DETERMINATION OF FRACTURE TOUGHNESS IN PBGA PACKAGES USING FINITE ELEMENT AND
EXPERIMENTAL METHODS
by
TAN CHEW PHENG
Thesis submitted in fulfillment of the
requirements
for the
degree
ofMaster of Science
U.S.M, Penang
November 2006
14,976 Words
,\'�t1"wkd!;CI1lCI1I'
ACKNOWLEDGEMENTS
I would like to express my
appreciation
and thanks to my supervisor Assoc. Prof. Dr. IshakHj.
Abdul Azid and myco-supervisor,
Assoc. Prof DrMani Maran for their
grateful support, encouragement
and invaluableguidance
in
helping
me to cope in everystage
of theproject
I also wish to thank my friends from the School of Mechanical
Engineering,
Universiti SainsMalaysia especially
Mr.Ong Kang
Eu and MrLee Boon
Leong
for theirhelp
in thisproject.
I also wish to thank Prof. K. N Seetharamu and Prof. M. V. V.Murthy
for theirguidance during
theearly
stage of thisproject.
I would like to thank every lab staff of the School of Mechanical
Engineering especially
Mr. Baharom B.Awang
and Mr. Md. Ashamuddin Hasim for their technicalsupport during
myexperiments
for this project I also would like to thank Mr. Md. Zandar Md. Saman from the School of Materials & MineraiResources
Engineering
inproviding
materials.Finally,
I wish to thank myfamily
members for their moral supportespecially
myparents.
Thanks also to those whohelped
me eitherdirectly
orindirectly
incompleting
myproject.
1,\1\ ('lIEWr/lENG McchFI1�.US1\(()(" � 11 -
Tableofcontents
TABLE ,,)F CONTENTS
Page
Acknowledgements
IITable of Contents III
List of Tables VIII
List of
Figures
XList of Plates xv
List of Abbreviations XVI
List of
Symbols
XIXList of
Equations
xxvAbstrak XXXI
Abstract XXXIII
CHAPTER 1 - INTRODUCTION 1-6
1.1
Background
of research 11.2 Problem statement 3
1.3
Objectives
41.4
Scope
of research 5CHAPTER 2 - LITERATURE REVIEW 7-21
2.1 Introduction 7
2.2 Review of related research on vapor pressure effect 7
2.3 Review of fracture
toughness analysis techniques
102.4 Review of related research on fracture
toughness
method 152.5 Review on
holographic interferometry techniques
18TAN cutw PHENGMcchEng,USM 0(,' - III -
Table ofcontents
2.6 Review of
holographic technique applied
for fracturetoughness
192.7
Summary
20CHAPTER 3 - ANALYSIS OF STRESS INTENSITY FACTOR IN 22-50 PLASTIC BALL GRID ARRAY
(PBGA)
3.1 Introduction 22
3.2 Theoretical
applications
of MCCI method 243.3 Methods of measurement 26
3.4
Methodology
293.5 Verification on Meei method 32-36
3.5.1 Thin
quad
flatpackage
finite elementmodeling
323.5.2
Sample
calculation on strain energy release rate 353.5.3 Results and discussions 36
3.6 Simulations and calculations 37-43
3.6.1
Temperature
distribution 373.6.2 Moisture distribution 39
3.6.3
Vapor
pressure determination 433.7 Results and discussions 44-49
3.7.1
Sample
calculation on strain energy release rate 45 3.7.2Sample
calculation on stressintensity
factor 453.7.3 Results and discussions 47
3.8
Summary
49CHAPTER 4- DETERMINATION OF FRACTURE TOUGHNESS 51-76 USING SIMULATION AND CONVENTIONAL TENSILE TEST
4.1 Introduction 51
4.2
Underlying
theories offracturetoughness
test 53-59TAN CIIEWPlIENG McchEng.USM 06' - IV -
Table()fcontents
4.2.1
Theory
of fracturetoughness
534.2.2 Fracture
toughness for single edge
notched tension 554.3
Sample specifications
604.4 Simulation evaluation 61-69
4.4.1
Ansys
simulation on thinplate
deformation 6'14.4.2 Parameter
study
624.4.3 Discussions 68
4.4.4
Epoxy
basedpolymer
thinplate
evaluation 694.5 Conventional tensile test evaluation 70-74
4.5.1 Procedures oftest 70
4.5.2 Results and discussions 70
4.6
Summary
75CHAPTER 5 - DETERMINATION OF FRACTURE TOUGHNESS 77 -118 USING HOLOGRAPHIC INTERFEROMETRY METHOD
5.1 Introduction 77
5.2
Underlying
theories of real-timeholographic interferometry
795.3
Methodology
82-895.3.1
Optical system
forholographic
measurement 825.3.2
Layout specifications
845.3.3
Expanding
beam withspatial
filter 855.3.4 Beam
intensity
ratio 865.3.5
Optical stability
test 875.3.6
Exposure
timeduring image recording
885.3.7
Hologram image processing
885.4
Holographic interferometry
calibration 89-102IAN CHEWPIlENGMcchEng.USM06' - V-
Tableoi"contents
5.4.1 Cantilever evaluation 89
5.4.2 Strain energy for linear deflection 94
5.4.3
Measuring fringe
order number 955.4.4
Perspective
distortion 975.4.S
Sample
calculations 985.4.6 Results and discussions 101
5.5 Uniform axial load evaluation
using holographic interferometry
103-1145.5.1 Mathematical
application
103S.S.2
Correcting
errors inholographic
fracturetoughness
test 1045.S.3
Holographic
fracturetoughness
test for aluminum 106 5.5.4Holographic
fracturetoughness
test for epoxy 1105.6 Results and discussions 114
S.7
Types
of errorduring holographic interferometry
115S.8
Summary
118CHAPTER 6 - CONCLUSIONS AND RECOMMENDATIONS 119-121
6.1 Conclusions 119
6.2 Recommendations and futurework 121
References 122-127
1 .
