 

 

 

 

.m

 

aim“...

 

 

.3.
1....va ..
u. .zuw.

a 199-.»
92:17").-
I: J. an
rm...

 

Lﬂ
.

hunt? UT

 

 

L M5. , -.

AWN? . rm.aﬁu.n¥m1.&

R: «» ¢ .

@va *
5:4

ﬁ.‘

 

, km...
550"”: 1.

5415:11.
. .21 .l)

 

V4 is nil.
. 0331'... . Pt.

I (II, I).

5.3,; @éés.

. x) .p:

$§§sﬁq ﬂﬁ. ..

 

.

32.}, lllllllllllllllllllllllllllllllllllllllllllllllllllll

'l ‘
( _ 31293 015610904

LIBRARY

Mlchlgan State
I University

 

 

 

This is to certify that the

thesis entitled

EFFECTS OF AGING ON FRACTURE BEHAVIOR IN A1 MATRIX
PARTICULATE REINFORCED METAL MATRIX COMPOSITES

presented by

Jong-kook Park

has been accepted towards fulfillment
of the requirements for

M.S. Materials Science

degree in

GM (F (ﬂu/93

Major professor

 

Date KQW {0" (99C

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

———- ____~_ﬁ___- __—

PLACE It RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

L__':L_J
__

___-
T—IF—l—T

MSU ioAn Afﬂrmdivo AotionlEquol Opportunity institution

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EFFECTS OF AGING ON FRACTURE BEHAVIOR IN A] MATRIX
PARTICULATE REINFORCED METAL MATRIX COMPOSITES

By
J ong-kook Park

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Materials Science and Mechanics

1996

ABSTRACT

EFFECTS OF AGING ON FRACTURE BEHAVIOR IN Al MATRIX
PARTICULATE REINFORCED METAL MATRIX COMPOSITES

By
J ong-kook Park

Crack propagation behavior is investigated in relation to various interface
microstructures evolved during solution heat treatment and artiﬁcial aging in 6061
aluminum based A1203 and SiC particulate reinforced metal matrix composites
(PRMMCs). Materials investigated were extruded bars of A1203/6O61Al, and
SiC/6061A] PRMMCs produced by a stir-cast melt process. Interface microstructure
characterization and crack proﬁle analyses were carried out using optical microscope with
digital imaging capability. Scanning and transmission electron microscope were used to
analyze microstructural details, such as interfacial composition, chemistry and fracture
surface morphology. Examinations of SiCp/6061Al composite interface, subjected to a
moisture (H20) environment, was also conducted using environmental scanning electron
microscope (ESEM) to ascertain the hydrophilic nature of the interface. Mechanical
testing was conducted on double notched, 4-point-bend and short bar fracture toughness
specimens in the under aged, peak aged, and over aged conditions. Interfacial
microstructure that formed during solution heat treatment and artiﬁcial aging played a
signiﬁcant role on observed microfracture mechanisms in PRMMCs. Aging conditions,
reinforcement type, and microstructural formation all inﬂuenced the crack growth
processes, such as crack initiation, microcrack formation ahead of the crack tip, and

microcracks/main crack link up mechanisms.

To my lovely wife and family

Acknowledgments

I would like to thank Prof. J. P. Lucas for his guidance, advice and ﬁnancial support
throughout the duration of this research. I would also like to extend my sincere gratitude
to my thesis committee members: Prof. K. Mukherjee, Prof. K. N. Subramanian and

Prof. T. L. Bieler.

iv

TABLE OF CONTENTS

TABLE OF CONTENTS ................................................................................................... iv
LIST OF TABLES ............................................................................................................. vi
LIST OF FIGURES .......................................................................................................... viii
CHAPTER I. INTRODUCTION
1.1 Fabrication of Particulate Reinforced Metal Matrix Composites ................................. 1
1.2 Aging Kinetics in Aluminum Alloys ........................................................................... 10
1.3 Aging Kinetics in MMCs ............................................................................................ 13
1.4 Interface in MMCs ...................................................................................................... 18
1.4.1 SiC / Aluminum interface ................................................................................................ 19
1.4.2 A1203 / Aluminum interface ............................................................................................. 20
1.5 Fracture of PRMMCs .................................................................................................. 21

CHAPTER II. MATERIALS AND EXPERIMENTAL PROCEDURES

2.1 Materials ...................................................................................................................... 30
2.2 Aging ........................................................................................................................... 31
2.3 Metallographic Preparation ......................................................................................... 31
2.4 Electropolishing and TEM Sample Preparation .......................................................... 32
2.5 Fracture Toughness Testing ........................................................................................ 33
2.6 Bend Testing ............................................................................................................... 34

CHAPTER III. RESULTS AND DISCUSSION

3.1 Microstructure Characterization .................................................................................. 38
3.2 Interface Characterization ........................................................................................... 42
3.2.1 A1203/6061Al interface .................................................................................................... 42
3.2.2 SiC/6061A] interface ........................................................................................................ 46
3.2.3 Observation of moisture effect on interfaces ................................................................... 48
3.3 Aging Kinetics ............................................................................................................. 53

3.4 Fracture Behavior ........................................................................................................ 57

3.4.1 Crack initiation ................................................................................................................. 57
3.4.2 Crack propagation ............................................................................................................ 62
3.5 Fracture Toughness ..................................................................................................... 72
CHAPTER IV. CONCLUSIONS
LIST OF REFERENCES .................................................................................................. 78
APPENDIX ....................................................................................................................... 83

vi

LIST OF TABLES

Table 1. Properties of reinforcement (particulate A1203), a matrix alloy (6061Al), and a
composite (15 "/o A1203/ 6061Al) [10,12~14]. ................................................... 9

Table 2. Nominal chemical composition (weight percent) 6061 aluminum alloys [17]. . 12
Table 3. Observation of accelerated aging in various MMC systems. ............................. 15
Table 4. Properties of matrix alloy (6061Al), and PRMMCs [10,12]. ............................. 30
Table 5. Descriptions of metallographic preparation for optical and SEM samples. ....... 32

Table 6. Average microhardness values for several PRMMCs during aging. Standard
deviation values are in parentheses. (Unit: Hv) ................................................ 55

Table 7. Data for the effect of aging on fracture mechanisms along the crack path for the
PRMMCs ............................................................................................................ 67

Table 8. Fracture toughness(K,,) values for the various PRMMCs as a function of aging
conditions. (Unit: MPGJE, unit of values in parentheses: Ksiw/i; ) ................ 75

LIST OF FIGURES

Figure 1. Schematic diagram for fabrication of particulate metal matrix composite by

vortex method [6]. ............................................................................................. 3
Figure 2. Schematic diagram for fabrication of particulate metal matrix composite by
compocasting [5]. .............................................................................................. 5
Figure 3. Fabrication flow chart of PRMMCs by powder metallurgy (PM) [11]. ............. 6

Figure 4. Schematic diagram for fabrication of particulate metal matrix composite by
spray deposition [5]. .......................................................................................... 8

