E_pair, Total Energy and Pressure showing as NaN.


I was trying to thermalize my system which has 5 types of atoms ( Cu, Cu, H, Cu and O). initially last two types of atoms have charge while other don’t have any charge. I am using this code for the thermalization:

-----------Initialize simulation-----------------

log copper_box_300k.log
units metal
dimension 3
boundary p p p
atom_style charge
atom_modify map array
atom_modify map array sort 0 0

-------------Create atoms and box-------------------------

read_data main2.R extra/atom/types 2
read_data main1.R add append offset 3 0 0 0 0 shift -0.5 -2.0 100
mass 1 63.546
mass 2 63.546
mass 3 1.008
mass 4 63.546
mass 5 15.999
pair_style comb3 polar_off
pair_coeff * * ffield.comb3 Cu Cu H Cu O
neighbor 0.5 bin
neigh_modify every 20 delay 0 check yes

velocity all create 700 25674 dist gaussian
minimize 1.0e-8 1.0e-8 1000000 1000000
timestep 0.001
reset_timestep 0
thermo 20
fix 1 all npt temp 700 700 0.1 iso 0.0 0.0 10
fix 3 all momentum 1 linear 1 1 1
fix 4 all qeq/comb 20 0.0001

write_data Cu34.R
run 500
print “Done”

When I am running this in lammps it shows the the E_pair, Total Energy and Pressure as NaN and terminates the program saying “Bad termination of one of your application process”. Can you please help me on this purpose?

It most probably means you have bad dynamics.

Which means bad initial structure (including bad initial charges), bad potential (COMB3 may not be fitted for your particular structure), and bad set up (your timestep is too large for starters, qeq frequency too small, etc).


Dear all,

I'm trying to combine fix setforce and fix aveforce in a slab
simulation to let the confining walls diffuse in the x and y
direction, while keeping them fixed in the z direction:

diff.txt (391 Bytes)

log.fail (12.4 KB)

in.ok (1.54 KB)

log.ok (13.9 KB)

in.fail (1.88 KB)


I am not sure if it will work;I think you can try setting velocity to zero before fix setforce.


Dear Vinit, dear all,

Thank you for your feedback.

My problem is that my script does work... unless I also define in the
script some variables, computes, fixes ave/time and fix ave/chunk
which should not affect the dynamics.

I can circumvent the problem by using two separate input files for
equilibration and production, but I would really love to understand
what is happening...

Best regards,