2.
Ansys
simulation program code for TQFPplastic
deformationAnsys
simulation program code for PBGAtemperature
distribution
Ansys
simulation program COC9 for PBGA wetness distributionAnsys
simulation program code for PBGAplastic
deformation128-213
128
135
Appendices
3.
4.
145
155
T /\NCIIEWPHENG MechEng.USM 06' - VI -
Tableofcontents
5.
Ansys
simulation program code for thinplate
deformation 164 6.Ansys
simulation on aluminumplate
in different crack ratio 1687. Simulation
analyses
on thinplate
fracturetoughness
1728. Cantilever deflection
using holographic interferometry
method 1769. Aluminum
plate
extensionusing holographic interferometry
192method
10.
Epoxy plate
extensionusing holographic interferometry
method 204
Publication List
I.A.
Azid,
C.P. Tan. B.L.Lee,
K.N. Seetharamu(2005).
Fracture MechanicAnalysis
forPackage
Delamination.Key Engineering Materials,
Vols.297-300.p.844-850.
TANCHEWPHENGMechEng.USM 06' - VU -
Listoftables
LIST OF TABLES
Page
Table 3.1 Dimensions of PBGA 23
Table 3.2 Material
properties
forthe main constituents of PBGA 23 Table 3.3 Shear modulus for interfacial delamination stress 24intensity
factorTable 3.4 Mathematical
applications
ofexperimental
methods 29 Table 3.5Analogy
between thermal and moisture 29Table 3.6 Dimensions ofTQFP 33
Table 3.7 Material
properties
for the main constituents ofTQFP 33 Table 3.8Comparison
MCGI method withPark,
Y.B. et als 36results for 1 MPa vapor pressure
Table 3.9 Thermal
properties
for PBGA 37Table 3.10 Moisture
properties
for PBGA(85°C/85%RH)
39Table 4.1 Material
properties
forAI, Cu,
MS and epoxy resins 51 Table 4.2 Thespecifications
and itsplane
condition 60Table 4.3 Critical load and
compliance
rate forAI,
Cu and MS 73Table 4.4 Fracture
toughness
used total extension forAI,
eu and 74MS
Table
5.1(a) Compositions
of chemical for solution A 88 Table5.1(b) Compositions
of chemical for solution B 88 Table5.1(c) Compositions
of chemical for bleach solution 88 Table 5.2 Values of deflection for6�m deflection, 6:r
delivered 99theoretically
Table 5.3 Coordinates of
point
a, b and c 100TAN CHEWPHENG McchEng.USM 06' - Vl11 -
l.istoftables
Table 5.4 Deflections obtained
experimentally
in casestudy
1 100 Table 5.5 Deflections obtainedexperimentally
in casestudy
2 101 Table 5.6 Extensionalong
AIplate,
2.45mm cracklength
loaded 107with 4N
using
HI methodTable 5.7 Material
properties
for D.E.R. 331 cured with Ancamide 110 260ATable 5.8 Extension
along
epoxyplate,
1mm cracklength
loaded 112with 1 N
using
HI methodTANCIIEWPHENG McchEng,USM 06' - IX-
l.rstoffigures
LIST OF FIGURES
Page
Figure
1.1 I R reflowsoldering profile
2Figure
1.2 Plasticcollapse
of popcorn failure 4Figure
2.1Cracking
mechanismduring
reflowsoldering
8Figure
2.2 Crackedbody subjected
to load F 10Figure
2.3 Load-extensiondiagram
11Figure
2.4 SENB for K determination 13Figure
2.5 CT for K determination 14Figure
2.6 Interiacial fracturetoug
hness test 14Figure
2.7 Schematicdiagram
of adhesion test 16Figure
2.8 Genericinterferometry
model 18Figure
3.1 Schematic of X PBGA 22Figure
3.2 Theprinciple
ofsuperposition
28Figure
3.3 Thespectrum
ofexperimental
methodsapplied
on 28fracture
problems
Figure
3.4 Cracktip
elements 31Figure
3.5 Cross sectional view ofTQFP 33Figure
3.6Typical plastic
deformation in TOFP 34Figure
3.7 Plastic deformation in TQFP(x1 00)
35Figure
3.8 Variation of SERR with delamination ratio for TQFP 36using
MCCI method at different vapor pressureloadings
Figure
3.9Temperature
distribution on PBGAduring
reflow 38Figure
3.10 Casestudy
ontemperature