Figure 5. Illustrative micrograph for crack propagation mechanism at the crack tip in
particulate reinforced metal matrix composite (15V/o (A1203)p/2014Al
composite), a) reinforcement/matrix interface debonding, b) through particle
fracture, c) crack propagates through reinforcement/matrix interface of
fractured particle, (1) matrix failure, e) fractured particle around the crack, f)
crack tip, g) reinforcement/matrix interface debonding ahead of the crack tip,
h) fractured particle ahead of the crack tip. ..................................................... 29

Figure 6. Optical micrographs showing various specimens for mechanical testing and
typical location of hardness indents, a) hardness indentation placed well
between the reinforcement particles, b) half-inch width (B=O.5) short bar
fracture toughness specimen, c) half-sectioned and mounted fracture
toughness specimen for crack propagation observation, (1) double-notched 4-
point bend bar specimen for crack initiation observation. .............................. 35

Figure 7. A schematic diagram showing the half-inch width short bar fracture toughness
specimen [52]. ................................................................................................. 36

Figure 8. A schematic illustration for the dimension of a double notched 4—point bend
test specimen. Linear dimensions given are in inches ..................................... 37

Figure 9. Three dimensional optical micrographs of PRMMCs showing partial anisotropy
and reinforcement clustering along the extruded direction, a)
10"/o(A1203)p/6O61Al, b) 20"/o (A1203)p/6O61A1, c) 10"/o SiCp I6061Al, d)
20"lo SiCp /6061A1 .......................................................................................... 39

Figure 10. Distribution in particle size and average particle size (APS) for the composite

materials, a) IOV/o (A1203)p/6061A1, b) 20"/o(Aleg)p/6061Al, c) 10"lo
SiCp/6061Al, d) ZOV/o SiCp/6061Al. ............................................................... 40

viii

Figure 11. Optical micrographs of PRMMCs, a) 10"/o (A1203)p/6061Al, b)

20V/o(Al203)p/6061Al, c) 1070 SiCp/6061Al, d) 2070 SiCp/6061Al. ............ 41

Figure 12. SEM micrographs showing matrix/reinforcement interface region of the

Figure 13.

Figure 14.

(A1203)p/6061Al composites, a) irregular interface structure between (A1203)p
and the matrix alloy indicating a chemical reaction by partial consumption of
(A1203)p, b) electrochemically extracted reinforcement particles showing
MgAl204 crystals formed on the A1203 particulates, c) an alumina particulate
covered nearly 100% with spine] (MgAl2O4) crystals (~ 1 mm), (1) individual
octahedral-shape MgAl2O4 crystals ................................................................. 43

TEM micrographs showing the interface of the (A1203)p /6061Al composites,
a) bright ﬁeld image of interface region showing formation of MgAl204 at the
Al203/6O61Al interface with irregular interface between the MgAl204 and the
A1203 particle, b) convergent beam electron diffraction pattern on the
MgAl204 spine] (FCC spinel structure), c) convergent beam electron
diffraction pattern on the A1203 particle (or-A1203 corundum structure). ....... 45

SEM micrographs showing matrix/reinforcement interface of the SiCp/6061Al
composites, a) arrows indicating formation of interfacial reaction products;
dark and short rod-like phase around the SiC particles, b) electrochemically
extracted SiC particles, c) an SiC particle showing extensive formation of
hexagonal platelets of aluminum carbide (Al4C3) growing at the SiC/6061A]
interface, (I) magniﬁed image showing the hexagonal, A14C3 phases with
rejected elemental silicon on its surface. ......................................................... 47

Figure 15. SEM micrographs showing the moisture sensitivity of the SiC/6061A]

interface, a) metallographically prepared SiCp/6061Al composite showing
interfacial reaction products (indicated by arrows) around SiC particles, b) the
interfacial reaction products are dissolved (indicated by arrows) by immersing
into water for 48 hours showing pits. .............................................................. 49

Figure 16. Observation of moisture effect on the SiC/6061A] and Al203/6061Al

composite interface using ESEM: a) as extracted SiC/6061A]; as electro-
chemically extracted SiC particle showing sharp edges in the hexagonal shape
of A14C3 platelets, b) 65 hours moisture environment of SiC/6061A]; dull
edges and cracks in swollen A14C3 platelets indicating dissolution of the
interface reaction product, c)120 hours moisture environment of SiC/6061Al;
remarkably dissolved A14C3 platelets showing the hydrophilic property of the
interface in 2070 SiCp/6061Al composite. The SiC particle indicated by the
arrow, however, remains the same thorough out the moisture absorption, d) as
extracted Al203/6061Al; as electro-chemically extracted A1203 particle
showing formation of MgAl204 at the interface, e) 65 hours moisture

ix

environment of A12O3/6061Al; no change on the interface, 1‘) 120 hours
moisture environment of Al2O3/6061Al; no change on the interface. ............ 50

Figure 17. Magniﬁed images for observation of moisture effect on the SiC/6061A]
interface using ESEM showing severe dissolution of hexagonal shaped Al4C3
platelets, a) as extracted A14C3 particles, b) 20 hours moisture absorption, c)
65 hours moisture absorption, (1) 120 hours moisture absorption (all
micrographs are at the same magniﬁcation) ................................................... 52

Figure 18. Microhardness variation for several PRMMCs and the complimentary
unreinforced matrix alloy as a function of aging time (aging temp. 180°C).
Accelerated aging is indicated by the curve peaks shifting to the left ............. 54

Figure 19. Observation of crack initiation processes in 20"/o (A1203)p/6O61Al composite
by conducting 4-point bend testing on double-notched bar specimens, a)
before the test (U A), b) after the test (UA) showing extensive plastic
deformation around the notch tip (a magniﬁed view(5x) of area I is shown in
Figure 20), c) before the test (OA), d) after the test (0A), the arrows indicating
the void nucleation at the reinforcements and link up by matrix failure. ........ 58

Figure 20. Magniﬁed images of notch tip area in PRMMCs from post-tested 4-point bend
bar specimens, a) magniﬁed image from Figure 19b, arrows indicating
fractured particles around the notch tip and tendency for void initiation at
sharp edges of the reinforcements, b) magniﬁed image from Figure 21d
showing preferential crack propagation path through SiCp cluster regions and
SiCp clusters link up via matrix failure. .......................................................... 59

Figure 21. Observation of crack initiation processes in 2070 SiCp/6061Al composite by
conducting 4-point bend testing on double-notched bar specimens, a) before
the test (UA), b) after the test (UA) showing preferential crack initiation at the
interface of the particles, 0) before the test (0A), (1) after the test (0A)
showing preferential crack propagation through SiCp cluster regions, and link
up of the microcracks by matrix failure (a magniﬁed view(5x) of area II is
shown in Figure 20) ......................................................................................... 61

