IHPC-IMS Program onAdvances & Mathematical Issues
in Large Scale Simulation(Dec 2002 - Mar 2003 & Oct - Nov 2003)
Seminar:Multiscale Modeling of Heterogeneous
Granular Systems
Alberto M. CuitiñoMechanical and Aerospace Engineering
Rutgers UniversityPiscataway, New [email protected]
Institute of High Performance Computing Institute for Mathematical Sciences, NUS
Singapore 2003 cuitiño@rutgers
Collaborators
• Gustavo Gioia • Shanfu Zheng
Singapore 2003 cuitiño@rutgers
Rutgers1. Harvard University
2. William and Mary
3. Yale University
4. Princeton University
5. Columbia University
6. University of Pennsylvania
7. Brown University
8. Rutgers University (1766)
9. Dartmouth
RutgersUniversit
y
1766 Rutgers Founded as
Queen’s C
ollege
1864 Named New
Jerse
y’s Land Gran
t Colle
ge
1989 Rutgers is
electe
d into
Associa
tion of A
merican
Universitie
s
RU
1970 University
of Medici
ne and Dentist
ry Founded
2003
Richard L. M
cCormick
19th president of R
utgers
50,000 Students
10,000 Faculty
and Staf
f
175 Academ
ic Depart
ments
1869 First A
merican
College F
ootball Gam
e,
(6) Rutgers
vs. Prin
ceton (4
)
1914 School of E
ngineering Nam
ed Separate S
chool
1956 The C
olleges
and Schools of R
utgers
Become th
e Stat
e Univers
ity of N
ew Je
rsey
Singapore 2003 cuitiño@rutgers
Rutgers
RutgersUniversity
Philadelphia
NYC
New Brunswick
Camden
Newark
Singapore 2003 cuitiño@rutgers
College AveLivignstonDouglassBusch
Rutgers,New Brunswick
Cook
Singapore 2003 cuitiño@rutgers
Rutgers,Busch
StadiumGolfScience and Engineering
Hairston Leads Rutgers Past Navy 48-27SEPTEMBER 27, 2003
Singapore 2003 cuitiño@rutgers
Rutgers,Mechanical and Aerospace
Rutgers,Engineering
Entrance
Doyle D. Knight
Michael R. Muller
Timothy Wei
Abdelfattah M.G. Zebib
Norman J. Zabusky
Jerry Shan
Tobias Rossman
S. Bachi
Fluid Mechanics
Zhixiong (James) Guo
Yogesh Jaluria
Constantine E. Polymeropoulos
Kyung T. Rhee
Stephen D. Tse
Thermal Sciences
Haim Baruh
Hae Chang Gea
Noshir A. Langrana
Constantinos Mavroidis
Madara M. Ogot
Dajun Zhang
Design and Dynamics
Haym Benaroya
William Bottega
Alberto Cuitiño
Mitch Denda
Ellis Dill
Andrew Norris
Kook Pae
Assimina Pelegri
George Weng
Solids, Materials and Structures
Singapore 2003 cuitiño@rutgers
Current Research
• Granular Systems (G. Gioia and S. Zheng)• Crystal Plasticity• Multiscale Modeling • Foam Mechanics• Folding of Thin Films• Microelectronics• Digital Image Correlation• Computational Material Design
(Ferroelectric Polymers)
Support from NSF, DOE, DARPA, FAA, NJCST, IFPRI, CAFT is gratefully acknowledged
Singapore 2003 cuitiño@rutgers
Damage due to Electromigration in Interconnect Lines
Sequence of pictures showing void and hillock formation in an 8µm wide
Al interconnect due to electromigration
(current density 2x107 A/cm², temperature 230°C)
Thomas Göbel (t.goebel@ifw-dresden .de), 18.04.02
Singapore 2003 cuitiño@rutgers
E 0,
V
T
Schimschak and Krug, 2000
Schimschak and Krug, 2000
Singapore 2003 cuitiño@rutgers
Grain Boundary Effects
Grain 1 Grain 2
VOID TRAPPINGby GRAIN BOUNDARY
Initial DefectVOID MOTION
@ GRAIN BOUNDARY
VOID RELEASEFrom GRAIN BOUNDARY
e-Atkinson and Cuitino ‘03
Singapore 2003 cuitiño@rutgers
Goal
Understand and quantitatively predict the MACROSCOPIC
behavior of powder systems under compressive loading based on
MICROSCOPIC properties such as particle/granule behavior and spatial arrangement
Load