distribution for LEO 39TANCHEWPHENG McchEng.USM 06' - X -
LI�lofligures
Figure
3.11 Transientwetness distribution in PBGA for 34 hours 40Figure
3.12 Transient wetness distribution in PBGA for 168 41 hoursFigure 3.13(a)
Transient wetness distribution in LED for 34 hours 42Figure 3.13(b)
Transient wetness distribution in LED for 168 hours 42Figure
3.14Typical plastic
deformation in PBGA 44Figure
3.15 Plastic deformation in PBGA(x1 000)
45Figure 3.16(a)
Variation of SERR with delamination ratio for PBGA 47Figure 3.16(b)
Variation of SIF with delamination ratio for PBGA 47Figure 3.17(a)
Variation of SERR with vapor pressure for PBGA 48Figure 3.17(b)
Variation of SIF with vapor pressure for PBGA 48Figure
4.1 3-Ddrawing
of thinplate specimen
52Figure
4.2 Conventional tensile test 52Figure
4.3 Fracturetoughness
with thickness for steel 54(rr, =2068.43MPa)
Figure
4.4 Linear elastic thinplate
55Figure
4.5 Plate subjected
with a load 55Figure
4.6 Freebody diagram
of thinplate
in leftportion
57Figure
4.7 AIplate
deformation with crack ratio 0.1(x100)
61Figure 4.8(a)
Variation of fracturetoughness
with cracklength
62ratio on total extension and crack
edge
extension forAI
Figure 4.8(b)
Variation of fracturetoughness
with totalplate length
63for AI
Figure 4.8(c)
Variation of fracturetoughness
withplate
thickness 63for AI
Figure 4.8(d)
Variation of fracturetoughness
withapplied
load for 64AI
T ANCHEW PHENG MechEng.USM06' - Xl -
l.ist oi"figures
Figure 4.9(a)
Variation offracturetoughness
with cracklength
64ratio on total extension and crack
edge
extension for euFigure 4.9(b)
Variation offracturetoughness
with totalplate length
65 for CuFigure 4.9(c)
Variation of fracturetoughness
withplate
thickness 65for Cu
Figure 4.9(d)
Variation of fracturetoughness
withapplied
load for 66eu
Figure
4.1O(a)
Variation of fracturetoughness
with cracklength
66ratio on total extension and crack
edge
extension for MSFigure 4.10(b)
Variation of fracturetoughness
with totalplate length
67for MS
Figure
4.1O(c)
Variation of fracturetoughness
withplate
thickness 67for MS
Figure
4.1O(d)
Variation offracturetoughness
withapplied
load for 68MS
Figure
4.11 Variation of fracturetoughness
with cracklength
69ratio for epoxy
Figure 4.12(a)
Total extension-load curve for AI 71Figure 4.12(b)
Variation ofcompliance
with cracklength
ratio for AI 71Figure 4.13(a)
Total extension-load curve for eu 72Figure 4.13(b)
Variation ofcompliance
with cracklength
ratio for eu 72Figure 4.14(a)
Total extension-load curve for MS 73Figure 4.14(b)
Variation ofcompliance
with cracklength
ratio for 73MS
Figure
5.1 Load-extension mechanism 78Figure
5.2 Path of incident and reflected rays atpoint
b 80Figure
5.3 3-D reflected ray atpoint
o 80Figure
5.4Optical system
for HI measurement 83TANClIEWI'll[NG McchEng.USM06' - XII
-
List offigures
Figure
5.5Optical layout
with itsspecifications
84Figure
5.6Optical system
forstability
test 87Figure
5.7 Cantilever deflected in anangle, ¢
90Figure
5.8 Illumination and observationangle
in x- zplane
91Figure
5.9 Cantilever rotate into anangle, /3
93Figure
5.10 Freebody diagram
of cantilever withapplied load,
F 94Figure
5.11 Lineprofile
offringes pattern
for cantilever deflection 96Figure
5.12Comparison
of unitpixel
and unit dimension in the 97analysis
ofperspective
distortionFigure
5.13Ansys
simulation on cantilever deflection(x200)
for 99Bum deflection, oZF
Figure 5.14(a) Comparison
ofholographic
result with deflection 101theory
for cantilever in0ZF =6�m
Figure 5.14(b) Comparison
ofholographic
result with deflection 102.