Figure 22. SEM images of the actual crack propagation in 2070 (A1203)p/6061Al
composite (UA) and a corresponding illustrative drawings of the crack path
proﬁle and damage enclave (indicated by dashed lines) consisting of fractured
particles left in the crack wake (note that ‘X’= transparticle fracture, and ‘0’:
interface debonding). ....................................................................................... 63

Figure 23. SEM images of the actual crack propagation in 20"/o SiCp/6061Al composite
(UA) and a corresponding illustrative drawings of the crack path proﬁle and
damage enclave (indicated by dashed lines) consisting fractured particles left
in the crack wake (note that ‘X’= transparticle fracture, and ‘0’: interface

debonding), a) 20"/o (A1203)p/6061Al composite (UA), b) 2070 SiCp/6061Al
composite (UA). .............................................................................................. 64

Figure 24. Effect of aging on fracture mechanisms along the crack path for the
PRMMCs. 20"/o (A1203)p/6061Al composite shows the transparticle fracture
as dominant fracture mechanism irrespective of aging conditions. 20"/o
SiCp/6061Al composite shows the interface debonding as dominant fracture
mode without signiﬁcant variation due to aging. ............................................ 66

Figure 25. SEM micrographs showing the crack tip of the PRMMCs, a) crack tip of 2070
(A1203)p/6061Al composite(OA) showing the extensive particle cracking
around the crack tip, b) 20"lo SiCp/6061Al composite (PA) showing
preferential crack propagation through SiC particulate clustered region and
microcrack link up between the particulate clusters. ...................................... 69

Figure 26. SEM fractography of 2070 (A1203)p/6061Al composite(OA), a) morphology
of fracture surface, b) cracked A1203 particles, c) both transparticulate
fractured and interface debonded A1203 particle, d) transparticulate fractured
A1203 particle. ................................................................................................. 70

Figure 27. SEM fractography of 20"/o SiCp/6061Al composite(OA), a) morphology of
fracture surface, b) interface debonded SiC particles, c) transparticulate
fractured SiC particle, d) interface debonded SiC particle. ............................. 71

Figure 28. Comparison of fracture toughness (K21) variation for the PRMMCs as a
function of aging time (three specimens were tested to obtain each data point).74

xi

Chapterl

INTRODUCTION

1.1FabrieationofParticulateReinforoedMetalMatrixComposites
A composite material is deﬁned as a material system composed of two or more
constituents, which are essentially insoluble in each other, that form a new material with

speciﬁc properties usually different from the constituents [1].

Generally, a composite material contains one or more reinforcement phases or
constituents for strengthening. The matrix material conﬁnes the reinforcements and
allows for stress redistribution. Aluminum, magnesium, and titanium are commonly used
as matrix materials. Alumina, boron, graphite, and silicon carbide are representatives of
ceramic reinforcements used in metal matrix composites (MMCs). All of these
reinforcements can be produced as continuous ﬁbers, but silicon carbide and alumina are
usually available as discontinuous whiskers. Silicon carbide, boron carbide, alumina, and
titanium carbide are often used in the form of a discontinuous particulate. Discontinuous
reinforced metal matrix composites (DRMMCs) contain either whiskers or particulates

reinforcement or both [2].

Discontinuous whisker and particulate reinforced metal matrix composites have
shown superior isotropy and ductility with low fabrication cost, as compare to continuous
ﬁber reinforced metal matrix composites. With the smallest aspect ratio, the particulate

reinforced metal matrix composites (PRMMCs) have shown exceptional isotropy for

formation of near-net-shape microstructure and great forrnability for large volume
applications. PRMMCs are also amenable to conventional casting, powder metallurgy
processes and metal working. Inexpensive particulate reinforcements such as SiC and
A1203 have been the likely choices for use in PRMMCs. PRMMCs exhibit superior
mechanical properties, such as high stiffness-to-weight and strength-to—weight ratio,
improved wear resistance and low thermal expansion coefﬁcient, as compared to

unreinforced aluminum alloy [2~4].

Fabrication processes for PRMMCs have been developed to produce optimum
interfacial strength and uniform particulate distribution with the lowest level of
contamination and inclusions. Due to a high surface tension of the molten matrix, efforts
have been made on the fabrication of PRMMCs to improve the wettability between
ceramic reinforcements and matrix without excessive interfacial reaction. The wettability
can be improved by the pre-treatments of reinforcements. such as coating, pre-heating
and chemical cleaning. Various fabrication methods of the PRMMCs including ingot
metallurgy (IM) (i.e. casting), powder metallurgy (PM) and spray deposition are briefly

introduced in the following paragraph [5~11].

Figure 1. shows the vortex method for fabrication of PRMMCs. The reinforcements
are mixed and dispersed into the molten matrix by vigorous stirring. The molten alloy is
stirred at high speed to create a vortex, while adding reinforcements into the vortex.
Optimum wetting can be achieved by continuous stirring after completion of adding
reinforcements. This method can utilize the conventional casting apparatus with

minimum modiﬁcation.

 

_
— reinforcement
n i particles
1
inert gas —> — 1°. ./
l
.0 O 3 O

thermocouple tunnel for

A? 7/. m °/ 22:22.2.

heater 0 t '-
o
o

0 :

 

 

 

agitator

 

 

 

 

Figure 1. Schematic diagram for fabrication of particulate metal matrix composite by vortex
method [6].

However, the large amount and ﬁne size (< 50 um) of reinforcements are difﬁcult to
incorporate with the vortex method due to the high surface tension of molten matrix
alloy. The recent development by DURALCAN® shows a stir casting method without the

formation of a vortex, which contains an arrangement of a reactor and a special stirrer

under a vacuum environment [5,6].

The schematic drawing of an apparatus for compocasting is shown in Figure 2. The
compocasting consist of vigorous stirring of a semi-solid alloy upon casting, while adding
the particulate reinforcements to the surface. The ceramic reinforcements are
mechanically entrapped into the matrix, which is in between a solidus and liquidus state,
even without wetting due to the agitation. The interface bonding can be achieved by
abrasive cleaning effect on the particulate reinforcements surface from the continuous
stirring throughout the process. Complementary techniques such as gravity casting,
injection, squeeze casting, rolling and extrusion can be applied after the compocasting.
The compocasting can be advantageous for the fabrication with reinforcements having
particularly low wettability and a large aspect ratio. The entrapment of impurities and
gases upon agitation can cause the excessive porosities and high level of inclusions. The

temperature control for large batch is also a limitation of the compocasting [5,8,9].

The fabrication flow chart of PRMMCs by powder metallurgy (PM) is illustrated in
Figure 3. The near-net-shape distribution of ﬁne particulate reinforcements can be

achieved by employing the powder metallurgy resulting improvement of ductility and

toughness.

 

   

particulate feed

stirrer through

thermocouple

 

resistance or
inducﬁon
heated 2;:;2;2;2
crucible

 

inert gas or
vacuum proction 3? ..
chamber W 535 ................... :2: ................ 52353..

 

 

mold

 

 

 

 

 

 

Figure 2. Schematic diagram for fabrication of particulate metal matrix composite by
compocasting [5].

