Kloss Abq 2011 Final

Enhancing LAMMPS CapabiIities Enhancing LAMMPS CapabiIities
GranuIar ModeIs, Coding Concepts, CAD InteroperabiIity

and CoupIing to Continuum Methods
2nd LAMMPS Workshop AIbuquerque Aug 2011 2nd LAMMPS Workshop, AIbuquerque, Aug 2011
Christoph Kloss*, Christoph Goniva**, Stefan Amberger,
Alice Hager, Andreas Aigner, Michael Friedl, Stefan Pirker***
*christoph.kloss@jku.at, ** christoph.goniva@jku.at, ***stefan.pirker@jku.at
aII CD Laboratory on ParticuIate FIow ModeIIing aII CD Laboratory on ParticuIate FIow ModeIIing
Johannes KepIer University Linz, Austria
www.Iiggghts.com | www.cfdem.com | www.particuIate-fIow.at
OutIine
Open Source DEM and CFD-DEM
OutIine OutIine
1 Introduction 1. Introduction
Our Group, Mission Statement
2. ModeIs
GranuIar Contact ModeIs, Heat Transfer, SPH, MuItisphere ModeI
3. CAD InteroperabiIity
Mesh Import, ParticIe Insertion, Mesh regions, Wear Prediction
4. Features and Coding Concepts
Load BaIancing, Per-ParticIe Properties, Transport Equations g, p , p q
5. CoupIing to Continuum Methods
OpenFOAM ResoIved and UnresoIved CFD DEM MuItiphysics
Christian Doppler Laboratory on Particulate Flow Modelling www.particulate-flow.at 3
OpenFOAM, ResoIved and UnresoIved CFD-DEM, MuItiphysics
I. Introduction
About Us / Mission Statement
About Us
Mission Statement Mission Statement
CD Lab ParticuIate FIow ModeIIing (head: Stefan Pirker)
6 P t D 7 PhD t d t 5 t t d t 6 Post-Docs, 7 PhD students, 5 master students
Goal: Particulate Flow Modelling... Goal: Particulate Flow Modelling...
Research focus:
Fluid flow CFD, LB, SPH
Granular flow CFD, DEM/MD
Heat transfer
Experimental Validation Experimental Validation
...of industrially relevant processes ...of industrially relevant processes
AppIication exampIes incIude:
Blast furnaces, cyclones, pneumatic conveying, fluidized beds, dryers, wet scrubbers,
conveyor and chute systems, tablet pressing, powder sintering, hopper flow,
soil sampling, energy storage, river bed erosion, blood flows...
The LIGGGHTS/CFDEM Codes
Overview Overview
LIGGGHTS = An Open Source C++ MPI paraIIeI DEM code LIGGGHTS = An Open Source, C++, MPI paraIIeI DEM code
LAMMPS IMPROVED FOR GENERAL GRANULAR AND GRANULAR
HEAT TRANSFER SIMULATIONS
CFDEM = CoupIing of LIGGGHTS to CFD code OpenFOAM
CFD-DEM
WWW.LIGGGHTS.COM | WWW.CFDEM.COM
The web pIatform now has about 600 registered users
PubIished in [1, 2, 3, 4, 10, 11, 12, 13, 14, 16]
The LIGGGHTS+CFDEM Codes
Cooperations Cooperations
D

