reax/c with pair_style hybrid

Lukose_K_P · September 22, 2014, 6:14pm

Dear All,

I am seeking help for the following problem:

Trying to use both reax/c and lj using pair_style hybrid (or hybrid/overlay):

pair_style hybrid lj/cut/coul/cut 15.0 reax/c NULL

I only want to use reax for the 6th atom type. I have also defined bond, angle and dihedral styles between other atom types and between other atom types and atom type 6 (not between atoms of type 6). The following shows the assignment of pair_coeffs:

pair_coeff 1 1 lj/cut/coul/cut 0.070 3.55
pair_coeff 2 2 lj/cut/coul/cut 0.066 3.50
pair_coeff 3 3 lj/cut/coul/cut 0.250 3.55
pair_coeff 4 4 lj/cut/coul/cut 0.030 2.42
pair_coeff 5 5 lj/cut/coul/cut 0.030 2.50
pair_coeff * * reax/c ffield.reax NULL NULL NULL NULL NULL C

This shows the error “Non-existent ReaxFF type”.

When i change the order to

pair_coeff * * reax/c ffield.reax C C S H H C #Note that the arguments are different
pair_coeff 1 1 lj/cut/coul/cut 0.070 3.55
pair_coeff 2 2 lj/cut/coul/cut 0.066 3.50
pair_coeff 3 3 lj/cut/coul/cut 0.250 3.55
pair_coeff 4 4 lj/cut/coul/cut 0.030 2.42
pair_coeff 5 5 lj/cut/coul/cut 0.030 2.50

and use hybrid or hybrid/overlay, it runs, however gives undesired results as the parameters for atom types 1-5 are dummy.

For the following definition of pair_coeffs and use with “hybrid” (not “hybrid/overlay”):

pair_coeff * * reax/c ffield.reax C C S H H C
pair_coeff 1 1 lj/cut/coul/cut 0.070 3.55
pair_coeff 2 2 lj/cut/coul/cut 0.066 3.50
pair_coeff 3 3 lj/cut/coul/cut 0.250 3.55
pair_coeff 4 4 lj/cut/coul/cut 0.030 2.42
pair_coeff 5 5 lj/cut/coul/cut 0.030 2.50
#pair_coeff 6 6 lj/cut/coul/cut 0.070 3.55
pair_coeff 1 2 lj/cut/coul/cut 0.0679 3.525
pair_coeff 1 3 lj/cut/coul/cut 0.132 3.55
pair_coeff 1 4 lj/cut/coul/cut 0.0458 2.985
pair_coeff 1 5 lj/cut/coul/cut 0.0458 3.025
pair_coeff 1 6 lj/cut/coul/cut 0.070 3.55
pair_coeff 2 3 lj/cut/coul/cut 0.1284 3.525
pair_coeff 2 4 lj/cut/coul/cut 0.0445 2.96
pair_coeff 2 5 lj/cut/coul/cut 0.0445 3.0
pair_coeff 2 6 lj/cut/coul/cut 0.0679 3.525
pair_coeff 3 4 lj/cut/coul/cut 0.0866 2.985
pair_coeff 3 5 lj/cut/coul/cut 0.0866 3.025
pair_coeff 3 6 lj/cut/coul/cut 0.132 3.55
pair_coeff 4 5 lj/cut/coul/cut 0.030 2.46
pair_coeff 4 6 lj/cut/coul/cut 0.0458 2.985
pair_coeff 5 6 lj/cut/coul/cut 0.0458 3.025

it prints “Setting up run …” and then terminates (I also tried to run on single processor).

I appreciate any help!

Regards,
Lukose

akohlmey · September 23, 2014, 2:11pm

Dear All,

I am seeking help for the following problem:

Trying to use both reax/c and lj using pair_style hybrid (or
hybrid/overlay):

pair_style hybrid lj/cut/coul/cut 15.0 reax/c NULL

please generate a small test input (using only 2 or 3 atom types and a
small number of atoms) that exhibits the same problem and post it
here, so people can look into it. thanks,
axel.

Lukose_K_P · September 24, 2014, 12:02am

Hi All,

I am including a simple problem.
This input has four atom types. My original intention is to use reax
only for the second atom type.

units real
atom_style full

#pair_style hybrid/overlay reax/c NULL lj/cut/coul/cut 20.0
pair_style hybrid reax/c NULL lj/cut/coul/cut 20.0
bond_style harmonic
angle_style harmonic
dihedral_style opls

read_data data.input

bond_coeff 1 513.075 1.386
bond_coeff 2 290.71 1.726
bond_coeff 3 447.90 1.436
bond_coeff 4 370.49 1.080

angle_coeff 1 86.33 111.46
angle_coeff 2 39.565 110.82
angle_coeff 3 35.250 124.04
angle_coeff 4 86.33 91.36

dihedral_coeff 1 0 9.51 0 0
dihedral_coeff 2 0 9.51 0 0
dihedral_coeff 3 0 9.51 0 0
dihedral_coeff 4 0 0 0 0