theory
for cantilever inau
=1Ourn
Figure 5.14(c) Comparison
ofholographic
result with deflection 102theory
for cantilever inau
=14J.lm
Figure 5.15(a)
Extension distributionalong plate using
HI method 105Figure 5.15(b)
Extension distributionalong plate using Ansys
105simulation
Figure
5.16Correcting
errors inholographic analysis
106Figure
5.17 Load-extension curve for AIplate
with cracklength,
106as=
4mmFigure
5.18 Lineprofile
offringes pattern
for AIplate,
2.45mm 107crack
length
loaded with 4NFigure
5.19 Extension distributionalong
AIplate
108Figure
5.20 Strain at center crack in differentloading
for AIplate
108Figure
5.21 Variation ofcompliance
with cracklength
ratio for AI 109plate
IANcur.w1'1I[,NG MeehEng.USM 06' - Xlll -
l.istoffigures
Figure
5.22 Load-extension curve for D.E.R. 331 with crack 111length,
as= 4.5mmFigure
5.23 Lineprofile
offringes pattern
for epoxyplate,
1mm 112crack
length
loaded with 1NFigure
5.24 Extension distributionalong
epoxyplate
112Figure
5.25 Strain at center crack in differentloading
for epoxy 113plate
Figure
5.26 Variation ofcompliance
with cracklength
ratio for 113epoxy
plate
Figure
5.27Safety
conditions for EMC(O.E.R. 331)
on PBGA 114TANCHEW PHENG McchEng.USM 06'
- XIV -
list ofplates
LIST OF PLATES
Page
Plate 5.1 Illustration of load frame 79
Plate 5.2 Real-time
interferometry set-up
for cantilever evaluation 83Plate 5.3 Cantilever beam 90
Plate
5.4(a)
Cantileverimage
96Plate
5.4(b)
Initialfringes pattern
96Plate
5.4(c)
Loadedfringes pattern,
casestudy
1 96Plate
5.4(d)
Loadedfringes pattern,
casestudy
2 96Plate 5.5
Graph
paper asrectangular grid pattern
on cantilever 97beam
during analysis
ofperspective
distortionPlate 5.6
Fringes pattern
for AIplate,
2.45mm cracklength
loaded 107with 4N
Plate 5.7
Fringes pattern
for epoxyplate,
1mm cracklength
111loaded with 1N
Plate
5.8(a) Hologram
withfringes
in its initial condition 117 PlateS.8(b) Hologram
withfringes
in its initial condition 117TANCHEW PHENGMechEng.USM06' - XV -
List ofabbreviations
LIST OF ABBREVIATIONS
Page
1-0 One-dimensional 8
2-D Two-dimensional
8,20,22,25-28
3-D Three-dimensional
20,26,28,52,78-92,121
Ag
Silver 23AI Aluminum
51,60-64,
70-77,104-109
Anhyd. Anhydrides
88ASTM American
Society
forTesting
and Materials 558S1 British Standards Institute 55
BT Bismaleimidetriazine
22,23,37
-40CCD
Charge-coupled
device83,95
CI Conventional
interferometry
19C-SAM C-mode
scanning
acousticmicroscopy
8CPU Central
processing
unit 83CTE Coefficient of thermal
expansion 15,17
Cu
Copper 17,24,51,60-66,70-75
O.E.R Dow epoxy resins
(trademark
ofthe Dow49,77,110-121
ChemicalCompany)
OHI
Digital holography interferometry 19,20
EMC
Epoxy molding compound
1-6,16,23,30�34,
77,114-121
ESPI Electronicspeckle pattern interferometry
19FE Finite element
3,9,25,30,78
TANCHEW PHENG McchEng,USM06' - XVI -
List ofabbrevintions
FEA Finite element
analysis 4,9121-23,29,34,35.40, 45,49,52,59,119,121
FEM Finite element method 18,25,26
GI
Grating (moire) interferometry
19HI
Holographic interferometry 4-7,18-21,68,74-89,
98-107,111-121IC
Integrated
circuit1,8,9,17
IR InfraRed 1
JEDEC Joint Electron Device
Engineering
Council 40LED
Lig
htemitting
diode38,39,41,42
LEFM Linear elastic fracture mechanics 15.25,69
MGGI Modified crack closure
integral 5,24,26,30- 32,36,49,119,121
NOT Non-destructive test 18
PBGA Plastic ball
grid
array2-8,22,23,36-49.
77,114-120
PCB Printed circuit board
22,23,37
-39PQFP Plastic
quad
flatpackage
53QFN Quad flat non-lead 9
RH Relative
humidity 5,22,40-43
SERR Strain energy release rate
4,15-17,22-27,30, 35,36,45-49,58
Si Silicon
23,33
SiC Silicon carbide 49
SIF Stress
intensity
factor4-6,13-17,22-31,
46-53,59,114,118-121
SMD Surface mount device 7
SMT Surface mount
technology
1,15TANCHEW PH ENG McchEng,USM 06' - XVll -
Listof abbreviations
SOJ
TQFP
VPR
Small outline J-Ieaded
Thin
quad
flatpackage Vapor phase
reflowTANCHEW PHENG MechEng,USM 06'
40,43 5,15,16,32-36,49
1,9
- XVlll -
List of syrnbo!s
LIST OF SYMBOLS
Page
t;
Reflowsoldering temperature (OC)
2Ix
Soldering
time(s)
2as Control
sample's
cracklength (m) 6,70-73,106-113
t Thickness
(m) 10-15,26-35,45,52-69
78,95,98,109,114
F
Applied
load(N) 10-14,31,35,45,52-73,
94-99,105-113
8 Crack
edge
extension(m) 10-12,55-62,108,112
a Crack
length (m) 10-14,25,26,31-36,
45-73,78,109-114
W
Sample
width(m) 10-15,52-73,78,
95,98,109-114
G Strain energy release rate(J/m2) 11,12,25,26
U Stored strain energy
(J) 11,12,56,57
C
Compliance,
theslope
of extension-load12,71-73,108-114 (mIN)
Ke
Critical stressintensity
factor(MPa
m1/2) 12-17,49-54,59- 69,74,1
09,114Ge
Critical strain energy release rate(J/m2) 12,15,58,59
E
Young's
modulus(GPa) 12,15,23,26,32,33,46,
51-73,94,98,109-114
Fe
Criticalloading (N) 12-15,58,59,
70-74,106-114
L
Length
distance(m) 13-15,33,52-69,90-113
S Distance between two
pin support (m)
13K Stress
intensity
factor(MPa
m1/2) 13,14,28
TANCHEW PHENGMcchEng,USM 06' - X1X -
Li,1 ofsymbols
Moment inertia of
multilayer (dimensionless)
15E" Effective
Young's
modulus(GPa) 15,32,46
Jr Pi
(=3.142) 15,25,31,32,46
f2 Bimaterial constant
(dimensionless) 15,31.32,46
J Crack
driving force,
Jintegral (J/m2) 15,16,27,32.49,53 KJI
Mode II stressintensity
factor(MPa
m1/2) 16,26
KlIC
Critical mode II stressintensity
factor(MPa
16m
1/2)
v Poisson's ratio
(dimensionless) 23,32,33,45, 46,51,62-73
ji Shear modulus
(GPa) 24,31,32,45,46
ljI Work ofthe external tractions per unit 25 thickness
(Jim)
Strain energy of the
body
per unit thickness 25(jim)
Normal stress
(N/m2) 25,29,56
r Shear stress
(N/m2) 25,29
v
Opening displacement
on crack surfaces(m) 25,26,29,31,35,45
u
Sliding displacement
on crack surfaces(m) 25,26,29 GI
Mode I strain energy release rate(J/m2) 26,31-36,45-48 G"
Mode II strain energy release rate(J/m2)
26KI
Mode I stressintensity
factor(MPa
m1/2) 26,32,46-49,114
TJ
Multiplier depending
Poisson's ratio u and26,46,59 geometry (dimensionless)
77=
1/(I-u2)
forplane strain;
17;:: l forplane
stress
�.