L
a
b
C
D
Further (formaI and informaI) cooperations for LIGGGHTS and CFDEM
The LIGGGHTS Code
ReIation to LAMMPS ReIation to LAMMPS
present
LIGGGHTS
2.0
???
p
LIGGGHTS
...
LIGGGHTS
1.3.1
15.06.2011
LIGGGHTS
1.0
18.04.2010
LIGGGHTS
beta 1
19. Feb. 2010
CurrentIy ~30.000
Iines of code (LOC)
LAMMPS LAMMPS LAMMPS
( )
LAMMPS
24-Jan-2010
LAMMPS
30-Mar-2010
LAMMPS
15-Jun-2011
....
Sandia NationaI Labs AIbuquerque US
Sandia NationaI Labs, AIbuquerque, US
+ community contributions
II. ModeIs
LIGGGHTS ModeIs - ReIeased
SPH - Smoothed ParticIe Hydrodynamics
SPH is well-suited for geometrically complex systems, and cases where
i l diff i f Fi it V l h i t t bl
SPH Smoothed ParticIe Hydrodynamics
numerical diffusion of Finite Volume approach is not acceptable
ReIeased in LIGGGHTS 1.4 - commands pair sph, fix sph/density/continuity,
fix sph/density/corr fix sph/pressure/summation fix waII/sph fix waII/region/sph fix sph/density/corr, fix sph/pressure/summation, fix waII/sph, fix waII/region/sph
PrincipIe of SPH SIoshing Tank with SPH in LIGGGHTS vs.
p g
VoIume of FIuid (VOF) Method in OpenFOAM
Hertz-MindIin (HM) GranuIar Contact ModeI Hertz MindIin (HM) GranuIar Contact ModeI
ij t ij t ij n ij n
vt t k vn n k F ) ( ) ( o o + =

force ngential ta
vel relative
ngential ta
j
overlap
ngential ta
j
force normal
vel relative
normal
j
overlap
normal
j
. .
( ) ( )
8 2
2 2
o o R G S R Y S
n t n n
= =
* * * *
,
,
* *
4
= R Y k
n n
o
( ) ( )
( )( ) ( )( ) 1 2 2 1 2 2 1
1 1 1
2
2
2
1
2
1
2 2
v v
t
|
Y Y Y
e
e
+ +
=
+
=
*
, ,
) ( ln
) ln(
,
,
*
0
6
5
2
3
> = m S
R Y k
n n
n n
|
o
( )( ) ( )( )
1 1 1 1 1 1
1 2 2 1 2 2 1
2
2 2
1
1 1
v v v v
Y Y G
+ = + =
+
+
+
=
*
,
* *
8
6
= R G k
n t
n n
o
|
n restitutio of e coeff ratio Poisson
modulus G...Shear modulus s Y...Young'
2 1 2 1
v
m m m R R R
+ = + =
* *
,
.
*
0
6
5
2 > = m S
t t
|
n restitutio of e...coeff. ratio ...Poisson v
Hertz-MindIin (HM) Contact ModeI Hertz MindIin (HM) Contact ModeI
ij t ij t ij n ij n
vt t k vn n k F ) ( ) ( o o + =

