Subject: Bond Atom Missing Error in SWM4-NDP Water Simulation on Graphene Surface
Dear LAMMPS Community,
I am simulating a water nanodrop placed at the center on a graphene surface and want to evaluate the contact angle. I am using the polarizable SWM4-NDP water model (5000 molecules), with the graphene sheets (14.9 nm × 14.9 nm) modeled as fixed and non-polarizable. The simulation box is periodic in the x, y, and z dimensions, with sufficient vacuum space provided in the z-dimension (perpendicular to the graphene surface). The z-dimension box length is 20nm.
However, I am encountering the following error at step 175580: “Bond atoms *** *** missing on proc ** (…/ntopo_bond_partial.cpp:60).” The bond in question is between the Drude core (oxygen) and the Drude particle. After searching the archives, I understand that potential causes for this error could be a large cutoff (1.5 nm), large timestep (1 fs), or bad dynamics.
For reference:
- The cutoff value and timestep are based on the original SWM4-NDP paper (Lamoureux et al., 2006) and a more recent publication (Misra et al., JPCC 2017).
- I have also tried a smaller timestep of 0.5 fs but still encounter the error at the different step (after ~0.1 ns).
The error consistently occurs at varying steps depending on the number of processors and the use of GPUs. For instance, I’ve used proc = 8, 16, and 18 with the command ‘-sf gpu -pk gpu 0,’ and the error still occurs at different steps. Up to the point of error the total energy, water temperature are approximately constant with time. I did not encounter the error when the number of water molecules was 2000,4000.
I am currently using the LAMMPS version from August 2, 2023 on ubuntu 2020.04.6. I would appreciate any insights into the possible source of this error or suggestions on how to resolve it. Below is the input I am using
package gpu 0 pair/only on
units real
boundary p p p
atom_style full
bond_style harmonic
angle_style harmonic
special_bonds lj/coul 0.0 0.0 0.5
pair_style lj/cut/coul/long 15.0
kspace_style pppm 1.0e-4
log WCA-0.log
read_data data-p.lmp
#######################################################
pair_coeff * * 0 0
pair_coeff 2 2 0.210939 3.183950 # ODw ODw SWM4-NDP
pair_coeff 2 4 0.107200 3.3215 # CA1 ODw (Misra et.al 2017)
pair_coeff 2 5 0.107200 3.3215 # CA2 ODw (Misra et.al 2017)
###########################################################
group ATOMS type 1 2 3
group CORES type 2
group DRUDES type 6
group SOL type 1 2 3 6
group GRA type 4 5
variable TK equal 298.0
variable TDK equal 1.0
fix DRUDE all drude N C N N N D
delete_bonds ATOMS multi
comm_modify vel yes
velocity ATOMS create ${TK} 12345
velocity DRUDES create ${TDK} 12345
variable posx equal -xcm(SOL,x)
variable posy equal -xcm(SOL,y)
compute TEMP SOL temp/drude
dump dmp all dcd 1000 WCA-0.dcd
fix TEMP SOL langevin/drude ${TK} 100. 1256 ${TDK} 20. 13977 zero yes
fix RIGID ATOMS rigid/nve/small molecule
fix NVE DRUDES nve
fix hold SOL spring tether 1000 0.0 0.0 NULL 0.0
thermo_style custom step etotal ke pe ebond evdwl ecoul elong press c_TEMP[1] c_TEMP[2] v_posx v_posy
thermo 1000
run 1000000
Best regards,
Binu Varghese
PhD scholar
IIT Madras
)