Normal force(N)
26Fe
Shear force(N)
26T AN CHEWPHENGMcchEng,USM 06' - XX-
Li51ofsymbols
o
Strain energydensity (J/m3)
27T Traction vector
(N/m2)
27u
Displacement
vector(m)
27ds Element of arc
length
around the contour r 27(m)
r Contour in crack
driving
force 27(dimensionless)
N
Fringe
order number(dimensionless) 29,79-81,90-96,
100-107,112r. Optical sensitivity
constant(N/m.lines)
29fi Fringe sensitivity (miIines)
29B Additional non-strain-related motion
(lines)
29la
Photoelastic constant for material(N'rn.lines)
29T
Temperature (OC)
29w Wetness fraction
(dimensionless) 29,40
p Material
density (kg1m3) 29,30,37
k Thermal
conductivity (W/mOC) 29,30,37
D Moisture
diffusivity (m2/s) 29,30,39
C\O{ Saturated concentration
(kg/rn3) 29,30,39
Cp
Specific
heat(J/kgoC) 29,30,37
pr
Vapor
pressure in the delaminationduring 30,35,43-48
reflow(Pa)
�(�)
Saturation water vapor pressure30,38,43 corresponding
to reflowsoldering
temperature (Pa)
RHIII Relative
humidity
at moistureconditioning 30,43 (dimensionless)
lV"Jr
Wetness in the delaminationduring
reflow30,40,43 (dimensionless)
TAN CHEWPHENG McchEng.USM 06' - XXi -
Listofsymbols
K
Multiplier depending
Poisson's ratio v andgeometry (dimension less)
K=
(3
-4u)
forplane strain;
K=(3-u)/(l+v)
for
plane
stressf
Length
of diepad (m)
e
Length
ofchip (m)
ar Thermal
diffusivity (m2/s)
h Convective heat transfer coefficient
(W/m20C)
Wo Initial wetness fraction
(dimensionless)
ao Initial delamination
length (m)
ae Allowed
length (m)
K/c
Plane strain fracturetoughness (MPa
m1/2)
(Jrs Tensile
strength (MPa)
(Jy Yield
strength (MPa)
5' Total extension
(m)
5e
Critical crackedge (m)
c Linear strain
(dimensionless)
M External moment
(Nm)
/ Moment inertia
(rn")
6:;.-
Deflection at load F(m)
q Distributed load
(N/m)
F;
Virtual load(N) MI
Virtual moment(Nm) 58
Deflection of beam(m)
5{)
Crackedge en.largement by plate
deformation
(m)
TAN CHEWPHENGMechEng,USM 06' - xxii -
31,32,45,46
33.35,36 33,45-48,114
37
37
40
49
49
49,51-55,62,66,114
51,59-69,11052-56,62,71-73, 78,105-112
55
56,106-113
56,5756-59,94-98 57,58,94-102
57,58 57,58 57,58,94
58
58
LISIofsymbols
A Area atthe center of thin
plate (m2)
59K/c
Meanplane
strain fracturetoughness (MPa
60m
1/2)
ar Mean
yield strength (MPa)
60<:
Maximumplane
strain fracturetoughness 62,64,69 (MPa
m1/2)
KIC,,"n Minimum
plane
strain fracturetoughness 62,64,69 (MPa
m1/2)
p Pressure
loading (MPa)
63,65,67r;
Tensileloading,
70% less than critical 70-74loading (N)
Laser
wavelength (nm) 79-81,90-93.99.103
Illumination vector
(dimensionless) 79-82,90-94
o Observation vector
(dimensionless)
79-82,90-94d
Displacement
vector(m) 79-81,90-94,99-112
¢>
Deflectionangle (0) 80,90,94
r
Angle
between vectord and x- zplane (0)
80PLD Path
length
difference(m)
84Lb
()y Pathlength
ofobject
beam(m)
84LIe! Path
length
of reference beam(m)
84BR Beam
intensity
ratio(dimensionless)
86I'd
Reference beamintensity (Lux)
