflect wall/region
101 harold-l4>

Then my guess is you are trying to run
with reax/c and getting the error. If you
are using pair_style reax you should not

get that error.

Steve

Steve,

We compiled the latest distribution of LAMMPS. That distribution has an examples/reax directory. In that directory there is a Fluorographene
subdirectory. We ran that calculation and got the error:

pair_style reax/c NULL
ERROR: Invalid pair style (force.cpp:168)

You are right in.FC contains that line and thus reax/c.

In the examples/reax there is a rdx example. We ran that and got the error:

pair_style reax 6.0 1 1 1.0e-6
ERROR: Invalid pair style (force.cpp:168)

That input does not specify reax/c.

The second message below indicates that reax was compiled.

Thanks,
Brett

Steve,

Please disregard the attached email to your group--if you can. The code was somehow compiled wrong on that computer. I am sorry to have bothered lammps users.

Brett

Brett I. Dunlap
Code 6189
Chemistry Division
US Naval Research Laboratory
Washington, DC 20375-5320
Phone: (202) 767-3250
Fax: (202) 767-1716
Email: [email protected]

Ray can verify that example works with only
the original ReaxFF compiled in. I don’t
think it should be possible to get that error

if your LAMMPS executable was compiled
with the REAX package installed, as indicated
by lmp_foo -h. Can you verify that style_pair.h
has the regular reax pair style listed in it?

Steve

Brett,

The only way to get that error from in.reax.rdx is if your executable
was compiled without pair_reax.h in style_pair.h. Check your build
procedures.

Aidan

Classification: UNCLASSIFIED
Caveats: NONE

Have the LAMMPS developers corrected this bug yet? Another user wants to use REAX/LAMMS and it is compiled within LAMMPS, but she is getting that Invalid pair error.

Here are the pair style compiled with LAMMPS AUG2012:

Pair styles: adp airebo beck bop born/coul/long born/coul/wolf born buck/coul/cut buck/coul/long buck comb coul/cut coul/debye coul/long coul/wolf dpd dpd/tstat eam/alloy eam/fs eam eim gauss hbond/dreiding/lj hbond/dreiding/morse hybrid hybrid/overlay lcbop lj96/cut lj/charmm/coul/charmm lj/charmm/coul/charmm/implicit lj/charmm/coul/long lj/cubic lj/cut/coul/cut lj/cut/coul/debye lj/cut/coul/long lj/cut/coul/long/tip4p lj/cut lj/expand lj/gromacs/coul/gromacs lj/gromacs lj/smooth lj/smooth/linear morse **** reax **** rebo soft sw table tersoff tersoff/zbl yukawa

-James

James,

I have just verified that if 1) the REAX package is successfully compiled,
2) the executable is correctly linked/copied to where the input file is,
and 3) “pair_style reax” is used in the input script, then it works
without a problem.

So if it didn’t work for you or her, one of the above had to be
unsatisfied.

Two suggestions, though. Please update to a recent version since many
bugs, related or unrelated to reax or reax/c, were fixed in the last two
years. Second, I suggest you use “pair_style reax/c” (USER-REAXC
package), which is easier to compile and runs faster.

Ray