Temperature and Energy increases with "nve/limit" command

Dear Lammps users,

  I am working on dynamics of fluids in carbon models. The carbon
model remains fixed during simulation. I am using "NVE/limit 0.05" for
the fluid molecules for equilibration and to remove any high energy
configurations. However, the temperature and energy (both KE and PE)
increase during the simulation.

I am attaching the input script . Any help will be appreciated.

Dear Lammps users,

  I am working on dynamics of fluids in carbon models. The carbon
model remains fixed during simulation. I am using "NVE/limit 0.05" for
the fluid molecules for equilibration and to remove any high energy
configurations. However, the temperature and energy (both KE and PE)
increase during the simulation.

​there is no reason why this should not happen. nve/limit only applies
restrictions when atoms reach the threshold. otherwise it will behave just
like fix nve.
a more effective way to remove initial high potential energy spots would be
to run a minimization first.

​axel.​

Dear Axel,

Many thanks for your reply. I found that the some atoms were
experiencing large forces during the simulation. So, I tried using a
soft potential. Below is shown part of my input script :

pair_style lj/cut/coul/long/soft 2.0 0.5 10.0 10.0 12.0

read_data lammps_data.dat

pair_coeff * * 0.0 0.0 0.95
pair_coeff 1 8 0.0546 3.08 0.95
pair_coeff 2 8 0.00 1.4 0.95
pair_coeff 3 8 0.0744 3.275 0.95
pair_coeff 4 8 0.00 1.4 0.95
pair_coeff 5 8 0.106 2.88 0.95
pair_coeff 6 8 0.0955 2.9 0.95
pair_coeff 7 8 0.04 2.610 0.95

pair_coeff 1 9 0.0934 3.205 0.95
pair_coeff 2 9 0.00 1.525 0.95
pair_coeff 3 9 0.127 3.4 0.95
pair_coeff 4 9 0.00 1.525 0.95
pair_coeff 5 9 0.181 3.005 0.95
pair_coeff 6 9 0.163 3.025 0.95
pair_coeff 7 9 0.0685 2.735 0.95

pair_coeff 8 8 0.053654 2.8 0.95
pair_coeff 8 9 0.0917 2.925 0.95
pair_coeff 9 9 0.1569 3.05 095

This is working fine. When I switch to full strength of the potential
from 0.95 to 1.0, I get an error

Dear Axel,

Many thanks for your reply. I found that the some atoms were
experiencing large forces during the simulation. So, I tried using a
soft potential. Below is shown part of my input script :

pair_style lj/cut/coul/long/soft 2.0 0.5 10.0 10.0 12.0

read_data lammps_data.dat

pair_coeff * * 0.0 0.0 0.95
pair_coeff 1 8 0.0546 3.08 0.95
pair_coeff 2 8 0.00 1.4 0.95
pair_coeff 3 8 0.0744 3.275 0.95
pair_coeff 4 8 0.00 1.4 0.95
pair_coeff 5 8 0.106 2.88 0.95
pair_coeff 6 8 0.0955 2.9 0.95
pair_coeff 7 8 0.04 2.610 0.95

pair_coeff 1 9 0.0934 3.205 0.95
pair_coeff 2 9 0.00 1.525 0.95
pair_coeff 3 9 0.127 3.4 0.95
pair_coeff 4 9 0.00 1.525 0.95
pair_coeff 5 9 0.181 3.005 0.95
pair_coeff 6 9 0.163 3.025 0.95
pair_coeff 7 9 0.0685 2.735 0.95

pair_coeff 8 8 0.053654 2.8 0.95
pair_coeff 8 9 0.0917 2.925 0.95
pair_coeff 9 9 0.1569 3.05 095

This is working fine. When I switch to full strength of the potential
from 0.95 to 1.0, I get an error

----------------------------------------------------------
Step c_mytemp PotEng KinEng Fmax E_vdwl E_coul E_pair E_bond E_angle
E_long E_tail

  100000 302.94912 -28974.945 1914.4316 202.90136
-2193.2657 -1299.8602 -29793.114 401.92992 416.23932
-26299.988 0

  100100 306.07354 -28993.423 1934.1758 193.19966
-1846.9499 -1704.2272 -29852.518 412.88093 446.21454
-26301.341 0

ERROR on proc 0: Bond atoms 6303 6305 missing on proc 0 at step 100130
(../ntopo_bond_all.cpp:63)
------------------------------------------------------------
---------------------------------------------

My question is after equilibration using (lamda = 0.95) I switch back
to full potential (lambda = 1) and got this error. I will appreciate
your help.