ERROR on proc 9: Internal error (../tri_mesh_I.h:109)

Submitted by zen on Tue, 11/26/2019 - 15:50

Hi There,

I am trying to simulate many particles with rotating mesh, there are both interactions between particle-particle and particles to the wall. The particles are attempt into different sizes of groups from less than 100 up to 10000, some of these were successfully compiled but some were getting errors before end up with final run steps and the errors were mostly shown as 'ERROR on proc 9: Internal error (../tri_mesh_I.h:109)'. I am trying to figure this out by changing the timestep and run steps accordingly, but that was showing less confidence from the compilation. Please, could anyone help with those errors? Many thanks in advance.

Cheers,

Zen

zen | Mon, 12/02/2019 - 17:12

Hi mschramm,

Thanks for your prompt reply, please find the simplified input script below. As I was suggested that no matter how many particles have inserted and time step has been given, the errors always happened quite often. Please suggest if there was a potential error.

atom_style granular

#atom_modify sort 0 0
atom_modify map array

hard_particles yes

boundary m m m

newton off

communicate single vel yes

units si

region reg block -0.4 0.4 -0.4 0.4 -0.4 0.4 units box

create_box 1 reg

neighbor 0.002 bin

neigh_modify delay 0

#Material properties required for new pair styles Hertz
fix m1 all property/global youngsModulus peratomtype 1e10
fix m2 all property/global poissonsRatio peratomtype 0.22
fix m3 all property/global coefficientRestitution peratomtypepair 1 0.66
fix m4 all property/global coefficientFriction peratomtypepair 1 0.62

fix 11 all property/global minSeparationDistanceRatio scalar -
#(value=value for the minimum separation distance, recommended as 1.01)
fix 12 all property/global maxSeparationDistanceRatio scalar -
#(value=value for the maximum separation distance, recommended as 1.1)
fix 21 all property/global surfaceLiquidContentInitial scalar -
#(value=value for the initial surface liquid volume in % of the solid volume)
fix 22 all property/global surfaceTension scalar -
#(value=value for the surface tension of liquid (SI units N/m))
fix 31 all property/global fluidViscosity scalar -
#(value=value for the fluidViscosity (SI units Pas))
fix 32 all property/global contactAngle peratomtype -
#(value_i=value for contact angle of atom type i and fluid)
fix 41 all property/global maxLiquidContent peratomtype -
#(value_i=value for maximum liquid content of an atom of type i in % of particle volume)

#New pair style - contact model is hertz + washino capillary viscous model (Hertzian with cohesion)
pair_style gran model hertz tangential history cohesion washino/capillary/viscous
pair_coeff * *

timestep 0.0000001

fix gravi all gravity 9.81 spherical 0.0 210.0 # v_gx v_gy v_gz spherical v_theta v_phi vector 0.0 0.0 -1.0
#-------------------------------------------------------------------------

#-------------------------------------------------------------------------
#import mesh from cad:
fix cad1 all mesh/surface file meshes/NAME.stl type 1 rotate axis - - - angle -

#use the imported mesh as granular wall
fix granwalls all wall/gran model hertz tangential history cohesion washino/capillary/viscous mesh n_meshes 1 meshes cad1

#-------------------------------------------------------------------------
#region for insertion
region bc block - - - - - - units box

#-------------------------------------------------------------------------
group nve_group region reg

#distributions for insertion
fix pts1 all particletemplate/sphere 15485863 atom_type 1 density constant - radius constant -
fix pdd1 all particledistribution/discrete 15485867 1 pts1 1.0

#particle insertion
fix ins nve_group insert/pack seed 32452843 distributiontemplate pdd1 &
maxattempt 100 insert_every 3000 overlapcheck yes all_in yes vel constant 0.0 0.0 -0.2 &
region bc particles_in_region 100 ntry_mc 1000000

#apply nve integration to all particles that are inserted as single particles
fix integr nve_group nve/sphere

#output settings, include total thermal energy
compute rke all erotate/sphere
thermo_style custom step atoms ke c_rke vol
thermo 1000
thermo_modify lost ignore norm no

run 1

dump dmp all custom/vtk 10000 post/cohesion_*.vtk id type type x y z ix iy iz vx vy vz fx fy fz omegax omegay omegaz radius
dump dump all mesh/stl 10000 post/dump*.stl

#insert particles
run 1
unfix ins

#run
run 1500000 upto

variable omega1 equal 6.283

#Moving mesh
fix move all move/mesh mesh cad1 rotate/variable origin - - - axis - - - omega v_omega1

run 20000000

mschramm | Tue, 12/10/2019 - 17:19

Hello,
Unfortunately, nothing is jumping out at me. Without the cad file I wont be able to do a full debug. Below are some general areas that I have to double check every so often.

fix cad1 all mesh/surface file meshes/NAME.stl type 1 rotate axis - - - angle -
Is you cad dimensions in the correct format? (there is no scale keyword)

Looking at the source code of tri_mesh_I.h, it seems that a point in the mesh is not being resolved correctly.

zen | Wed, 12/11/2019 - 16:38

Hi Mschramm,

Thanks for your reply, as the reply or new comment, cannot attach additional files, then I have opened the new topic with attached a new full input script and mesh in zip format which is named in 'Internal error (../tri_mesh_I.h:109)'.

Thanks!

saman | Thu, 12/12/2019 - 20:48

Hi zen
According to the software documentation:
For washino / capillary / viscous cohesion model this model can ONLY be used as part of pair gran, not as part of a fix wall/gran.

zen | Mon, 01/06/2020 - 16:47

Yes, you are right Saman. Although it compiled for fewer particles within a wet mesh and gives a reasonable appearance, technically, it restricted by liggghts program. Thank you for the notice!