 

F , ﬂ

I matrix alloy powder I Iceramic reiniorcementsl

precision
m ixinglblending

 

 

 

 

I , a I
canning vacuum hot pressing

Icold isostatic pressing I

     

 

 

 

 

 

 

 

 

 

 

 

 

 

l l
Ihot isostatic pressingl I hotpressing I sintering
I
billet
I rolling I I forging I I extrusion I

 

 

   
 

 

 

 

heat treatm ent

I final machining I

Figure 3. Fabrication ﬂow chart of PRMMCs by powder metallurgy (PM) [ I 1].

 

The deleterious reaction between reinforcement/matrix interface at high temperature can
be avoided by this process, since it is carried out at temperatures below the melting point
of the matrix. In spite of producing high quality PRMMCs, the PM is still limited to low
volume and high performance applications due to the relatively expensive fabrication

costs associated with this method [5,11].

The spray deposition technique for formation of PRMMCs is shown in Figure 4.
The ﬁne structures without macro segregation can be achieved by a high speed injection
of semi-solidiﬁed droplets to the substrate. Due to the rapid solidiﬁcation rate, the
excessive interface reaction can be avoided. The uniform dispersion of particulate
reinforcements can be achieved from the dispersion. However, the application of the
spray deposition techniques has been limited due to the expensive processing with
signiﬁcantly sensitive deposition parameters, such as substrate temperature, deposition

rate, melt superheat, atomizing gas temperature and ﬂow rate [5,10].

 

stopper
rod

    
 
 

induction
heated ladle

i

\ .

deposition
chamber

inert gas

reassess
eerie!

i

 

...............
...............
...............

atomiser

 

 

 

_ particulate
injector

 

 

 

 

 

 

particulate

IIUIdZIed bed substrate

(tube, billet, plate, etc)

 

 

 

Figure 4. Schematic diagram for fabrication of particulate metal matrix composite by spray
deposition [5 I.

PRMMCs have commanded ample interest due to the potential advantages over
monolithic metal alloys in engineering applications. It is known that the addition of
ceramic particles into a matrix alloy can generate admirable improvements of mechanical
properties, such as speciﬁc strength and stiffness, friction, and wear properties. Also, low
coefﬁcient of thermal expansion is achivable. Table 1 shows an example of the change of
various properties due to the addition of 1570 particulate A1203 into 6061 aluminum
alloy. Signiﬁcant improvements are found in thermal expansion, wear resistance, elastic
modulus and ultimate tensile strength with minimal increase in density. Tensile
elongation and fracture toughness, however, are considerably reduced by the addition of

the particulate reinforcements.

Table 1. Properties of reinforcement (particulate A1203), a matrix alloy (6061/11), and a
composite (15 "/o A1203/6061A!) [10,12~14].

 

A1203 (particle) 6061 aluminum (A1203)p 15 "/o/6061Al

 

 

(T6) (T6)
density (g/cm3) 3.98 2.68 2.86
coefﬁcient of thermal 7 .0 23.4 19.8
expansion (1047K)
*volume loss by wear test --- 8.75 0.0174
(mm’)
elastic modulus (Gpa) 460 69.0 88.9
tensile elongation (%) --- 20 6
ultimate tensile strength 8 0.31 0.37
(Gpa)
fracture toughness 4 27.0 22.0
(Mpa «lb—t )

 

*volume loss (mm3) by block-on-ring wear test (ASTM G77). 4140 steel (HRC = 50-60) was used for the
ring material. Tests performed in 10W40 motor oil for 2 hrs. with a 667-N load at 36 rpm [12].

 

10

Engineered variability of mechanical properties is achieved in PRMMCs by the
appropriate selection of the type, volume fraction, size, shape of the reinforcement phase
and matrix alloy. Such ﬂexibility allows the PRMMCs to meet performance
requirements deemed difﬁcult using monolithic materials. Tailored PRMMCs enable the
automotive, aerospace and military industry to achieve improved materials performance
with traditional fabrication methods at reasonable costs. However, the lack of fracture
toughness and tensile ductility are the major limitations of PRMMCs in structural
application. Efforts have been made to improve the low ductility and toughness in

PRMMCS [15].

1.2 Agng Kinetics in Aluminum Alloys

Strengthening of aluminum alloys by heat treatment can be achieved by the
following processes: solution heat treatment for dissolution of soluble phases, quenching
for development of supersaturation and age hardening for precipitation of solute atoms.
The solution heat treatment sets the maximum practical amount of the soluble hardening
elements such as copper, magnesium or silicon into solid solution in the aluminum
matrix. Quenching is performed in order to produce a supersaturated solid solution (SSS)
which preserves the solid solution state formed during solution heat treating at
temperatures of 450~550°C. Dislocations are generated from the quenching of the
solution heat treated aluminum alloys due to quenching strains, condensation of excess
vacancies and differential contraction in constituents or dispersoid particles. Age

hardening is achieved from either natural aging at room temperature (about —20°C to

60°C) or artiﬁcial aging at a moderately elevated temperature (less than about 250°C).

11

During natural aging and at initial periods of artiﬁcial aging, redistribution of the solute
elements in the supersaturated solid solution (SSS) occurs forming clusters or GP
(Guinier-Preston) zones with lattice structure coherency. These precipitates nucleate
preferentially at heterogeneous sites such as grain boundaries, dislocations, and dispersoid
particles having greater disorder and higher energy. The GP zones generally form a disk,
a needle, or a sphere like shape, and are located parallel to some low-index plane of the
matrix lattice, in the size range of 10 ~ 100 A in diameter. The alloying elements
determine the shape, orientation, and size of the GP zones. The formation of GP zone
produces coherency strain by the distortion of lattice planes. The distorted region
provides additional interference to dislocations when they move to cut through the GP
zones. The strengthening effect is caused by the additional stress required to move a
dislocation through the GP zones. The maximum strength of the alloy results from a high
density of ﬁne GP zones or precipitates interacting with dense dislocations. The GP
zones are characteristically metastable. Hence, the GP zones are converted or dissolved
into equilibrium precipitates which become incoherent with crystal structure of the parent
lattice, as aging temperature or time is increased. This transition of precipitates occurs
when the strength of the interfacial bond is exceeded by growth of the GP zones
accompanying increase in coherency strains. The transition to equilibrium precipitates
increases interparticle spacing and results in dislocations to loop around rather than to cut
through the precipitates. The effect of strengthening accordingly decreases with the
growth of the equilibrium precipitates. Hence, the following three mechanisms are
associated with precipitation hardening. The ﬁrst is ‘strain hardening’ in which the strain

ﬁelds impede the dislocation motion and result in an increase of ﬂow stress. The strain

12

ﬁelds are caused from the nucleation of the coherent particles with slightly different
lattice parameters. The second is ‘chemical hardening’ which is caused by an increased
number of solute-solvent bonds from the presence of the solution-rich zones. The
additional stress is required for dislocation motion. The last mechanism is ‘dispersion
hardening’ in which the dislocation motion is blocked by obstacles from the dispersion of

ﬁne precipitates [16].