86Ir/hj Object
beamintensity (Lux)
86eo
Observationangle (0) 91,92
o,
Illuminationangle (0) 91,92,100
e
Rays angle (0) 92,93,103
TANruewPHENGMcchEng.USM06' - XXlH -
Listofsymbols
fJ
Rotationangle (0) 93,103
!3,mg Image turning angle (0)
97,98x' Cantilever's coordinate
along
x' axes(m) 97,98
X
Image pixel along
X axes(pixel) 97,98
y
Image pixel along
y axes(pixel)
97,98r
Scaling
factoralong
X axes(m/pixel)
98� Correcting
scale forholographic analysis 105,107,112 (dimensionless)
8u
Initial crackedge (m)
108TANCHEWPHENG McchEng.USM06' - XXIV -
l.istofequations
Equation
2.1Equation
2.2Equation
2.3Equation
2.4Equation
2.5Equation
2.6Equation
2.7Equation
2.8Equation
2.9Equation
2.10Equation
2.11LIST OF EQUATIONS
Gtda=OFF'
=OFE+EFF'E'-OF'E'
G-JUI
taa FV=-F.o1 2
c=!_
F
G=
[2
(aci
2t Da
)
Ke =�GcE'
=
F;
(8C)EI
2t aa
K=
3FS1/-:;- [1.93 (.::)V2
-3.07(!:)l!Z
+14.53(.::)51
-zs.u:
.1�':
+25.8 "-;
l
21 W .. It' JIl JI" II. ii J
=_6�_[1.93(�)V2
tW,,2 IV-3.o7'l.::)Ji: +14.53(")��
JI' \11 -25.11,II",:�:+25.S".H":l
�F
[ ()V2 (
\)12( .5� .
'= ":
�
=- 11.58.E..
-18.42.E..)
+87,181�, -150.66" +JS·U·JtWV2 IV w ,,,"': II H_
K -
_fL_
•f(!!_)
c
-
tWl/2
W( ) [
1'2 J"�,' "
, "-
�
f E_ =
11.58(�).' -IU2(!:)
-
+87.18:( -d
-
-150.66 ,I
•
+15H ,1
-
i
W W II' \WJ ,II JI J
Ke
=t�J' ./(;)
f( �1_)=[177.6(!:.))lI2 _1113(!:.}l'� +3934.2'::"';'=
-6102 .1 ': ....)S33,� ,0:-JV JI' IV \11 II "_
TANCHEWPHENG McchEng.USM 06' - XXV-
Page
11
11
11
12
12
12
12
12
13
13
14
LISI ofequations
Equation
2.12Equation
2.13Equation
3.1Equation
3.2Equation
3.3Equation
3.4Equation
3.5aEquation
3.5bEquation
3.6Equation
3.7Equation
3.8Equation
3.9Equation
3.10Equation
3.11G
=_1
(g(FcLJ2J
__l(_!3_(FeLJ2J
e 2£II f3 W 2£' " 3 W
I 2 l)
15
Ke
=JGcE· cosh(nn)
15G=dlf_dc!>
da da
25
G = lim
_1_
r
(7(
Sa- r,O)
v(r,
Jr)
dr&l-O2Lla
+lim
_1_
r
r(
Doa- r,O)
u(r,
K)
dr60-02�a
25
K2
K 2 G=GI +G11 =�I• +_11•E E
26
26
GI =lim�l-Fy
- 2D.a1 .s, 2626
(
. auJ
J =
Sr Udy
- Taxds27
p,. = PK"
(T ).
RHm • Wd.Tr30
31
31
32
K)
=JGIE· cosh(nn)
32TANCHEWPHENG McchEng.USM06' - XXVI -
List ofequations
Equation
4.1aEquation
4.1 bEquation
4.2aEquation
4.2bEquation
4.3Equation
4.4Equation
4.5Equation
4.6Equation
4.7Equation
4.8Equation
4.9Equation
4.10Equation
4.11Equation
4.12Equation
4.1354
54
F
CJ=::---
(W-a)t
56
8'
E=-
L
56
(J=E'E 56
U=_l_
2EIfM2dx
56
g
=8U
:F 8F
=_1
ElfM(8M'Lv
8Fr
57
F q=w
57
57
BMx_:r
--=X
8F;
57
l
( qx' J
8_F
• i =- F .x+M. +-(x')dx
El
I I
2
58
�
8 == qa :Fj 8EI
58
8 ;::;
Fa4
:fj 8WE]
58
8B =
2(0
:Fj)
= 4WEIFa4
58
58
TANCHEW PHENGMechEns.USM06' - XXVll
-
Listofequations
Equation
4.14p2a3
58Ge-- c 2tWEI
Equation
4.151 =
t(L/2)3
5912
Equation
4.16 Fe =Aars
59=(W-a)tars
Equation
4.17a3 (W _a)2
f 2 59Ge
= (J'rs2WEI
Equation
4.18Ke =JGeE'
59=W'[(; r -(; rl(;� r
u�Equation
5.1( o-i)
. d=N•A 80Equation
5.2[::=}=[:�}A
81
o) -I
N) Equation
5.3a0=0 i+o j+o.k=
(Xc)i+(Yc)j+(z,)k
81x y •
J(
XC/ +(Yc )2 +(
Z,)�
Equation
S.3b •_
'. '. .
