Rigid

LAMMPS LAMMPS Mailing List Mirror
1

Dear all,
I have simulated a system consist of nanotube which it was filled by water inside( SPCE model), my question is:
I used rigid command for nanotube and I expect that nanotube’s atom don’t move , but it didn’t happen…my input is following. what can I do for fix this problem?

Thanks, for your attention,

echo both
boundary p p p
units metal
atom_style full
bond_style harmonic
angle_style harmonic
read_data 16_202_water2.data
group water type 3 4
group nanotube type 1 2
#neighbor 2.0 bin
neigh_modify every 1 one 10000 page 200000 #include all binsize 2.0 #delay 10 check yes exclude all
fix 1 nanotube rigid molecule
#potential===================================================================================================
pair_style lj/charmm/coul/long 10 20

pair_coeff 3 3 0.0 0.0 0 2.5
pair_coeff 4 4 0.1553 3.166 0.25 7.9
pair_coeff 3 4 0.0 1.583 0.0 2.5

#pair_coeff 1 1 0.115918 4.43 0.2 11.075
#pair_coeff 2 2 0.498327 3.34 0.6 8.35
#pair_coeff 1 2 0.240344 3.885 0.36 9.75

pair_coeff 1 4 0.134172 3.798 0.195 9.475
pair_coeff 2 4 0.278191 3.253 0.40 8.1325
pair_coeff 1 3 0.0 2.215 0 5.5375
pair_coeff 2 3 0.0 1.67 0 4.175
kspace_style pppm 1e-4
delete_atoms overlap 1.5 water nanotube mol yes

2

Dear all,
I have simulated a system consist of nanotube which it was filled by water
inside( SPCE model), my question is:
I used rigid command for nanotube and I expect that nanotube's atom don't
move , but it didn't happen..my input is following. what can I do for fix
this problem?

​there are multiple problems here:
- you are confusing "rigid" with "immobile". a rigid object can still
translate and rotate, only it does it as a whole (i.e. rigid) object.
- as stated in the documentation, fix rigid is not active during
minimization
- you are applying time integration to some atoms multiple times in the
script segment after the minimization
- there are two common ways to immobilize atoms: 1) do not include them in
time integration 2) initialize their velocities to zero and set forces to
zero
- you should not need to override the neighbor list settings for "one" and
"page", if your system is set up correctly

​axel.​