Table 2. Nominal chemical composition (weight percent) 6061 aluminum alloys [1 7].

 

Al Cu Mg Si Mn Cr Zn Fe Ti Others

 

 

balance 0.15 ~ 0.8 ~ 0.4 ~ 0.15 0.04 ~ 0.25 0.7 0.15 0.15

0.4 1.2 0.8 max. 0.35 max. max. max. max.

 

The composition range of 6061 aluminum alloy (Al-Mg-Si) is in Table 2 and, the

precipitation sequence for the alloy may be described as follows:
SSS —> GP —> B’ —9 B (Mg2Si)

The first stage of precipitation tends to occur in the <001> direction from the
supersaturated solid solution with a ﬁne needle-like shape. These GP zones are about 60
A in diameter and 200 to 1000 A in length with a minor coherency strain. However,
others contended that GP zones are initially spherical in shape and eventually evolve into
needle-like shapes [18]. With an increase in aging time, GP zones are transformed into [3’
transition precipitates which have a three dimensional rod-like shape with a highly

ordered Mg2Si structure. The detailed sequence of precipitation consists of SSS,

 

l3

vacancy-silicon clusters, vacancy rich coherent Al-Mg-Si GP zones, partially coherent
and disordered needle-shape phase in <100>, partially coherent and ordered needle-shape
phase aligned in <100> directions, semicoherent hexagonal rod-shape phase, and
equilibrium Mg2Si platelets. Hardening of the alloy system results from the additional
hindrance to dislocation motion by the stress required to break magnesium-silicon bonds

in the zones and transition precipitates [19].

ranging rum in 111108

The nucleation and growth kinetics of precipitation in an aluminum alloy can be
signiﬁcantly changed by the addition of a ceramic reinforcement to the matrix.
Consequently, such change in the aging kinetics of the matrix can alter the overall
properties of MMCs. Numerous investigations have clearly veriﬁed that age hardening
behavior of the MMCs, which are fabricated by either powder metallurgy (PM) or ingot
metallurgy (IM), is quite different from that of the unreinforced matrix [20~36]. The
accelerated aging has been extensively observed in discontinuously (i.e. whisker or

particulate) reinforced MMCs as compared to the unreinforced matrix alloy.

Results of earlier study on 23% Sij/6061Al MMC (PM) [20] showed that
accelerated aging was primarily due to the high dislocation density generated from the
thermal expansion mismatch between the reinforcement and the matrix, and due to the
presence of a highly diffusive interface in the MMC. Consequently, both factors
increased the diffusivity of precipitate-forming elements, and thus change the aging

kinetics inﬂuencing solute/vacancy diffusion.

14

Dislocation generation due to the thermal expansion mismatch was observed by in-
situ High Voltage Electron Microscope (HVEM) study on 2070 SiC..,/6061Al (PM)
composite. In this work, the generation of dislocations at Al/SiC interface and
subsequent nucleation of precipitates on the dislocations were clearly observed upon

cooling from the annealing temperature [21,22].

An investigation of 13.2 ‘70 Sij/2124Al MMC (PM) [23] revealed an increase in
dislocation density due to Sij reinforcements. Transmission electron microscope (TEM)
showed dislocations punched out from the whisker end upon cooling. The increased
dislocation density caused by thermal expansion mismatch provided the heterogeneous

nucleation sites for strengthening precipitation with resultant accelerated aging [23].

Another investigation [24], employing differential scanning calorimetry (DSC), also
found that the precipitation and dissolution kinetics were accelerated in SiC reinforced
MMCs, such as 8 and 20% SiC../2124Al (PM), 20" c Sij/2219Al (PM) and 20%
SiCp/6061Al (PM). However, the overall age hardening sequence of the MMCs was not
changed by the addition of SiC reinforcements, as compared to their unreinforced matrix
alloys. Effect of solution heat treatment on 6061 Al alloy is minimum as compared to
other aluminum alloys. Hence, the monolithic 6061A] alloy is known as quench-
insensitive alloy [16]. In addition, the quench-insensitive 6061 aluminum alloy was
observed to become quench-sensitive by the addition of SiCp, which promoted the

precipitation of GP zones or the intermediate phase during quenching [24].

Two principal mechanisms regarding accelerated aging have emerged from the

theoretical and experimental studies of 10% Sij/6061Al (PM) composite [25]. One is a

15

dislocation dominant mechanism involving large-sized reinforcement phase (whisker
radii >1 um) and high dislocation density (>10l4 m'2) which states that the dislocations
serve as heterogeneous nucleation sites for precipitation. The other is a residual stress
dominant mechanism involving small-sized reinforcement phase (whisker radii < 0.25
um) and low dislocation density (<10l3 m’z) which suggests that residual stress induces

the diffusion of solute atoms [25].

Accelerated aging is consistently observed in various MMC systems. Some are

listed in table 3.

Table 3. Observation of accelerated aging in various MM C systems.

 

 

Reinforcement / Matrix Processing Method Particle Size Reference
30% SiC/7091A] powder metallurgy ~ 5 um [27]
ZOw/o SiCp/8090Al powder metallurgy ~ 3 pm [28]
17"/o SiCp/2014Al powder metallurgy --- [29]
13.9% SiCp/6061Al powder metallurgy --- [30]
15v/0(A1203)p/6061A1 powder metallurgy --- [31]

10 & 15"/o (A1203)p/6061Al ingot metallurgy 0.5 ~ 25 um [32]

 

The size and shape of the reinforcement control the intensity of dislocation generation
and resultant aging kinetics. It was reported that the intensity of dislocation generation

was lowest for small, nearly spherical particles (< 1 mm), but increased with increasing

l6

particle size (1~5 pm). The sharp comer of particulates or whisker ends also showed

intense generation of dislocations due to an increased local strain [21,22].

The interfacial reactions in MMCs can inﬂuence the aging kinetics by changing the
chemical composition of the matrix. The rejection of silicon was found from the
interface reaction by remelting 10"/o SiCp/6061Al (IM) composite. A hardness increase
was observed by the presence of the additional silicon in solid solutions [33]. On the
other hand, migration of matrix magnesium to the interface was found in the 10%
(A1203)p/6061Al (IM) composite. During the solidiﬁcation, magnesium in the matrix
diffused to the interface to form a reaction product MgAl2O4 resulting in possible change

in matrix aging kinetics [34].

At the constant volume fraction, the average dislocation density was decreased by
increase of reinforcement size, hence, accelerated aging was considerably less in MMCs
with relatively large reinforcements [21,22,26]. Consequently, mechanical properties
such as hardness and strength were improved in a MMC with a small particle or whisker
size (~5 um) which has higher dislocation density or a smaller mean free path between

heterogeneous nucleation sites [35].