_
(
-Xa)i+(
-Ya)j+{ -Za)k
81l-lxl+1yJ+lzk-
2 2 .,�(Xa) +(Yo) +(zoY
Equation
5.3c d=d)+dyj+d;k
81Equation
5.4[
Ox,-ix 0)'1 -iy
O"-
i,
f' J [N, J
81 0x2 -
�x 0y2
-
ly
0,,= i, d,
=
�' (A J
0x3 - Ix
0y]
-
iy
o., l:d:
3Equation
5.5(
Ox,-
�x
0,,= i: J(d}( N, }AJ
810x2 - Ix 0:2 lz
d, Nz
Equation
5.6a .(xc)i+(Zc)k
820=0xl+O.k- =
J(
XC)2 +( Zc)2
TANCIIEWPHENGMechEng.USM 06' - XXVlll -
List ofequations
Equation
5.6bEquation
5.7Equation
5.8Equation
5.9Equation
5.10Equation
5.11Equation
5.12Equation
5.13Equation
5.14Equation
5.15Equation
5.16Equation
5.1782
PLD =
Lob}
-Lrej
=0
84
86
d;:
_Lsin¢
d( L(l-cos¢)
90
(O; -iJ(d;:)=(N){A)
o -i :;::
(zc)
_(-za)
z z
J(Xc)2 +(zc)2 J(Xa)2 +(za)�
=cos
80
-(
- cos� )
91
91
91
d = NA
z
1+ cos
�
91
92
(sinO. +sinB,
cose;+COSll.)(�:J =(N)(A)
9292
dz = NA 2cos8
92
Equation
5.18a d. =dx
cosf3
+d;
sinfJ
93Equation
5.18bd..
=-dx
sinf3
+dz
cosf3
93Equation
5.19d;
=0 93Equation
5.20 d, =-dz tanfJ
93TAN CHEW PHENGMcchEng.USM06' - XXIX -
LISI ofequations
Equation
5.21 d. 93d,. ;:::_-_
cos
f3
Equation
5,22 NA 93d.' ;:::
2cos()cos
p
Equation
5.23 Mx-x =F.x+MI 94Equation
5.248Mx_x
94--=x
of
Equation
5.250,
F =;/ f (
F•x+M;) (
x)dx
94Equation
5.26F(L -0.015)3
94b:F
=3£1
Equation
5.27Wt3
951=- 12
Equation
5.28a x'cosPimg
98r=
X
Equation
5.28b-f)
98Amg
=tan XEquation
5.29d=,
=0 103Equation
5.30d;:::�
x 103tan
f3
Equation
5.31d,=�
r 103sin
f3
Equation
5.32 d = NA 103x
'
2cos()sin
fJ Equation
5.33d,= N 104
x
0.7644x
106 Equation
5.34C= ae
(8 )
108 aL o
TANCHEWPHENGMechEng,USM 06' - XXX-
\ 1"11,'�
PENGUKURAN KEKUATAN PATAH DALAM PAKEJ PBGA DENGAN MENGGUNAKAN KAEDAH UNSUR TERHINGGA DAN EKSPERIMEN
ABSTRAK
Pakej
elektronikkhususnya jenis 'plastic
ballgrid array' (PBGA) mempunyai kegagalan
retakan dalaman yang unik yang berlaku disebabkan proses memateri.Kewujudan kelembapan setempat
yang meresap ke dalampakej menghasilkan
tekanan wap. Semasa proses memateriJenis
aliran(215°C), kelembapan
yang meresap akan mengewap dan kemudianmengakibatkan
retakan'popcorn'.
Olehitu,
kaedahberangka diperlukan
untukmemperihalkan rintangan patah
dalam bentuk faktor keamatantegasan (SI F) geometri. Kemudian, kajian
ke atas kekuatanpatah
bahanepoksi
sebatianacuan
(EMC)
menentukan keadaansempadan
untuk retakanpakej,
Dalam
kajian ini, perisian Ansys
telahdigunakan
untuk melakukansimulasi taburan
kelembapan
dan terma demi menentukan tekanan wappada bahagian
retak.Daya
tekanan wap ini ditindak ke atas PBGA untukmemperolehi
SIFgeometri dengan
penggunaan kaedahpenutupan
retak terubahsuai(MCCI). Didapati
bahawa kenaikan tekanan wap danpenambahan
saiz
lekangan menghasilkan
SIFgeometri
yangtinggi. Perhubungan daya- pemanjangan
celah retakan yangdiperolehi
menerusi kaedahholoqrafik
interferometri
(HI)
dandaya
kritikal D.E.R. 331sebagai
EMC telahdigunakan
untuk menganggar kekuatan
patah.