force ngential ta
vel relative
ngential ta
j
overlap
ngential ta
j
force normal
vel relative
normal
j
overlap
normal
j
. .
( ) ( )
8 2
2 2
o o R G S R Y S
n t n n
= =
* * * *
,
,
* *
4
= R Y k
n n
o
( ) ( )
( )( ) ( )( ) 1 2 2 1 2 2 1
1 1 1
2
2
2
1
2
1
2 2
v v
t
|
Y Y Y
e
e
+ +
=
+
=
*
, ,
) ( ln
) ln(
,
,
*
0
6
5
2
3
> = m S
R Y k
n n
n n
|
o
( )( ) ( )( )
1 1 1 1 1 1
1 2 2 1 2 2 1
2
2 2
1
1 1
v v v v
Y Y G
+ = + =
+
+
+
=
*
,
* *
8
6
= R G k
n t
n n
o
|
n restitutio of e coeff ratio Poisson
modulus G...Shear modulus s Y...Young'
2 1 2 1
v
m m m R R R
+ = + =
* *
,
.
*
0
6
5
2 > = m S
t t
|
n restitutio of e...coeff. ratio ...Poisson v
LIGGGHTS ModeIs - UnreIeased
ParticIe Bonds
ParticIe bonds are inter-particIe
ParticIe Bonds
connections abIe to resist
tension/compression tension/compression,
shear,
torsion and
bending
up to a certain breakage limit. Once broken, the bond connection is up to a certain breakage limit. Once broken, the bond connection is
lost permanently. Particle bonds and related models can be used to
modeI soiIs, ,gIued particIes", particIe breakage, crack formation in
b ( l tti b d l) t beams (lattice beam model) etc...
Looking for co-workers to make modeI ready for reIease!
Asaf et. al, Soil & Tillage Research 92 (2007) 227242 Determination of discrete element model parameters required for soil tillage
Christian Doppler Laboratory on Particulate Flow Modelling www.particulate-flow.at
g ( ) p q g
Shmulevich: Soil & Tillage Research 111 (2010) 4153 State of the art modeling of soiltillage interaction using discrete element method
Zhang, Li: Journal of Terramechanics 43 (2006) 303316 Simulation on mechanical behavior of cohesive soil by Distinct Element Method,
13
Pics from: Latham, S, WeatherIy, D.,: Scripting ParaIIeI Discrete EIement SimuIations with ESyS_ParticIe, https://twiki.esscc.uq.edu.au/bin/view/ESSCC/ESySParticIeDownIoad
Lattice Beam ModeI
Dynamic fracture roughness in a concrete microstructure Dynamic fracture roughness in a concrete microstructure
LIGGGHTS wiII incorporate a beam Iattice
modeI (e.g. *,**) used to study fracture and
f t ti h fragmentation phenomena.
* G. LiIIiu, J.G.M. van Mier, 3D Iattice type fracture modeI
for concrete, Engrg. Fract. Mech. 70 (2003) 927-941.
**J.G.M. van Mier, E. SchIangen, A. Vervuurt, Lattice type
Jean-Francois Jerier
EPF Lausanne
ComputationaI SoIid
Mechanics
, g , , yp
fracture modeIs for concrete, in: H.B. MhIhaus (Ed.),
Continuum ModeIs for MateriaIs with Microstructure,
John WiIey & Sons, 1995, pp. 341-377.
Numerical concrete sample
based on a X-ray micro-
tomography image (25mm
3
)
The crack paths (red)
crossing the sampIe
500 000 l di h d 4 illi b hi l
Timoshenko beam theory leads to the
interaction between the spheres
500.000 polydisperse spheres and 4 million beams represent this sample
Each simulation calculated on 60 processors in 3 days
LIGGGHTS ModeIs
Capturing non-sphericity Capturing non sphericity
GranuIar particIes outside Iaboratories are rareIy sphericaI