The difference of aging kinetics was reported between powder metallurgy (PM),
such as cold packed vacuum sintering, and ingot metallurgy (IM), such as stir casting.
Even though the kinetics of aging in ingot metallurgy (IM) processed materials are
remarkably slower than those observed in powder metallurgy (PM) processed materials
[24], accelerated aging was observed for the MMCs fabricated by both the PM and the 1M

process [20~32].

17

Furthermore, the type and amount of precipitation in the matrix were also affected
by the amount of reinforcement. It was found that the addition of 15% (A1203),, to
6061A] increased the dislocation density slightly more than that of IOV/o (A1203)p.
Consequently, microhardness for as-quenched matrix was elevated with the addition of
alumina particles. The increased strength of these materials apparently resulted from the

high dislocation density in the composite matrix [32].

Solution treatment temperature also affects aging kinetics in MMCs. The effect of
solution heat temperature was observed by using two different temperatures (529°C and
557°C) on 13.9"/o SiCp/6061Al (PM) MMC [30]. The increase in hardness was more
pronounced for composite with solution heat treated at 557°C, as compared to its
monolithic 6061 aluminum. Therefore, greater dislocation density in the matrix was
achieved from the higher solution heat treatment temperature under solidus temperature

upon quenching.

The inﬂuence of an aging temperature was also examined on 17% SiCp/2014Al
(PM) MMC employing various aging temperatures from 150°C to 195°C. Accelerated
aging increased with aging temperature over the range of 150°C to 195°C. Upon aging at
the highest temperature (195°C), the maximum accelerated aging was accomplished by
the heterogeneous nucleation of fast growing 8’ precipitation (CuAlMg2) on the dense
dislocations. As the aging temperature was decreased, heterogeneous nucleation of GP
zones, with low growth rate, became dominant over S’ formation. Clearly, the extent of

accelerated aging in matrix depends on the aging temperature [29].

18

Accelerated aging is associated with the coefﬁcient of thermal expansion mismatch
(ACTE) between the matrix and the reinforcement with a result of an increased
dislocation density and/or enhanced diffusion. The strain ﬁelds generated by ACTE, upon
quenching, are relaxed by the formation of dislocations in the matrix. The increased
dislocations due to the ACTE can serve as sites for the heterogeneous nucleation of
strengthening precipitates and as a short circuit of the diffusion path for solute atoms
[36]. Furthermore, the elastic ﬁelds from residual stress around the reinforcements can

induce the enhanced diffusivity at relatively low dislocation density and/or large particle

size [25].

1.4 interface in MMCs

The nature of the interface in metal matrix composites is the most important for
tailoring mechanical properties since the maximum mechanical strength of MMCs is
achieved by the proper load transfer from the matrix to the reinforcement. The efﬁciency
of the load transfer strongly depends on the bonding strength of the interface and the
optimum bonding strength can be attained from the formation of adequate reaction layer
at the interface. Eventually, the interface reaction can improve the wettability of
ceramic/metal system from the development of interfacial phases on the interface.
However, the poor adhesion of the interface can result from the formation of brittle
interrnetallic compound due to a massive reaction. Interface debonding, upon loading,
can be caused by poor adhesion resulting in strength decrease of MMCs. Hence, the
interface reaction may have either beneﬁcial results, by increasing the ductility and the

fracture toughness, or deleterious effects, by causing crack initiation points. The interface

19

reaction is controlled by the matrix composition, the type of reinforcement, the

fabrication method and the heat treatment [35~48].

1.4.1 SiCI Aluminum interface
Intermetallic compounds form at a SiC/aluminum interface [37~42]. A typical

reaction is given as follows:
4A] (1) + 3SiC (s) —> Al4C3 (s) + 3Si (s), (1)

where the (l) and (8) represent liquid and solid respectively. The reaction product, Al4C3,
is a brittle and hydrophilic compound which is detrimental to mechanical properties of
MMCs. Excessive formation of Al4C3 at interface of a composite may result in overall
degradation of mechanical properties, such as reduction of reinforcement and interfacial
strength and an increase in corrosion sensitivity. [38,39]. The change of matrix
composition was also observed in that the hardness of the matrix was increased due to the
additional silicon rejection from the interface reaction [33]. The formation of A1403 is

mainly due to the prolonged contact between SiC and molten aluminum [37,40,42].

From the interface investigation on compocasted SiCp/6061Al MMC [40], with as-
received and oxidized SiCp, no reaction products were found at the as-received SiCp
interface. During this time, the MgAl2O4 was formed at the interface of oxidized SiCp by
the reaction between SiO2 layer and molten aluminum. However, a signiﬁcant amount of
A14C3 was found at the interface between as-received SiCp and the matrix by remelting
the composite at 800 °C for 20 min. Thus, the formation of A14C3 depends on the casting
route. Compocasting, which has a reduced fabrication time at a high temperature (above

650°C), can prevent the formation of the interfacial product.

20

The morphology of the SiC/aluminum interface was recently revealed by the
observation of electrochemically extracted SiC particles from a spray formed 20%
SiCp/2024Al composite. Massive reaction products formed during a 24-hour heat
treatment at 560°C and 640°C. Elevated growth rate was noted at higher temperature,
640°C. SEM examinations showed that SiC particles were covered with Al4C3 platelets,
having a HCP crystal structure. Dendrites of elemental Si were also present. Si exhibits a
diamond cubic structure. It was found that the A14C3 was formed at moderately elevated

temperature, ~450°C [4 l ].

The extent of formation of A14C3 at the interface can be altered by controlled
processes; such as an addition of silicon to the matrix alloy to prevent the interfacial
reaction by increased free silicon in the matrix [42], a suitable fabrication route without
prolonged contact between SiC particles and molten aluminum [38,40,43], and formation

or coating of some oxide layer like SiO2 and Ti02 on the SiC reinforcement [44].

1.4.2 Alp, [Aluminum Inmrface
Wettability and bonding at the A1203/aluminum interface are improved by the
formation of spinels (MgAl204) from the interface reaction. Spine] promotes
strengthening of the interface by forming strong bonds with metals and ceramics
[34,45~47]. The interface product, MgAl2O4, forms when sufﬁcient contact occurs over

time between alumina and molten aluminum alloy [45].

Pertaining to interface in A1203 short ﬁber (Safﬁl) and Al-Mg (2~3%) alloy
composite, magnesium diffusion toward interface was observed upon solidiﬁcation and

heat treatment at 500°C. Magnesium diffused and reacted with the alumina ﬁbers and

21

silica coatings forming interfacial reaction products, such as MgO, MgAl2Si, Mg2Si and
Si. Magnesium diffuses continuously to the interface during heat treatment where it in

turn reacts to produce Si and Mg2Si precipitation in the matrix [47].