pair_coeff * * reax/c ffield.reax C C S H #NULL C NULL NULL
pair_coeff 1 1 lj/cut/coul/cut 0.070 3.55
#pair_coeff 2 2 lj/cut/coul/cut 0.070 3.55
pair_coeff 3 3 lj/cut/coul/cut 0.250 3.55
pair_coeff 4 4 lj/cut/coul/cut 0.030 2.42
pair_coeff 1 2 lj/cut/coul/cut 0.070 3.55
pair_coeff 1 3 lj/cut/coul/cut 0.1323 3.55
pair_coeff 1 4 lj/cut/coul/cut 0.0458 2.985
pair_coeff 2 3 lj/cut/coul/cut 0.1323 3.55
pair_coeff 2 4 lj/cut/coul/cut 0.0458 2.985
pair_coeff 3 4 lj/cut/coul/cut 0.0866 2.985
#pair_coeff * * reax/c ffield.reax C C S H #NULL C NULL NULL

group greax type 2
compute 1 all pair reax/c
fix 1 all nvt temp 300.0 300.0 200.0
fix 2 greax qeq/reax 1 0.0 10.0 1.0e-6 reax/c
dump 1 all xyz 1 out_reax.xyz
dump_modify 1 element C C S H

timestep 1.0
thermo 1000
run 1000

This prints "Setting up run ..." without continuation.

The data input is
single molecule

        14 atoms
        14 bonds
        14 angles
        16 dihedrals

           4 atom types
           4 bond types
           4 angle types
           4 dihedral types

-100 100 xlo xhi
-100 100 ylo yhi
-100 100 zlo zhi

Masses

  1 12.010
  2 12.010
  3 32.065
  4 1.008

Atoms

       1 1 1 -0.1819 0.356343 -1.288038 3e-06
       2 1 2 0.0748 -0.0 0.0 1e-06
       3 1 3 -0.1496 1.090491 0.709526 -0.0
       4 1 2 0.0748 2.180983 0.0 2e-06
       5 1 1 -0.1819 1.825342 -1.286756 5e-06
       6 1 2 0.0748 -2.343965 0.869384 1e-06
       7 1 3 -0.1496 -3.434455 0.159858 -2e-06
       8 1 2 0.0748 -4.524945 0.869384 -0.0
       9 1 1 -0.1819 -4.168605 2.157421 4e-06
      10 1 1 -0.1819 -2.699605 2.156141 5e-06
      11 1 4 0.1561 2.489037 -2.150137 6e-06
      12 1 4 0.1561 -0.187334 -2.231614 5e-06
      13 1 4 0.1561 -2.074203 3.047653 8e-06
      14 1 4 0.1561 -4.712283 3.100997 5e-06

Bonds

       1 1 1 2
       2 3 1 5
       3 4 1 12
       4 2 2 3
       5 2 3 4
       6 1 4 5
       7 4 5 11
       8 2 6 7
       9 1 6 10
      10 2 7 8
      11 1 8 9
      12 3 9 10
      13 4 9 14
      14 4 10 13

Angles

       1 1 1 2 3
       2 2 1 5 4
       3 3 1 5 11
       4 2 2 1 5
       5 4 2 3 4
       6 1 3 4 5
       7 3 5 1 12
       8 4 6 7 8
       9 2 6 10 9
      10 1 7 6 10
      11 1 7 8 9
      12 2 8 9 10
      13 3 9 10 13
      14 3 10 9 14

Dihedrals

       1 2 1 2 3 4
       2 1 1 5 4 3
       3 3 1 5 4 3
       4 4 2 1 5 11
       5 2 2 3 4 5
       6 1 3 2 1 5
       7 3 3 2 1 5
       8 4 4 5 1 12
       9 2 6 7 8 9
      10 4 6 10 9 14
      11 1 7 6 10 9
      12 3 7 6 10 9
      13 1 7 8 9 10
      14 3 7 8 9 10
      15 2 8 7 6 10
      16 4 8 9 10 13

When i use
pair_coeff * * reax/c ffield.reax NULL C NULL NULL
it prints "Non-existent ReaxFF type".

I appreciate any suggestion.

Best,
Lukose

Ray_Shan · September 24, 2014, 5:10pm

ReaxFF is a pure atomic potential meaning you must not define bonds, angles, and dihedrals. Please remove bond_, angle_, and dihedral_ styles from your input script and try again.

Ray

Lukose_K_P · September 24, 2014, 5:50pm

Hi,

Thanks for the suggestion!

The same problems persists even without bonds, angles and dihedrals.

Lukose

Ray_Shan · September 24, 2014, 5:58pm

That is not possible. What version of LAMMPS are you using?

I can run your script without any problems (of course with bonds, angles and dihedrals removed).

Ray

Lukose_K_P · September 25, 2014, 5:04am

I use 1Feb14 version of LAMMPS. What version did it work for?

Thanks for helping!
Lukose

Ray_Shan · September 25, 2014, 12:58pm

I use 1Feb14 version of LAMMPS. What version did it work for?

Since the 4May14 version, which was when reax/c was enabled to work with
hybrid or hybrid/overlay.

Please note that the doc page and advertised features/capabilities always
refer to the most current version, thanks.

Ray

Lukose_K_P · September 25, 2014, 9:52pm

Thanks a lot! That helped!
It also worked with bonds, angles and dihedrals assigned for non-reax
atoms (of course there is a warning!).