LIGGGHTS ModeIs
MuIti-Sphere Approach
Step 1 - ParticIe Image by Laser Scanner
Step 2 - Generate MuIti-Sphere SimuIation ModeI
LIGGGHTS ModeIs
Step 3 - LIGGGHTS SimuIation
Show case cylinder: Gravity tilted so that angle of repose forms
SphericaI SphericaI
Non-SphericaI
LIGGGHTS ModeIs
Capturing non-uniformity Capturing non uniformity
In reaIity, no particIe is Iike any other...
Fix particIedistribution/discrete takes particletemplates as input
and serves as input for insertion commands
Fix particIetempIate/sphere and fix particIetampIate/muItisphere Fix particIetempIate/sphere and fix particIetampIate/muItisphere
fix particIedistribution/discrete
FixParticIetempIate **Iist;
Iist[0]
Iist[1]
Iist[2]
fix particIetempIate/sphere fix particIetempIate/sphere fix particIetempIate/muItisphere
50 mass%
uniform size distribution
35 mass%
-
50 mass%
= 2000 kg/m
15 mass%
= 1500 kg/m
35 mass%
= 3500 kg/m
III. CAD InteroperabiIity p y
CAD InteroperabiIity
Importing a Mesh Importing a Mesh
IndustriaI scaIe granuIar probIems need CAD/Mesh interface!
Rotary dryer, ~1M particIes
Work with Diego Peinado
(Intrame Madrid)
Command fix mesh/gran
reads a trianguIar mesh from g
STL or VTK
Command fix move/mesh/gran Command fix move/mesh/gran
transIates/rotates etc the mesh
(Iike fix move)
Generating a Non-triviaI ParticIe Packing Generating a Non triviaI ParticIe Packing
Packing in tetrahedraI mesh
Command region tetmesh
Reads tet mesh from VTK to be
used as a region used as a eg o
Command fix insert/pack
Works simiIar to fix pour Works simiIar to fix pour
Can use arbitrary regions
(aIso region tetmesh)
Generating a ParticIe Stream Generating a ParticIe Stream
ParticIe insertion at surface
Command fix insert/stream Command fix insert/stream
Extrudes trianguIar surface mesh
ParticIes are inserted in packages
Iike with fix pour and are integrated
by the fix with constant veI untiI
veIocity
by the fix with constant veI untiI
they reach the surface mesh. BCs
are enforced at the surface mesh.
IndustriaI AppIication
Wear Prediction at Transfer Chute Wear Prediction at Transfer Chute
Work with Andre KatterfeId (Univ. Magdeburg, Cepartec)
Goniva, C, Katterfeld, A, Kloss, C: Simulation of dust emission and transport and chute wear Goniva, C, Katterfeld, A, Kloss, C: Simulation of dust emission and transport and chute wear
Proc. of 16. Fachtagung Schttgutfrdertechnik Magdeburg, Sept 2011
Wear prediction (Finnie modeI)
Fix mesh/gran/stressanaIysis Fix mesh/gran/stressanaIysis
25 k particIes, 4 processors:
10 sec reaI time (200k steps)
LB on: 1146 sec CPU time
LB off: 1514 sec CPU time
IV. Features and Coding Concepts g p
Dynamic Load-BaIancing
What is Load-BaIancing What is Load BaIancing
How to distribute Ioad between processors?
Without dynamic Ioad baIancing:
With dynamic Ioad baIancing:
Process 0
Process 0
Process 1
Process 1
advancing simuIation time advancing simuIation time
Dynamic Load-BaIancing
SiIo Test Case SiIo Test Case
Strategy: "SimpIe" LB based on number of owned atoms
Speed-up
LB off LB off LB on
Coding Concepts
Strengths and Weak Points for LAMMPS Strengths and Weak Points for LAMMPS
Strengths of coding in LAMMPS
C d t d d d t d Code easy to read and understand
Code easy to modify