A detailed interface characterization of 10% (A1203)p/6061Al (IM) composite
revealed the following. The morphology of MgAl2O4 single crystal was revealed by
electrochemical dissolution of the conductive matrix alloy around the alumina particles.
The alumina particles were almost fully covered with MgAl2O4 (spinel) of average size 1

um. MgAl2O4 single crystals were grown at the surface of the alumina particle by the

following chemical reaction:
4 2
Mg (1) + §Al203(s) —> MgAl2O4(s) + 3A](l). (2)

The reaction occurs between molten aluminum and solid alumina particles when
segregation of magnesium at the interface region results. Another possible reaction

between SiO2 and molten matrix was also considered as the following reaction:
28i02 (s) + 2A] (1) + Mg (1) -—> MgAl2O4 (s) + 2Si (s). (3)

However, the reaction between SiO2 and molten matrix was suggested to be less

signiﬁcant in this observation [34].

1.5 Fracture of PRMMCS
It is generally known that the embedded ceramic particles in a matrix alloy, which
forms a composite material, usually generate admirable improvements of mechanical and

physical properties, such as high speciﬁc strength and stiffness, enhanced friction and

22

wear capabilities, and low coefﬁcient of thermal expansion. However, tensile elongation
and fracture toughness of PRMMCs are reduced considerably by the addition of the
particulate reinforcements. Compared to the matrix alloy, the lack of the fracture
toughness and tensile ductility limits the use of particulate reinforced metal matrix
composites (MMCs) in structural applications in spite of other advantageous properties.
Even though PRMMCs show signiﬁcantly lower fracture toughness and tensile ductility,
fracture still occurs by a ductile rupture mechanism similar to that of matrix material.
The failure process of unreinforced alloys and MMCs alike arises by a sequence of void

nucleation, growth and coalescence [3,4].

In PRMMCs, void nucleation tends to occur by the following processes; (i) by
transparticle cracking and/or (ii) by particle/matrix interface debonding [49~53]. The
crack propagates by the matrix failure resulting from link-up of microfracture defects and
coalescence of rnicrovoids. Void nucleation is strongly inﬂuenced by the composite
material’s microstructural composition including particulate shape, size and volume
fraction, and matrix temper by aging [52,54~61]. The interfacial bonding strength
between reinforcement and matrix critically affects void nucleation of the PRMMCs. The
formation of brittle intermetallic compounds and other phases at the reinforcement/matrix
interface, such as A]4C3, can cause premature void nucleation at the interface resulting in
decreased load transfer efﬁciency [38,39]. Reinforcement content (Vf) and matrix heat
treatment also inﬂuence void nucleation, and consequently, the resultant mechanical
properties of MMCs. Fracture toughness usually decreases with an increase of

reinforcement volume fraction. Also, studies show that unlike the matrix alloy, neither

23

the fracture toughness nor tensile elongation recover upon over aging of the MMCs

[50.53.56.59].

Void nucleation in particulate reinforced metal matrix composites were studied
using acoustic emission monitoring during tensile testing [57]. Three volume fractions
(5, 10 and 20 vlo) and three sizes (nominally 3, 10, 30 um ) of SiCp reinforced in 1070A]
and 5050A] were investigated to observe the effects of microstructural parameters on the
fracture behavior. It was observed that the void nucleation occurred from the onset of
plastic deformation and continued throughout the plastic regime. Both particle cracking
and interface debonding were observed as a mode of the void nucleation. The rate of
damage initiation increased with particle size and volume fraction. However, the damage
initiation was independent from the matrix composition and heat treatment. It was
concluded that the failure strain or ductility of the composites was not controlled solely
by void nucleation at the reinforcements, but local failure processes occurring in the

matrix during void growth and coalescence [57].

The fracture toughness of discontinuous reinforced metal matrix composites
(DRMMCs) with different reinforcement type, particulate size and matrix alloy was
observed for SiC/A356A], B4C/A356Al, SiC/A201Al and B4C/A201Al [52]. At a
constant reinforcement volume fraction, the fracture toughness increased with an increase
in particle size and interparticulate spacing. The fracture toughness of the B4C/A356Al
composite was shown to be substantially lower compared to the SiC/A356Al composite.
Apparently, the lower fracture toughness in RC reinforced MMCs was due to the

extensive reaction at the reinforcement/matrix interface and as a consequence matrix

24

microstructure change due to redistribution of alloying elements during the molten-state
of the composite processing. Discontinuous crack growth for fracture of DRMMCs was
also suggested from employing the acoustic emission and metallographic technique. The
discontinuous crack growth process occurred by a rather strict sequence of failure: the
initial transparticular fracture or reinforcement/matrix interface debonding and following

ductile failure in the interparticular ligaments by void coalescence.

The effect of particle size was studied with the ~50 vlo SiCp/Mg(4“'lo) -Al (IM)
composite with a particle size of 3.5 ~ 165 um. The particle size effect is suggested to be
exclusively related to the incidence of damage in the composite. In four point ﬂexural
tests, particle cracking was shown to occur in the larger particles when the probability of
the cracking increased as the particle size increased. Particle cracking reduces the ﬂow

strength of the materials and subsequent crack coalescence controlled the ultimate

strength and ductility [58].

The effect of particle size on fracture toughness was also studied on 20%
SiCP/1100Al with 2.4, 3.2, 8 and 20 um average particle size. Fracture toughness, K2,
was not improved with the increase of particle size in this particular study [61]. Hence,
the he fracture toughness was independent of the SiC particulate size up to 20 um
average size. For constant particle size, the me fracture toughness was indicated to be
decreased with increased volume fraction that retained smaller interparticle space. The
fracture toughness was not changed with an increase of the particle size at the same

volume fraction. The ch fracture toughness was mainly dependent on the volume

25

fraction of SiC particles [61]. Other studies, however, have reported an increase in

fracture toughness with increasing particle size [26, 52]

The effect of aging conditions on fracture properties were studied with the 15 &
20"lo SiCp/Al-Zn-Mg-Cu (PM) composite. Unlike an unreinforced matrix alloy,
signiﬁcant differences in fracture initiation and growth toughness were observed due to
aging conditions. Both fracture initiation and growth toughness of materials in the UA
condition were about two times higher than that of in the 0A condition. The toughness
decrease between materials in the UA and the 0A conditions was due to the transition in
the local fracture mode form SiC particle cracking to the SiCp/matrix near interface
debonding. Void nucleation by particle cracking impedes void coalescence by a
constraining matrix ﬂow, leading to a increase in ductility and fracture toughness. The
SiCp volume fraction affected the fracture initiation toughness which decreased linearly

with an increase in the volume fraction of SiCp [53].

From a study of tensile behavior on 20% SiCp/6061Al (IM) composite, the T4
tempered composite showed more particle cracking than the T6 tempered composite
throughout the gauge length. More particle cracking in the T4 condition was due to
extensive plastic strain distribution in the gage length of the composite. In contrast,
localized strain was found for material in the T6 condition so that the fracture was
initiated in the clusters of reinforcement particles. Fracture in a reinforcement clustering
zone was a result of, not only particle cracking, but also fracture of the matrix ligaments
with high triaxial stress due to the elastic misﬁt and the plastic constraint of the particles.

