Lubricate pair style - outputting particle forces

Dear LAMMPS users,

I am using the lubricate pair style to simulate simple shear in dense suspensions. I find that when I output the particle forces (i.e. dump fx fy fz, see input script below), I get non-zero values. Is this what I should expect? It seems that from the equation of motion,

F(dissipative) + F(brownian) + F(applied) = 0,

one would expect the total force on each particle to be zero? And for the case of zero browian and zero applied forces, the forces generated due to the lubricated contacts should balance, giving zero net force on each particle?

Many thanks,

Chris

atom_style sphere
boundary p p p
newton on
communicate single vel yes

region reg prism 0 14 0 14 0 14 0 0 0 units box
create_box 1 reg
lattice sc 0.1

create_atoms 1 box

set region reg diameter 1

pair_style lubricate 1 0 0 1 2
pair_coeff * *

timestep 0.00001

velocity all create 1 4928459 dist gaussian

fix 1 all nve/sphere

compute 1 all erotate/sphere
compute 3 all temp/deform
compute 4 all pressure 3
compute 5 all contact/atom

thermo_style custom step c_3 c_4 c_4[4] vol atoms
thermo 100
thermo_modify lost ignore norm no
compute_modify thermo_temp dynamic yes

dump id all custom 10000 hello.dump id x y z vx vy vz fx fy fz c_5

fix 2 all deform 1 xy erate 0.1 remap v

run 1000000

write_restart restart.restart

It seems that from the equation of motion,
F(dissipative) + F(brownian) + F(applied) = 0,
one would expect the total force on each particle to be zero? And for the case of zero browian and zero applied forces, the forces >generated due to the lubricated contacts should balance, giving zero net force on each particle?

I’m not expert on FLD, but I don’t think that is correct.
Maybe you are thinking of pair lubricateU ? For lubricate
it’s just adding a force and optionally the FLD term. You integrate
with fix nve/sphere. If there were no net force, the particles
wouldn’t move.

Steve

Chris,

You might try contacting Higdon or Kumar, they have been pretty receptive to my questions. Kumar’s thesis is probably the best source of info on the FLD model (they made it). You can find that here, https://www.ideals.illinois.edu/handle/2142/16032.

On your question of the total cancellation of forces. Yes, that is correct, and I am sure you’ve found it recurring throughout SD literature. The reasoning is the same as the overdamped Langevin particle. The approximation made is that particles relax to the fluid velocity almost instantaneously, and thus inertia really doesn’t carry a particle anywhere. So essentially you just solve for the velocity after setting F_hydro=Other forces, and integrate first order ODEs rather than second order.

As for why you are finding non-zero forces, I was really hoping someone else would chime in on this.

Hope some of this helped your sanity.

Thanks Eric and Steve. Ok, I’m glad I wasn’t wrong about those forces cancelling. So the question is why am I calculating these forces as non-zero. Do you have a script that is using the lubricate model (even a hybrid with other models would be useful) that you would be willing to share?

Many thanks,

Chris