Good encapsulation for many modelling tasks Good encapsulation for many modelling tasks
(fixes, computes, pair styles)
But:
Have to write lots of lo-level code (error prone), some of which is redundant
Abstraction Layer I: Per-ParticIe Properties
Motivation Motivation
We need a Iot of different per-particIe properties for our modeIs
Temperature, heat flux, heat source (e.g. from chemical reactions)
M i t t t ( t i ) li id f fil h i ht Moisture content (water in pores), liquid surface film height
Particle Reynolds number
Drag force Magnus force Saffman force exerted by surrounding fluid Drag force, Magnus force, Saffman force,... exerted by surrounding fluid
Surface dust content, dust emitted to fluid
Energy stored in elastic deformation, energy dissipated by contacts...
....
N d d f ti I ti f d I Need dozens of new per-particIe properties for our modeIs
Why not store new properties in Atom cIass?
Error prone: Have to implement low-level code via a new AtomVec class
(pack/unpack, interproc exchange, restart...)
Have to change many cIasses to make new property fully functional Have to change many cIasses to make new property fully functional
(at least Atom, AtomVec, Dump, Set, ComputePropertyAtom, FixAdapt)
This is a recurring task, so we can add an abstraction layer
SoIution: Store it in a new fix property/peratom cIass
I l t I I I d I Implement Iow-IeveI code onIy once
Look-up mechanism via Modify class so each other class can use it
Every modeI (fix) can now request to store a new per-particIe property Every modeI (fix) can now request to store a new per particIe property
Lazy storage: only store a property if any of the models really need it
Need to touch onIy one cIass for new property
(d t i l t d l f fi t / t )
(dump, set... implemented only once for fix property/peratom)
Usage Usage
//register dragforce
if(!dragforce)
{
char* fixarg[11];
fixarg[0]="dragforce";
fixarg[1]="all";
fixarg[2]="property/peratom";
fixarg[4]="vector";
Type of fix
fixarg[5]="no";
fixarg[6]="yes";
fixarg[7]="no";
fixarg[8]="0.";
fixarg[9]="0.";
fixarg[10]="0.";
dragforce = modify->add_fix_property_peratom(11,fixarg);
}
Usage Usage
if(!dragforce)
{
char* fixarg[11];
fixarg[1]="all";
fixarg[4]="vector";
Name of property
fixarg[5]="no";
fixarg[6]="yes";
fixarg[7]="no";
fixarg[8]="0.";
fixarg[9]="0.";
fixarg[10]="0.";
}
Usage Usage
if(!dragforce)
{
char* fixarg[11];
fixarg[1]="all";
fixarg[4]="vector";
fixarg[5]="no";
fixarg[6]="yes";
fixarg[7]="no";
fixarg[8]="0.";
Type and size of data
fixarg[9]="0.";
fixarg[10]="0.";
}
Usage Usage
if(!dragforce)
{
char* fixarg[11];
fixarg[1]="all";
fixarg[4]="vector";
fixarg[5]="no";
fixarg[6]="yes";
fixarg[7]="no";
fixarg[8]="0.";
Settings for forward comm,
reverse comm, restart
fixarg[9]="0.";
fixarg[10]="0.";
}
Abstraction Layer II: "Transport Equations"
There is another recurring task: SoIving an ODE for each particIe
for a quantity, based on a fIux and a source
E g solve for heat transfer moisture evoIution of surface dust E.g. solve for heat transfer, moisture, evoIution of surface dust...
i p
Q Q
dT
c m
- -
+
,
reactions g e
sources to due
generation heat
source pi
contacts by
conduction heat
j i contacts
pj pi
p
p p
Q Q
dt
c m

