0% found this document useful (0 votes)
387 views

Ansys Workbench Moisture-And-Creep

The document discusses moisture diffusion modeling capabilities in ANSYS MAPDL and solder joint modeling in Workbench. It outlines enhancements in coupled-field diffusion analyses in ANSYS 14.0 and 14.5, including modeling of moisture, sodium, and electromigration diffusion. It provides an example of moisture diffusion modeling and discusses future plans to further couple diffusion with additional fields like electric and structural nonlinearities.

Uploaded by

zhoufl
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
387 views

Ansys Workbench Moisture-And-Creep

The document discusses moisture diffusion modeling capabilities in ANSYS MAPDL and solder joint modeling in Workbench. It outlines enhancements in coupled-field diffusion analyses in ANSYS 14.0 and 14.5, including modeling of moisture, sodium, and electromigration diffusion. It provides an example of moisture diffusion modeling and discusses future plans to further couple diffusion with additional fields like electric and structural nonlinearities.

Uploaded by

zhoufl
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

Moisture Diffusion

Modeling in MAPDL and


Solder Joint Modeling in
WorkBench

Prepared by: Matt Sutton, PADT


With Input from Elana Antonova, ANSYS Inc
And Ming Yao Ding, ANSYS Inc

December, 2011
1
Outline

• Coupled-diffusion at 14.0
• Standard diffusion at 14.5
• Electromigration and other coupled-diffusion
applications at 15.0
• Solder Joint Modeling in Workbench
• Conclusions

December, 2011
2
Coupled-Field Enhancements at 14.0

• New at 14.0
– Diffusion physics and coupled-diffusion analyses
• Thermal-diffusion
• Structural-diffusion
• Structural-thermal-diffusion
• Motivation
– Simulation of moisture diffusion
– Sodium migration in aluminum reduction cells

December, 2011
3
Elements for Coupled-Diffusion Analyses

• KEYOPT(1) controls physics DoFs


– KEYOPT(1)=100001 is structural-diffusion (U+CONC)
PLANE223
2-D 8-node – KEYOPT(1)=100010 is thermal-diffusion (TEMP+CONC)
quadrilateral – KEYOPT(1)=100011 is structural-thermal-diffusion
(U+TEMP+CONC)
• Material properties required
– MP,DXX (DYY, DZZ) for diffusivity
SOLID226 – MP,CSAT for saturated concentration
3-D 20-node – MP,BETX (BETY, BETZ) for coefficients of diffusion expansion
brick
– MP,CREF for reference concentration
– All materials can be temperature-dependent
• Boundary conditions
SOLID227 – D,,CONC and IC,,CONC for concentration
3-D 10-node – F,,RATE for diffusion flow rate
tetrahedron
• Results
– Concentration gradient (CG)
– Diffusion flux (DF)
– Diffusion strain (EPDI)

December, 2011
4
Coupled Effects in Structural-Thermal-Diffusion Analyses (14.0)

Thermal strain    (T  Tref )


th

Structural Thermal
 Thermal expansion
NLGEOM

Diffusion strain
 di   (C  Cref ) Diffusion

 - moisture expansioncoefficient
Cref - " reference" concentration

December, 2011
5
Example Model From Galloway and Miles.

Geometry and ¼ Symmetry


ANSYS Model

1. ANSYS model consists of solid226 with thermal


and diffusion DOFS
2. Die is modeled with solid70s (thermal only)
3. Material properties are temp dependent

December, 2011
6
Results for 85C/85% RH for 168 Hours
0.35%

0.30%
Percent Weight Gain

0.25%

0.20%

0.15%

0.10%

0.05%

0.00%
0 24 48 72 96 120 144 168
Time (hrs)

Model is run as a transient, but since the temperature is


constant and the thermal time scale is much smaller than
the diffusion time scale, you can turn thermal time
integration off.
Post process to get percent weight gain as a function of
time.

December, 2011
7
“Popcorn” Modeling
• Use cohesive zone elements to model
delamination
• Use Solid226 to model thermal/structural/diffusion INTER204
interaction 3-D 16-node
interface
Thermalstrain    (T  Tref )
th

SOLID226
Structural Thermal
 Thermal expansion 3-D 20-node
NLGEOM Brick

SOLID227
3-D 10-node
Diffusion strain tetrahedron
 di   (C  Cref ) Diffusion
 - moisture expansioncoefficient
Cref - " reference" concentration

December, 2011
8
Diffusion Analysis at 14.5

• New at 14.5
– Three high-order elements for a standard diffusion analysis
• PLANE238 – 2D 8-node quadrilateral
• SOLID239 – 3D 20-node hexahedral
• SOLID240 – 3D 10-node tetrahedron
• Motivation
– Prior to 14.0, a temperature-concentration analogy was used
to model diffusion
• Valid only for homogeneous materials
– For inhomogeneous materials, a normalized concentration
approach is available with the new elements
• Unlike temperature, concentration is discontinuous across material
interfaces since it is limited by saturated concentration, which is
different for different materials. Normalized concentration =C/Csat is
continuous across material interfaces, so this is the DoF used in
moisture diffusion problems.

