hbondchk failed while using Hybrid reaxFF and LJ potential

Hi,

I am trying to use hybrid LJ and reaxFF potential in my structure of Boron Carbon Hydrogen (all reaxff) and Argon(L J). But I am getting the error step1830-hbondchk failed: H=6 end(H)=250 str(H+1)=175. This error is occurring in different steps for different structure. All structures are almost same just different number of atoms. Mincap and safezon did not help. Changing neighbor distance did not help. Only time I did not get that error when I used atom_modify sort 0 0.0 and used lammps in serial in normal desktop computer. But while running in supercomputer with multiple cores the error occurred (for single core no error). The version of lammps I am using in all devices is lammps/16Mar18. I have attached the force field with param and control file, structure, and the input here.

I highly appreciate your help.

Regards,
Nirmal

Input script

units real

atom_style charge

#atom_modify sort 0 0.0

read_data data.BCH

######################################################hybrid potentials

#pair_style hybrid reax/c lmp_control safezone 1000 mincap 100000 lj/cut 11.0

pair_style hybrid reax/c lmp_control lj/cut 11.0

pair_coeff * * reax/c ffield.reax.hcb B C H NULL

##lj

pair_coeff 1 4 lj/cut 0.1254 3.4

pair_coeff 2 4 lj/cut 0.1254 3.4

pair_coeff 3 4 lj/cut 0.1254 3.4

pair_coeff 4 4 lj/cut 0.23983 3.4

data.BCH (31.8 KB)

ffield.reax.hcb (6.26 KB)

in.BCH (1.09 KB)

lmp_control (1009 Bytes)

param_bch.qeq (95 Bytes)

log.lammps (191 KB)

Hi,

I am trying to use hybrid LJ and reaxFF potential in my structure of Boron Carbon Hydrogen (all reaxff) and Argon(L J). But I am getting the error step1830-hbondchk failed: H=6 end(H)=250 str(H+1)=175. This error is occurring in different steps for different structure. All structures are almost same just different number of atoms. Mincap and safezon did not help. Changing neighbor distance did not help. Only time I did not get that error when I used atom_modify sort 0 0.0 and used lammps in serial in normal desktop computer. But while running in supercomputer with multiple cores the error occurred (for single core no error). The version of lammps I am using in all devices is lammps/16Mar18. I have attached the force field with param and control file, structure, and the input here.

you probably are using too many processors. your system is tiny. there is not much chance to do efficient MPI parallelization with LAMMPS on such a tiny system. also multi-threading will not be likely to provide much speedup beyond using a moderate number of threads. you cannot parallelize well, if there is not a lot of work to pass around.

axel.

Your potential file does not define any hydrogen bond type, so you can change the value of hbond_cutoff to 0.0 in the control file, thus disabling H-bond calculations. It should resolve your problem.

Michal

Hi Kanski,

After changing the hbond_cutoff to 0.0 it resolved the error. Thanks for the response. I was wondering how it would effect my calculation but it seems almost same in hydrogen bonding as before.

The issue pointed out by Axel, I am also not sure about that. But looking at the atoms interaction it seems hybrid potential is doing ok with the present input.

Thank you anyway for your valuable suggestion.

-Nirmal