Hi all!
I have been on-off using LIGGGHTS for a few months now, and I have to say I love reading all the new functions and opportunities that come out all the time!
I have recently started a new project in LIGGGHTS that I am a little stuck on how to approach and/or not sure it can be entirely possible!
I am trying to simulate the mixture of very small charged particles (10^5-6 particles on the order of 1-10 um) with some much larger particles (10^3 particles on the order of 5-10 mm). The geometry is a simple rotating drum, and I am quite familiar with implementing the mechanics of the drum in LIGGGHTS, but my error is in being able to run this simulation, I get conflicting neighbor-bin calculations. I am reaching out to see if anyone has ever done anything similar to this? Can LIGGGHTS handle different orders of magnitude of particle sizes like this? And I am still quite unclear on the declaration of a particle's charge, concerned if it is actually working in my original set up.
I will paste the code below, though I'm certain there are many other errors in it than what I am describing!
Thanks for all the help! Love the community!
Cheers,
Quincy
atom_style granular
atom_modify map array
boundary m m m
newton off
communicate single vel yes
units cgs
#centimeter accuracy
region reg cylinder z 0.0 0.0 12 -15 15 units box
create_box 1 reg
neighbor 0.001 bin
neigh_modify delay 0
#Material properties required for new pair styles
fix m1 all property/global youngsModulus peratomtype 5.e6
fix m2 all property/global poissonsRatio peratomtype 0.45
fix m3 all property/global coefficientRestitution peratomtypepair 1 0.3
fix m4 all property/global coefficientFriction peratomtypepair 1 0.5
fix m5 all property/global k_finnie peratomtypepair 1 1.0
fix m6 all property/global characteristicVelocity scalar 2.
fix m7 all property/global cohesionEnergyDensity peratomtypepair 1 600000
#New pair style
pair_style gran model hertz tangential history cohesion sjkr #Hertzian with cohesion
#pair_style gran model hertz tangential history
#Hertzian without cohesion
pair_coeff * *
timestep 0.0001
fix gravi all gravity 98.1 vector 0.0 -1.0 0.0
#the chute
fix cad all mesh/surface/stress file drum1n.stl type 1 wear finnie
fix inface all mesh/surface file meshes/insertion_face.stl type 1
fix granwalls all wall/gran model hertz tangential history mesh n_meshes 1 meshes cad
#region and insertion
group nve_group region reg
region bc cylinder z 0.0 0.0 11 -14 14 units box
group grp1 region reg
group grp2 region reg
#distributions for insertion
fix pts1 grp1 particletemplate/sphere 1 atom_type 1 density constant 89 radius constant 0.6
fix pts2 grp2 particletemplate/sphere 1 atom_type 1 density constant 89 radius constant 0.008
fix pdd1 all particledistribution/discrete 1. 2 pts1 0.9905 pts2 .0005
#particle insertion
fix ins nve_group insert/pack seed 5330 distributiontemplate pdd1 insert_every once overlapcheck yes &
particles_in_region 1000 region bc
#apply nve integration to all particles that are inserted as single particles
fix integr all nve/sphere
#output settings
compute 1 all erotate/sphere
#insert the first particles so that dump is not empty
run 1
dump dmp all custom 1000 post_cgs2/dump*.data_drop id type type x y z ix iy iz vx vy vz fx fy fz omegax omegay omegaz radius
dump dmpVTK grp1 atom/vtk 200 post_cgs2/dumpVTK1_*.vtu
dump dmpVTK2 grp2 atom/vtk 200 post_cgs2/dumpVTK2_*.vtu
dump dumpstress all mesh/gran/VTK 200 post_cgs2/dump*.vtk stress wear cad
run 1000 upto
#moving mesh
fix movecad1 all move/mesh mesh cad rotate origin 0. 0. 0. axis 0. 0. 1. period 16
#insert particles
run 7500000 upto
unfix ins
aaigner | Wed, 09/30/2015 - 11:11
Some suggestions ...
Hi Quincy,
Maybe I can help you with some suggestions ...
boundary m m mif you don't need it. If you 'lose' a particle it can freeze your whole system due to the increased ram usage. ;-). Use 'f' whenever possible.timestep 0.0001is big for your particles or not? Use fix check/timestep/gran to check this.RAJ KASHYAP | Wed, 09/30/2015 - 13:52
HOW TO DECIDE COHESION DENSITY
hey SIR,
If we are using cgs units than what is unit of COHESION DENSITY.Is this j/cm3 or cal/cm3.Suggest some value for cohesion energy density.
Thanks in advance.
quincy.suehr | Fri, 10/02/2015 - 22:48
Replies and Thanks
Raj and Andreas,
Thank you so much for your replies! This has been just the jump I needed to get back on track with my project! If my memory serves me right, CGS units for energy is in ergs which is 10^-7 J so the Cohesion energy density should be in the order of erg/cm3.
As far as the simulation goes, getting the neighbor parameters right is definitely tricky! I have had several errors with too many bins. Has anyone every had any thought of using the SJKR model as sort of a quasi-mesoscale Electrostatic adhesion force model with Cohesive energy? I read in another forum post Christoph was saying adding Van der Waals Forces in macroscopic particles couldn't be done with LJ potential coefficients.
Appreciate all the help!
V/R
Quincy
RAJ KASHYAP | Mon, 10/05/2015 - 12:47
stick to default value
Hey Quincy,
I am also new in this field, but to my understanding you can set the default value of bin i.e 0.1 for cgs.
hope this will work.
ckloss | Sun, 10/11/2015 - 17:27
Hi Quincy,
Hi Quincy,
a size difference of 10^3 will need some model change, this is (from my point of view) beyond the scope of DEM.
If the small particles do not need to be resolved, they could be modelled as scalar transport equation on top of the large particles or with a CFD-like approach
Best wishes
Christoph
flashfigo | Tue, 11/03/2015 - 09:33
Advisory
Hi Quincy,
I'm trying to simulate the charged particles(diameters 1e-6 to 1e-5 m on the order of 1-10 um)， but once I run the simulation the computer will stuck， and the message is "mpirun noticed that process rank 3 with PID 6645 on node flash-Z68M-D2H exited on signal 9 (Killed)."
Could you help me ?