+ =

. .
,
,

( )
pj pi j i contact
pj pi
pj pi
pj pi
y
T A
k k
k k
Q

-
+
= A
2 1
4
/
,
SoIution: A new fix transportequation/scaIar
Every modeI can request a scalar transport equation to be solved
pj p
The transportequation makes use of multiple fixes of type property/peratom
Abstraction Layer II: "Transport Equations"
There is another recurring task: SoIving an ODE for each particIe
for a quantity, based on a fIux and a source
E g solve for heat transfer moisture evoIution of surface dust E.g. solve for heat transfer, moisture, evoIution of surface dust...
i p
Q Q
dT
c m
- -
+
,
reactions g e
sources to due
generation heat
source pi
contacts by
conduction heat
j i contacts
pj pi
p
p p
Q Q
dt
c m

+ =

. .
,
,

( )
pj pi j i contact
pj pi
pj pi
pj pi
y
T A
k k
k k
Q

-
+
= A
2 1
4
/
,
SoIution: A new fix transportequation/scaIar
Every modeI can request a scalar transport equation to be solved
pj p
The transportequation makes use of multiple fixes of type property/peratom
V. CoupIing to Continuum Methods p g
CoupIing to Continuum Methods
Many problems are better modeled using a continuum approach
O i i t t li i fI id h i Our main interest lies in fIuid mechanics
Weapon of choice: OpenFOAM, www.openfoam.com Weapon of choice: OpenFOAM, www.openfoam.com
Open FieId Operation and ManipuIation
Leader in OpenSource CFD, market share Germany ~30%, US ??
Initiated by Henry WeIIer and Hrvoje Jasak at ImperiaI in the 90s
Constant deveIopment by universities and companies, Constant deveIopment by universities and companies,
~ 2.000.000 LOC according to Jasak (end of 2009)
Capable of doing Finite VoIume, Finite EIement, ParticIe Methods
(Lagrangian tracking DSMC MD) EIectromagnetics FinanciaI
(Lagrangian tracking, DSMC, MD), EIectromagnetics, FinanciaI
How does OpenFOAM (R) work? - PISO AIgorithm How does OpenFOAM (R) work? PISO AIgorithm
Pressure-Implicit Split-Operator (PISO) Algorithm
for (runTime++; !runTime.end(); runTime++) // use the runTime object to controI time stepping
{{
//Iinear momentum equation. The fIux of U, phi, is treated expIicity
fvVectorMatrix UEqn( fvm::ddt(U) + fvm::div(phi, U) - fvm::IapIacian(nu, U) );
soIve(UEqn == -fvc::grad(p)); // soIve using the Iast known vaIue of p, U approx. satisfies momentum.
for (int corr=0; corr<nCorr; corr++) // --- PISO Ioop---- take nCorr corrector steps
{
voIScaIarFieId rUA = 1.0/UEqn.A();
U UA*UE H()
momenum predictor
U = rUA*UEqn.H();
phi = (fvc::interpoIate(U) & mesh.Sf()) ;
for (int nonOrth=0; nonOrth<=nNonOrthCorr; nonOrth++)
{{
fvScaIarMatrix pEqn ( fvm::IapIacian(rUA, p) == fvc::div(phi) ); // set up the pressure equation
pEqn.soIve();
if ( O th N O thC ) hi E fI () // I t th t fI i if (nonOrth == nNonOrthCorr) phi -= pEqn.fIux(); // on Iast non-orth. corr., correct fIux using new p
} // end of non-orthogonaIity Iooping
U -= rUA*fvc::grad(p);
U.correctBoundaryConditions();
} // d f th PISO I
} // end of the PISO Ioop
} // end of the time step Ioop
38
How does OpenFOAM (R) work? How does OpenFOAM (R) work?
fvVectorMatrix UEqn q
(
fvm::ddt(U) + fvm::div(phi, U) - fvm::IapIacian(nu, U)
); );
soIve
(
UEqn == -fvc::grad(p) U q c g ad(p)
);
==
OpenFOAM is using the paradigm of object-oriented
i t k d t I PDE iI d bI
programming to make code to soIve PDEs easiIy readabIe
39
Motivation for using OpenFOAM is manifoId:
for postprocessing of DEM/MD simulations
f i l ti f I d fI id I fI (FV th d) for simulation of coupIed fIuid-granuIar fIow (FV method)
can use the full model portfolio of OF (FE, eIectrostatics,...)
particIe data
LIGGGHTS OpenFoam SoIver
More information and downIoads: www cfdem com
particIe data feedback
More information and downIoads: www.cfdem.com
Post-processing of DEM/MD data
Able to transfer arbitrary particIe properties to OpenFOAM fieIds
Can use arbitrary unstructured meshes
E l E l ti id hi h i i t t t Example: Evaluating voidage, which is an important process parameter
in multiphase flow reactors
SimuIation of CoupIed FIuid-GranuIar FIow: CFD-DEM Method
Able to transfer arbitrary particIe properties to OpenFOAM fieIds
Can use arbitrary unstructured meshes
Abl t f I d CFD DEM d I d CFD DEM Able to perform resoIved CFD-DEM and unresoIved CFD-DEM
-(f1 + f2) f4
f2
f
=
+
(f1 + f2)
-(f3 + f4)
of,1 of,2
f4
f2
f1
f3
DEM with force
coming from CFD
CFD with adapted voidfraction and
momentum sources
=
+
UnresoIved CFD-DEM
TheoreticaI background:
UnresoIved CFD DEM
Navier-Stokes equations for the fIuid in presence of a granuIar phase
( )
( )
u
f f
f f
t
o
o
c
= V +
c
c
0
( )
( ) ( ) ( ) g u u u u
u
f f f f f f
f f
p
t
o o o o
o
+ V + V = V +
c
c
t