Kekuatanpatah
D.E.R 331 ialah 0.488MPaAkhirnya,
kekuatanpatah
D.E.R. 331 ini telahdibandingkan dengan
SIFgeometri
untuk PBGA. Grafmenunjukkan pelbagai
keadaansempadan
1·\ '\('IIIW1'111.N(j Mcch1·:Il�.LJS,\.1 0(,' - XXXI -
selamat
pakej
telahdiperolehi. Daripada graf ini,
bolehdisimpulkan
bahawaPBGA masih selamat semasa proses
memateri,
iaitu tekanan wap 1.21MPa dan nisbahlekangan kurang daripada
0.6.Kegagalan pakej
elektronik inidapat
dikawal
dengan
pengurangan kesan tekanan wap danlekangan
antarapermukaan chip
dan EMC.- XXXII -
DETERMINATION OF FRACTURE TOUGHNESS IN PBGA PACKAGES USING FINITE ELEMENT AND EXPERIMENTAL METHODS
ABSTRACT
Electronic
packages especially
theplastic
ballgrid
array(PBGA)
has itsunique
crack failurearising
from thesoldering
process. The existence of moisture absorbed inpackage
from ambient condition causes vapor pressure.During
reflowsoldering
processes(215°C),
the moisture absorbedvaporizes
and
eventually
causes popcorncracking. So,
numerical method is used tocharacterize the fracture resistance in terms of
geometric
stressintensity
factor(SIF). Then,
theinvestigation
of fracturetoughness
on epoxymolding compound (EMC)
materialprovided boundary
conditions ofpackage cracking.
In this
research, Ansys
software was used to simulate moisture diffusion and thermal transfer to determine the vapor pressure at the crackregion.
Thevapor pressure
loading
wasapplied
on PBGA to obtaingeometric
SIFusing
modified crack closure
integral (MCCI)
method. It was found that the increased in vapor pressure andhigher
delamination size createdhigh geometric
SIF.Load-crack
edge
extension relation obtained fromholographic interferometry (HI)
method and critical load for D.E.R. 331 as EMC material were used to evaluate fracturetoughness.
The fracturetoughness
for D.E.R 331 was foundto be 0.488MPa m
1/2.
Eventually,
the fracturetoughness
for D.E.R. 331 wascompared
withPBGA's
geometric
SIF. Agraph
withpackage safety
conditions was obtained.From the
graph,
it can be concluded that the PBGA istypically
to be savedduring soldering
process, where vapor pressure is 1.21MPa and delarnmatronI \, {III \\ PIli N{I Mcch I II!,:. llSM (lCl' - XXXlIl -
Abstract
ratio is less than 0.6. This EP's failure can be controlled
through minimizing
theeffect
of
vapor pressureand reducin_g die/EMC interface
delamination.TAN ('I mw I'III�NQM.:�hEna.USM 06' - XXXIV-
ChapterI
CHAPTER 1
INTRODUCTION
1.1
Background
of researchElectronic
packaging (EP) designing
andmanufacturing
tend to becomea
challenging global industry.
Thechallenges emphasize
onmaking
the EPlighter, higher packaging density (miniaturization)
with thecomplexity
infunctions and increased electrical
performance
while the cost remains at lowlevel.
However,
thesechallenges
ofpackaging
are hard to achieve andbring
difficulties in
processing, handling
andunderstanding
of smallercomponents particularly
with the use of thinner dies. The matterarising
now is how toproduce
ahigh reliability
EP.Higher density
and morecompact developed systems
areturning
themanufacturers to surface mount
technology (SMT)
which needs full concern of failure mechanism insideplastic package
which is known as popcorncracking.
The
package
ishighly
sensitive to moisture andhumidity
conditions. Itsapplications mostly
include consumerelectronics,
householdappliances, computing, automotive, telecommunications, flight control, robotics, military equipments
and astronautics.Plastic failure
usually happens
when the moisture diffused in thepackage during storage
becomesvaporized
and exerts a pressure known as vapor pressure that could result in the popcorn effect.Wong,
E.H. et al.(1998)
found that vapor pressure is
responsible
for the eventual popcorncracking
ofplastic integrated
circuit(Ie) packages during soldering
process.Tay,
A.A.D. et al.(1994)
mentioned thatduring soldering by
vaporphase
reflow(VPR)
or infrared(IR)
reflow(Figure 1.1)
or wavesoldering techniques.
TANCHEW PHENG MechEng.USM 06' - l -
Chapter I
the EP is heated up to solder
melting temperature
in theregion
of215°C
andexposed
to humid environment.High
thermal stresses were induced within the EP as the result of enhanced thermal mismatch between dissimilar materials.Meanwhile,
theevaporation
of moisture causedhigh
stresstypically
betweenmolding compound (MC)
andchip
interfaces. These stressesmainly
contributed to thephenomenon
ofcracking
inside the EPespecially
theplastic
ball
grid
array(PBGA) during
the reflowsoldering
process.250--�---/---�-,---�
o T:
nl.l\.=24S0C±SOC�
� 200 Tr=183°C 's=70s
Q)
- - - - - - - - - - - - - - - - - -
'-
:J Rampup
ro"- ISO Ramp down
2-30C/S/
el)
c.. 2�3°C/s
E
2O)c:: 100
�, /
'i::el)
/
Dry-out Reflow
-o i Pre-heat
C<n 50
�
) Ambientcondition Cooling
o 2SoC
� Cl)
O
a::::
O 60 120 180 240 300
�oldering time, ts (s)
Figure
1.1: IR reflowsoldering profile (OSRAM catalog, 2006, p.97)
From thesecircumstances,
it isimportant
tostudy
the micro mechanics of interface delam ination and popcorncracking.
Park,
Y. B. et al.(1997)
notedthat the popcorn
cracking phenomenon
inside the surface mountedpackages
was known
by assuming
an inherentedge
crack at diepad/epoxy molding compound (EMC)
interface andsubsequently
interface delamination under thermal arid vapor pressureloadings. According
to<