December, 2011
9
New Elements for Diffusion Analysis

• Degrees of freedom
PLANE238 – CONC – concentration or normalized
2-D 8-node concentration (if Csat specified)
quadrilateral
• Material properties (MP)
– DXX, DYY, DZZ, CSAT Will be supported by
the 22x elements
SOLID239 • Surface loads (SF)
3-D 20-node – Diffusion flux (DFLUX)
brick
• Body loads (BF)
– Diffusing substance generation rate (DGEN)
• Boundary conditions
SOLID240
3-D 10-node
– D,,CONC and IC,,CONC for concentration
tetrahedron – F,,RATE for diffusion flow rate
• Results
– Concentration gradient (CG)
– Diffusion flux (DF)
December, 2011
10
Coupled-Field Enhancements at 14.5/15.0
(Subject to Change)
• Current development
– Support structural material nonlinearities (plasticity,
viscoelasticity)
– Couple diffusion with electric and electrostatic fields
• Structural-thermo-electric-diffusion analysis
• Electrostatic-diffusion analysis
• Motivation
– Enhance moisture migration analysis
– Electromigration in solder joints

December, 2011
11
Driving Forces of Electromigration

• Atomic concentration
• Thermal gradient
• Electric field
• Stress gradient

 c
Mass conservation   J  0
t
  cQ * cZ *e c 
Atomic flux J   D c  2 T     
 kT kT kT 
Q* - heat of transport
Z * - effectivecharge
 - atomic volume

December, 2011
12
More Coupled Diffusion Analyses in 15.0
(Subject to Change)

Structural  Thermal expansion Thermal


Thermoelastic damping 
Plastic heat 

Joule heat
Diffusion Electric diffusion Electrical

December, 2011
13
Solder Joint
Modeling in
Workbench

December, 2011
14
Solder Creep Models

• Anand’s Viscoplasticity model


– Most popular material
model for solder
– Originally developed for
metal forming applications

R. Darveaux, “Effect of Simulation Methodology on Solder Joint Crack


Growth Correlation”, ECTC 2000

December, 2011
15
Solder Creep Models

• Combined time
hardening/Double power
Law
– Found to fit SnAg
solder test data well.

Anand model

Syed, A, “Accumulated Creep Strain and Energy Density Based


Thermal Fatigue Life Prediction Models for SnAgCu Solder Joints”
December, 2011
ECTC 2004 16
Results of interest
• Strain Energy Density

Wave 
 W  V
V
• Accumulated creep strain

 ave 
   V
acc

V
– Averaging over a small volume prevents singularities.

December, 2011
17
Cycles to Failure Calculation

• Following taken from Syed, Ahmer • Where


“Accumulated Creep Strain and Energy
Density Based Thermal Fatigue Life
Nf = Cycles to Failure
Prediction Models for SnAgCu Solder
Joints”, ECTC 2004 C’ is a constant dependent on material

 
C’=0.0153 and C”=0.0019 for the Anand
N f  C" Wave 
1 1
N f  C  acc model

• Following taken from R. Darveaux, “Effect of Simulation Methodology


on Solder Joint Crack Growth Correlation”, ECTC 2000
• Crack Initiation: N  K W K 2
0 1 avg
Here, K1 through K4 are material
• Crack Growth: da
 K 3Wavg
K4 parameters
dN a is the joint diameter at the
a
• Characteristic life: W  N 0  interface (‘final crack length’)
da W is the plastic work per cycle
dN

December, 2011
18
Workflow

• Geometry
– Split solder region for
volumetric averaging

– Design Modeler

December, 2011
19
Workflow

• Input material properties in engineering data

December, 2011
20
Workflow

• Setup Simulation
– Check Connections and contacts
– Create Named Selections
• To identify location for volume
averaging
– Define four load steps for thermal
load (3 cycles)

December, 2011
21
Workflow
cycletime=4200
*do,AR98,1,3
!--------------------------------------------------
! Read in the result at time=CycleTime*n (end of cycle)

• Post Processing ! Select solder part


! Ensure that Named Selection "Solder" exists
!--------------------------------------------------

– APDL Command for


set,,,,,cycletime*AR98
cmsel,s,Solder
!--------------------------------------------------
volumetric averaging ! Get accumulated plastic work (nl,plwk) and volume (volu)
!--------------------------------------------------
etable,erase
and calculations etable,vsetable,nl,plwk
etable,volu,volu
!--------------------------------------------------
! Multiply acc. plastic work by volume
!--------------------------------------------------
smult,pwtable,volu,vsetable
!--------------------------------------------------
! Sum all values and put in parameters Sx_NAMEy where
! NAME = type of summation
! y = cycle number (1-3)
!--------------------------------------------------
ssum
*get,s_volu%AR98%,ssum,,item,volu
*get,s_plwk%AR98%,ssum,,item,pwtable
!--------------------------------------------------
! Calculate volume-averaged acc. strain energy density (plastic work)
!--------------------------------------------------
s_wavg%AR98%=s_plwk%AR98%/s_volu%AR98%
*enddo
my_solder_wdiff_2_1=s_wavg2-s_wavg1
my_solder_wdiff_3_2=s_wavg3-s_wavg2

December, 2011
22
Conclusions
• Moisture Modeling is now supported at 14.0
– Coupled Diffusion-Structural-Thermal capabilities exist today.
– You can turn on and off which physics you are interested in with
keyopts on the elements
– You can control transient time integration for particular DOFs to
account for disparity in time scales between physics
• More advanced coupling is coming in future releases
– Coupling to support electromigration
– More advanced support for existing couplings
• Solder joint reliability studies can now be performed in WB
virtually natively
– Advanced material property input supported
– Transient workflow is supported.
– Post processing calculations are supplied with simple snippet

December, 2011
23

You might also like