Need for MULTISCALE Study PARTICLES POWDERS (discrete) (continuum)
Singapore 2003 cuitiño@rutgers
Background
10-4 10-3 10-2 10-1
Normalized Compaction Force
0.4
0.5
0.6
0.7
0.8
0.9
1
Re
lativ
eD
en
sity
MacroscopicCompaction Curve
1st Stage 2nd Stage
Compaction Force
3rd Stage
0th Stage
Singapore 2003 cuitiño@rutgers
Stages
Mixing Die Filling Rearrangement
Large Deformation Localized Deformation
Singapore 2003 cuitiño@rutgers
Identifying Processes and Regimes
Mixing
Transport
Granulation
Characteristics:
• Large relative motion of particles
• Differential acceleration between particles
• Large number of distinct neighbors
• Low forces among particles
• Long times, relatively slow process
• Quasi steady state
Discharge
Die Filling
Vibration
Characteristics:
• Large relative motion of particles
• Differential acceleration between particles
• Large number of distinct neighbors
• Low forces among particles
• Short times
• Transient
Early Consolidation
Pre-compression
Characteristics:
• Limited relative motion of particles
• Low particle acceleration
• Same neighbors
• Quasi-static
• Low forces among particles
• Small particle deformation (elastic)
Consolidation
Characteristics:
• No relative motion of particles
•Low acceleration
• Same neighbors
•Quasi-static
• Sizable forces among particles
•Small particle deformation (elastic + plastic)
Compact Formation
Characteristics:
• No relative motion of particles
• Low acceleration
• Same neighbors
•Quasi-static
• Large forces among particles
• Large particle deformation
Singapore 2003 cuitiño@rutgers
Identifying Numerical Tools (which can use direct input from finer scale)
Mixing
Transport
Granulation
PD/DEM/MC
Discharge
Die Filling
Vibration
PD/DEM/MC
Ballistic Deposition
Early Consolidation
Pre-compression
PD/DEM/MC
Consolidation
GCC
Compact Formation
GQC
OUR SCOPE
Numerical tools appropriate for process
Singapore 2003 cuitiño@rutgers
Identifying Numerical Tools (which can use direct input from finer scale)
Mixing
Transport
Granulation
PD/DEM/MC
Discharge
Die Filling
Vibration
PD/DEM/MC
Ballistic Deposition
Early Consolidation
Pre-compression
PD/DEM/MC
Consolidation
GCC
Compact Formation
GQC
OUR SCOPE
Numerical tools appropriate for process
Singapore 2003 cuitiño@rutgers
Die Filling
Numerical Experimental
Numerical Experimental
Cohesion No Cohesion
Open Configuration Dense Configuration
Singapore 2003 cuitiño@rutgers
Identifying Numerical Tools (which can use direct input from finer scale)
Mixing
Transport
Granulation
PD/DEM/MC
Discharge
Die Filling
Vibration
PD/DEM/MC
Ballistic Deposition
Early Consolidation
Pre-compression
PD/DEM/MC
Consolidation
GCC
Compact Formation
GQC
OUR SCOPE
Numerical tools appropriate for process
Singapore 2003 cuitiño@rutgers
Rearrangement
Video ImagingGlass Beads, Diameter = 1.2 mmGioia and Cuitino, 1999
Increasing Pressure Increasing Pressure
Process by which open structures collapse into dense configurations• Cohesive Powders are susceptible to rearrangement while• Non-Cohesive Powders are not
X-Ray Tomography-Density MapsAl2O3 Granules. Diameter = 30 micronsLannutti, 1997
Punch
Singapore 2003 cuitiño@rutgers
A physical description
Energy landscape exhibits a Spinoidal Structure (nonconvex)
H H
Convexification implies coexistence of two phases
H
Total
Singapore 2003 cuitiño@rutgers
A relaxation mechanism
Particle Rearrangement Mechanism
Snap-Through of Rings (Kuhn et al. 1991) Ring Structures in Cohesive Powders