fs
K
f
fIuid voIume fraction
u
f
fIuid veIocity
, p stress tensor, pressure
f
fIuid density
K
fs
fIuid soIid momentum exchange term
comprises drag force, Magnus and Saffman force, virtuaI mass force,...
UnresoIved CFD-DEM
TheoreticaI background:
UnresoIved CFD DEM
ExampIe for fIuid soIid momentum exchange K
fs
:
!" #$%"&$ '())*+, -./$ 01"234$ 567&8"17 519 5%6"2:398"&%$ "78$93&8"17 ;<;8$=;>? @78> A> 15 B6%8":/3;$ #%1CD
E1% FG :(HI (H) E1%> FGD :(HI(H)
( ) =

,
4 2
1
1
2
f
p
d f
C o
t
_

|
|
.
|
\
|
+ =
Re
8 4
63 0
4 2
d
.
. C
d
F
=

i
f
o
f
K
( )
(
(

=
|
.
\
Re log 5 . 1
exp 65 . 0 7 . 3
Re
2
10 p
p
_
p f
u u
cell
V
fs
K
J$88$9 =12$%; 303"%3K%$ 39$ K3;$2 17 L388"&$ J1%8M=377 ;"=6%38"17 '$>4> N1&/ 372 O"%%+
(
(
2
exp 65 . 0 7 . 3 _
AppIication ExampIe
SoiI SampIing with "MoIe" / Work with DLR Bremen
modeIs physics
I
SoiI SampIing with MoIe / Work with DLR Bremen
goaI:
-penetration of heated
body into frozen soiI
phenomena:
-heat transfer from
modeIing:
-heat transfer by finite voIume
body to soiI
-soiI mechanics
y
(continuum) approach
-DEM approach for soiI soiI mechanics
phase change
DEM approach for soiI
modeIIing
temperature triggered -phase change
due to heating
b d ti
- temperature triggered
particIe bonds
I d 6 DOF d i
-body motion
45
- coupIed 6 DOF and moving
mesh
AppIication ExampIe
SoiI SampIing with "MoIe" SoiI SampIing with MoIe
Fetch DEM particle data
Transfer data to CFD
Calculate flow
p
Calculate moving mesh Calculate object motion
Calculate particle bonds
additionaI
modeIs
additionaI
modeIs Calculate particle motion Calculate heat transfer
Calculate moving mesh Calculate object motion modeIs
Transfer data to DEM
Calculate volume fraction
and momentum coupling
Fetch CFD data
Transfer data to DEM
start
AppIication ExampIe
SoiI SampIing with "MoIe"
moving mesh appIication: coupIing of Temperature and object motion
SoiI SampIing with MoIe
heated moIe
uniform Iattice bonds break moIe moves
uniform Iattice bonds resist moIe fixed
coId moIe
uniform Iattice bonds resist moIe fixed
time
Reduction of Dust Emission in a Transfer Chute Reduction of Dust Emission in a Transfer Chute
OriginaI Geometry
Reduction of Dust Emission in a Transfer Chute Reduction of Dust Emission in a Transfer Chute
Optimized Geometry
Thank you for your attention! Thank you for your attention!
Questions?
| www.cfdem.com | www.particuIate-fIow.at

Kloss Abq 2011 Final

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Kloss Abq 2011 Final

Caricato da

Copyright:

Formati disponibili

Enhancing LAMMPS CapabiIities Enhancing LAMMPS CapabiIities

GranuIar ModeIs, Coding Concepts, CAD InteroperabiIity

Christian Doppler Laboratory on Particulate Flow Modelling www.particulate-flow.at 15

Potrebbero piacerti anche