Simulation stopped without an error

Dear all user,
I am getting problem simulating a open chain glucose molecule inside zeolite in presence of water. First, I have imported 20 water molecule inside zeolite then minimized it and ran 10000 NVT step. Then I wanted to import a glucose molecule and again wanted to run minimization and equilibriation. But although lammps created the glucose molecule and did the minimization, it does not go for NVT simulation. I have tried several time and each time it stopped without an error. I have attached my input files. It would be highly helpful if anybody can suggest me the way to run the simulation. Thank you.

watermolecule.txt (328 Bytes)

glucose.txt (3.63 KB)

zeolite-2-mod.in (11.5 KB)

zeolite.txt (33 KB)

Dear all user,
I am getting problem simulating a open chain glucose molecule inside zeolite in presence of water. First, I have imported 20 water molecule inside zeolite then minimized it and ran 10000 NVT step. Then I wanted to import a glucose molecule and again wanted to run minimization and equilibriation. But although lammps created the glucose molecule and did the minimization, it does not go for NVT simulation. I have tried several time and each time it stopped without an error. I have attached my input files. It would be highly helpful if anybody can suggest me the way to run the simulation. Thank you.

LAMMPS usually does not stop without an error message. perhaps you are not looking in the right place for the error. also, please always report which LAMMPS version exactly you are using.

there are some issues with your input deck:

  • you are using a rigid water potential, but not using fix shake
  • your molecule files have Masses sections, but you are atom style “full”, where you don’t have per-atom masses, but per-type masses, which have to be specified with the “mass” command in the input
  • since you are keeping your zeolite framework immobile, you can speed up your calculation significantly (and avoid bogus pressure values) by excluding the zeolite zeolite pairs from the neighbor list.

when i make some modifications to your input to account for that, the input fails at the last run command with the following error:
SHAKE stats (type/ave/delta) on step 5146
1 0.958082 0.0244403 120
1 103.174 4.21442
Per MPI rank memory allocation (min/avg/max) = 31.39 | 31.57 | 31.75 Mbytes
Step Temp E_pair E_mol TotEng Press
5146 55.514076 -5.6921784e+16 152.67105 -5.6921784e+16 -7.7840921e+16
ERROR on proc 1: Bond atoms 838 839 missing on proc 1 at step 5147 (src/ntopo_bond_partial.cpp:64)
Last command: run 10000
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 1

please not the extremely negative pressure, which is an indication of some atoms getting too close, which is typically an indication of a problem with the force field parameters. looking through the log, that close contact is already present after the minimization following the insertion of the glucose molecule. so it has to be assumed, that there is something wrong with either the glucose parameters or the glucose zeolite interactions. the latter is more likely.

axel.

p.s.: below is the last part of the input following the various _coeff command, where i made some modifications.

group zeolite type 1 2
fix 2 zeolite setforce 0 0 0
neigh_modify exclude group zeolite zeolite

set type 1 charge -0.75 #set the charge on all atoms of type 1
set type 2 charge 1.50

#create_atoms 0 random 10 523452422 void mol wtr 5234765 #create 10 water atoms in the region void
create_atoms 0 random 20 34534533453 void1 mol wtr 5234765 #Need to corrected
#create_atoms 0 random 20 34534533453 void2 mol wtr 5234765 #Need to corrected
#create_atoms 0 random 1 34534533453 void2 mol glucose 5234765 #Need to corrected

neighbor 1.0 bin
neigh_modify delay 0 every 1 check yes
thermo 200
group water type 3 4
minimize .00001 .00001 100000 100000
fix 1 all nvt temp 300 300 300
fix 4 water shake 1.0e-4 20 1000 b 1 a 1
restart 1000 restart_zeolite
timestep .5
#fix 3 all viscous .8
run 5000
#unfix 3
unfix 4

create_atoms 0 random 1 34534533453 void2 mol glucose 5234765 #Need to corrected
#create_atoms 0 single 12.68 10.27 13.15 mol glucose 349576 # create a glucose molecule at a specified coordinate

group glucose type 5 6 7 8 9 10 11 12
group mobile union water glucose
minimize .00001 .00001 100000 100000
unfix 1

write_data combined.data
reset_timestep 0
fix 1 mobile nvt temp 300 300 300
fix 4 water shake 1.0e-4 20 1000 b 1 a 1
dump replay all custom 1 ./zeolitetest.lammpstrj id xu yu zu
run 10000

Thank you very much Axel,for your kind and huge support. I am very much grateful for your explanation about the issue. I got your point that the negative pressure is due to some mistake in force field parameter. It will be very kind of you if you can little elaborate the sentence " there is something wrong with either the glucose parameters or the glucose zeolite interactions "

Regarding glucose parameter:
I have collected various ff parameter from OPLS and used it for glucose. The equilibrium bond distance, bond angle etc. I have taken from the optimized structure the glucose molecule while the force constant values I took from OPLS data base. If I did something wrong? I know this is a very basic question but I am very beginner to LAMMPS so thought about this.

Regarding glucose zeolite interactions,
I have only LJ parameter for zeolite. I am not quite sure what to do about the interaction, is this a problem in the cross interaction which I have incorporated via “pair_modify mix arithmetic”.

Thank you once again for your help.

PS-The version of LAMMPS I am using is “LAMMPS (16 Mar 2018)”

Rameswar

Thank you very much Axel,for your kind and huge support. I am very much grateful for your explanation about the issue. I got your point that the negative pressure is due to some mistake in force field parameter. It will be very kind of you if you can little elaborate the sentence " there is something wrong with either the glucose parameters or the glucose zeolite interactions "

Regarding glucose parameter:
I have collected various ff parameter from OPLS and used it for glucose. The equilibrium bond distance, bond angle etc. I have taken from the optimized structure the glucose molecule while the force constant values I took from OPLS data base. If I did something wrong? I know this is a very basic question but I am very beginner to LAMMPS so thought about this.

well, the problem is likely due to an “imbalance” between non-bonded lennard-jones and coulomb interactions. if the lennard-jones parameters are not “repulsive enough”, then atoms can get very close and that leads to a so-called “coulomb catastrophe” since that interaction will be highly attractive.

Regarding glucose zeolite interactions,
I have only LJ parameter for zeolite. I am not quite sure what to do about the interaction, is this a problem in the cross interaction which I have incorporated via “pair_modify mix arithmetic”.

well, it is a bad idea to mix-and-match different force field parameters that were not determined to work with each other. particularly that balance between coulomb and lennard-jones is different for different parameterizations, and thus you cannot mix them, but need to have compatible parameter sets for both, lennard-jones and (partial) charges.

this is all that i can explain in the setting of a mailing list and given my time available. you need to discuss this with someone local, that has more experience in using force fields and doing MD simulations, starting with your postdoc adviser or colleagues. i would also recommend to start with something simpler, you input is quite complex and then move to this more complex system later.

axel