Numerical
Experimental
Singapore 2003 cuitiño@rutgers
Relaxation process
Numerical
Experimental
Singapore 2003 cuitiño@rutgers
Experiments and Theory
Al2O3
Theoretical Experimental
Kong et al., 1999
Singapore 2003 cuitiño@rutgers
2D: simulation and experiment
Singapore 2003 cuitiño@rutgers
Rearrangement Front
Experiment Simulation
Singapore 2003 cuitiño@rutgers
“Grains”
Singapore 2003 cuitiño@rutgers
2D Simulations (Size Distribution)
Singapore 2003 cuitiño@rutgers
3D Simulations
Singapore 2003 cuitiño@rutgers
Mueth, Jaeger, Nagel 2000
Comparison with Experiment
Experiment Simulation
Singapore 2003 cuitiño@rutgers
Further Predictions
Experiment Simulation
Singapore 2003 cuitiño@rutgers
Particle Rearrangement 3D
• Homogeneous particle size;
• r = 0.5 mm;
• Particles = 120,991
mmr 5.0
Singapore 2003 cuitiño@rutgers
Quantitative Predictions
Nc
0 10 20 30 40 504
4.5
5
5.5
6
6.5
u=2.5u=5.0u=7.5
u=9.1(d)
u=0.25
Nc = 6.27
h/
0 10 20 30 40 50
0
0.5
1
u=2.5u=5.0u=7.5
u=9.1
(c)
u=0.25
v/
0 10 20 30 40 50
0
0.5
1
u=2.5u=5.0u=7.5u=9.1
(b)
u=0.25
0 10 20 30 40 500.45
0.5
0.55
0.6
0.65 = 0.627
=0.51
u=2.5u=5.0u=7.5u=9.1
(a)
•Rearrangement front;
•Density increase;
•Relative movement stops;
•Contact number increase;
Singapore 2003 cuitiño@rutgers
Heterogeneous System(Same Material)
Without rearrangement After rearrangement
•Log-normal distribution; d = 2.16 ~ 9.10 mm; particles=13,134
Singapore 2003 cuitiño@rutgers
Multiphase Systems
Singapore 2003 cuitiño@rutgers
Identifying Numerical Tools (which can use direct input from finer scale)
Mixing
Transport
Granulation
PD/DEM/MC
Discharge
Die Filling
Vibration
PD/DEM/MC
Ballistic Deposition
Early Consolidation
Pre-compression
PD/DEM/MC
Consolidation
GCC
Compact Formation
GQC
OUR SCOPE
Numerical tools appropriate for process
Singapore 2003 cuitiño@rutgers
Constrain kinematics of the particles by overimposing a displacement field described by a set of the displacements in a set of points (nodes) and a corresponding set of interpolation functions (a FEM mesh)
A quasi-continuum approach
FEM Mesh Set of Particles Combined System
Granular Quasi-Continuum
Singapore 2003 cuitiño@rutgers
Governing Equations
PVW
Euler Equation
P
8m2P
P
8n2Vm21î wmn +
P
8m2Pfm áî um = 0
P
8n2Vm21
drmndî wmn
+ fm = 0
Local Equilibrium
Singapore 2003 cuitiño@rutgers
Force Fields
Indentation ()
Fo
rce
0 0.1 0.2 0.3
-1200
-1000
-800
-600
-400
-200
0
d
r1 r2
= r1 + r2 - d
Hertziancontact
Similarity solution
volumechanged
• Elastic contact follows Hertz contact law;
• Plastic deformation follows similarity solution; Contacts on each particle are independent.
• Volume change after inter-particle voids are filled in.
Singapore 2003 cuitiño@rutgers
Role of FF parameters - y
Force (kN)
Den
sity
(g/c
m3)
0 2 4 6 8 10
0.5
0.6
0.7
0.8
0.9
1
HDPE 100
E = 1 GPa
n = 3
= 0.3
0 = 1.1 g/cm3
y = 1 (in MPa)
y = 0.1
y = 0.01• Effect of yielding stress is significant;
• Lower y yield higher deformation under the same pressure and thus higher density;
• Solidification force differ significantly but the solidification density relative unchange.
Singapore 2003 cuitiño@rutgers
Role of FF parameters: hardening
Force (kN)
Den
sity
(g/c
m3)
0 2 4 6 8 10
0.5
0.6
0.7
0.8
0.9
1
HDPE 100
E = 1 GPa
y = 1 MPa
= 0.3
0 = 1.1 g/cm3
n = 3
n = 10
n = • Effect of hardening
parameter n is significant;
• Soft material (n) is easy to be solidified.
Singapore 2003 cuitiño@rutgers
Role of FF parameters:Poisson’s ratio
Force (kN)
Den
sity
(g/c
m3)
0 2 4 6 8 10
0.5
0.6
0.7
0.8
0.9
1
HDPE 100
E = 1 GPa
y = 1 MPa
n = 3
0 = 1.1 g/cm3
= 0.3 = 0.2
Singapore 2003 cuitiño@rutgers
Case of Study: Multiphase System
Singapore 2003 cuitiño@rutgers
Spatial Distribution
Phase I
Variant 1
Phase I
Variant 2
Phase I
Variant 3
Singapore 2003 cuitiño@rutgers
Spatial Distribution
Phase II
Variant 1-4
Singapore 2003 cuitiño@rutgers
Spatial Distribution
Phase III
Needs an uniform distribution
Singapore 2003 cuitiño@rutgers
Multiphase SystemRearrangement
Full mixed + Cohesion force
Singapore 2003 cuitiño@rutgers
Multiphase SystemPost-Rearrangement
Input for GQC
Singapore 2003 cuitiño@rutgers
Calibration of FF
stress (MPa)
De
nsi
ty(g
/cc)
0 1 2 3 4 5 6 70.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
Sample:mass = 0.75 gsize = 12126.6 (red)
12125.8 (blue)density = 0.789 g/cc (red)
0.898 g/cc (blue)Particles:
size = 0.216 ~ 0.91 mmnumber = 13,134
Detergent Granule 1
Singapore 2003 cuitiño@rutgers
Multiphase SystemComparison with Experiment
stress (MPa)
density
(g/cc)
0 0.5 1 1.5 2 2.5 30.6
0.7
0.8
0.9
1
1.1
1.2
1.3
Sample :mass = 0.075 (g)size = 553.3 (mm3)density = 0.91 (g/cc)particle size = 0.1~0.5 (mm)particles = 7,986
• Density diversity at initial state is mainly due to the irregular shape of real particles;
• At early stage of experiment the deformation is the mainly from the particle rearrangement.
MACROSCOPIC Behavior
Singapore 2003 cuitiño@rutgers
Multiphase System:density evolution
0
2
4
6
Z
0
2
4
6
8
10
X
0
2
4
6
8
10
Y
X Y
Z
dens: 0.68 0.78 0.87 0.97 1.06 1.16 1.25 1.35
Movie Here
Full Field Predictions
Singapore 2003 cuitiño@rutgers
Multiphase System:pressure evolution
X
0
2
4
6
8
10
Y
0
2
4
6
8
10
Z
0
2
4
6
X Y
Z
Movie Here
Singapore 2003 cuitiño@rutgers
Granular Quasi-Continuum
• Allows for explicit account of the particle level response on the effective behavior of the powder
• Provides estimates of global fields such as stress, strain density
• Is numerically efficient, can also be improved by using stochastic integration
• Provides variable spatial resolution
• Is not well posed to handle large non-affine motion of particles
• Particle deformation is only considered in an approximate manner (as in PD/DEM)
GOOD BAD
Singapore 2003 cuitiño@rutgers
Towards Computationally Aided Material Design
100 nm
CASCADE OF SCALES
MICRO NANO ATOMISTIC
20 m 0.1 nm
FORCE PARAMETERS
NANO COMPOSITE PARAMETERS
MICRO SCALE
ACCEPATLE MAXIMUM PORE SIZE
PORE SIZE
ACCEPTABLERANGE OF FORCE PARAMTERS
ACCEPTABLE RANGE OF
NANOCOMPOSITE PARAMETERS
TO NANO SCALE
Singapore 2003 cuitiño@rutgers
Summary and conclusions
• Powder compaction is a complex process where many dissimilar entities (particles) consolidate by various concurrent mechanisms.
• In the low pressure regime, rearrangement and localized particle deformation dominates the mechanical response.
• In this initial regime, compaction proceeds in a discontinuous fashion by an advancing front.
• The physics of the rearrangement can be traced to a spinoidal structure in the energy density of the system.
• This process can be theoretically described using the framework of non-convex analysis.
• The effect of particle deformability and die wall roughness are incorporated into the analysis in a clear and physical manner.
• Numerical simulations verify the theoretical model• Experimental studies validate the model and simulations• 3D simulations show a similar behavior that 2D ones, indicating the same physics
operates in 2D and 3D cases.