Crack initiation and growth were due to the fracture of the matrix in the clustering zone

26

together with any fracture of the particles in the zone. Subsequently, cracks linked by fast

fracture through the matrix to adjacent clustering zones, resulting in macroscopic fracture

[50].

Fracture behavior of 20% SiCpﬂXXXAl (PM) composite was studied focusing on
microstructural variables, such as the SiC particle size and matrix temper. A strong effect
of aging was correlated to an increased fracture occurring by particle/matrix interface
debonding in OA condition. The dominant mode of the fracture at UA condition was
particle cracking with preferential fracture of large particles. The increase of interface
debonding on OA condition was due to the precipitation at the SiC particle/matrix
interface which reduced the interfacial bond strength. The reduced interfacial strength
caused by the precipitation may affect the lack of recovery in fracture toughness in OA
condition. By conducting 4-point bend test on double notched bar composite specimens,
preferential crack initiation was observed in clustered particle regions. Consequently,
particle clustering plays a major role in determining the strength-fracture relationship in

PRMMCs regardless of matrix temper [59].

The inﬂuence of aging condition on the fracture toughness was studied on 15"/o
(A1203)p/2014Al (IM) and 15"lo (A1203)p/6061Al (IM) [56]. The fracture toughness of the
MMCs decreased with aging from UA to 0A condition, and no recovery of fracture
toughness was found in OA condition, unlike the unreinforced matrix alloy. The lack of
recovery of fracture toughness in OA condition was due to the strain localization caused
by the formation of ligaments upon the fracture of the reinforcement particles. The

ligaments showed earlier fracture in OA condition by the formation of secondary

27

microvoids within the ligaments which were due to the presence of large second-phase
precipitates in the matrix. Remarkably, no reinforcement/matrix interface failure was

found at any aging condition in these MMCs.

A comparison of fracture properties between particulate reinforced (SiCp) and
whisker reinforced (Sij) MMCs is observed from the investigation on SiCp/6061Al and
Sij/6061Al composites. The SiCp/6061Al MMC exhibited moderately superior
strength, ductility, fracture toughness and crack propagation resistance, as compared to

the Sij/6061Al composite [60].

Crack propagation was observed by in-situ deformation of 15"/o (A1203)p/6061A1
(IM). The in-situ observation was conducted on tensile specimens by scanning electron
microscope (SEM) focused on the crack tip region. Microcracks were observed both near
and ahead of the crack tip region. Microcracking ahead of the crack occurred in a region
which was considerably larger than the inter-particle spacing. Crack propagation
mechanisms were sequential processes involving initiation of microcracks, link-up of

microcracks with macrocrack by matrix failure, and growth of macrocracks [54].

Void nucleation occurs at the reinforcements in PRMMCs. The mode of void
nucleation has been observed by two dominant ways: through particle cracking and
particle/matrix interface debonding [49~54]. Figure 5 shows the illustration for the
fracture mechanism around the tip of a crack, such as reinforcement/matrix interface
debonding, through particle fracture, and matrix failure. Figure 5 also shows the void
nucleation in the plastic deformation zone nearby the crack. The crack propagates by the

matrix failure which result from link-up, growth and coalescence of microvoids. The

28

mode of void nucleation is intensely inﬂuenced by microstructural parameters in
PRMMCs [38,39,50,53,56,57]. The reinforcement size and the bonding strength of a
reinforcement/matrix interface are the signiﬁcant parameters controlling void nucleation
of the PRMMCs. The increase of particle size causes a change from interface debonding
to particle cracking resulting in an increase in fracture toughness [58]. The formation of
brittle intermetallic compound at reinforcement/matrix interface can cause the void
nucleation at the interface, resulting in a decrease in load transfer efﬁciency [38,39]. The
volume fraction of reinforcement and matrix heat treatment are also important parameters
for void nucleation. Fracture toughness is observed to be decreased with an increase of a
reinforcement volume fraction and is not recovered at 0A condition, unlike the
unreinforced matrix alloy [50,53,56]. The increase of particle size and volume fraction of
the reinforcement causes an increase in void nucleation rate. However, the void
nucleation rate is independent of matrix heat treatment which controls the yield strength
and ductility of PRMMCs. The ductility of PRMMCs, hence, is not inﬂuenced by void

nucleation but by matrix failure processes, such as void growth and coalescence [57]

29

.1

matrix

I

reinforcements

 

Figure 5. Illustrative micrograph for crack propagation mechanism at the crack tip in
particulate reinforced metal matrix composite (15% (A1203),/2014Al composite), a)
reinforcement/matrix interface debonding, b) through particle fracture, c) crack propagates
through reinforcement/matrix interface of fractured particle, d) matrix failure, e) fractured
particle around the crack, f) crack tip, g) reinforcement/matrix interface debonding ahead of the
crack tip, h) fractured particle ahead of the crack tip.

Chapter II

MATERIALS AND EXPERIMENTAL PROCEDURES

2.1 Materials

The materials investigated were the 6061 aluminum alloy based PRMMCs,
(A1203)p/6061Al and SiCp/6061Al. Both composites consisted of 10 and 20 volume
percent (‘76) reinforcement. The composite materials were fabricated by Duralcan Inc.,
San Diego, CA, using a stir-melt cast process, and received as extruded bars, 19 mm thick
and 77 mm wide. Unreinforced commercial 6061 aluminum alloy was also investigated

for comparison. Typical mechanical values of as-received material are given in Table 4.

Table 4. Properties of matrix alloy (6061/11), and PRMM Cs [10,12].

 

6061A] 10 '10 (A1203), 20% (A1203), 10 'r. Sic, zo'ro Sic,
(T6) I6061AI(T6) l6061Al(T6) I6061Al(T l6061Al(T6

 

 

6) )
elastic modulus (GPa) 69.0 81.4 97.2 -- 103.4
tensile elongation (%) 20 10 4 -- 5.5
ultimate tensile strength (MPa) 276 296 352 -- 496.4
fracture toughness (mph/:22.) 29.7 24.1 21.5 24.7 20.5

 

30

 

31
22 A9509
All specimens were solution heat treated at 550°C for 90 minutes and quenched in
ice brine. Artiﬁcial aging was conducted at 180°C, immediately after the solution heat

treatment to prevent the natural aging at room temperature. Aging was accomplished

over a 12-step time sequence ranging from 1 to 3,000 minutes.

Hardness testing was conducted using a Leco® M-400-G microhardness tester
equipped with a Vickers diamond pyramid indenter. A 10 gram load was used to obtain
the change in hardness due to aging. Hardness indentations were taken exclusively in
matrix regions avoiding contact between the indenter and the reinforcement phase. At
least 10 hardness readings were taken for each composite sample. Figure 6a shows that
indentation is placed well between the reinforcement particles indicating no contact

between the indenter and the reinforcement phase.

23 Metallographic Preparation

Metallographic preparation for PRMMCs is signiﬁcantly different in comparison to
the preparation of monolithic alloys. Polishing of PRMMCs is more difﬁcult due to the
large hardness difference between the metal matrix and the ceramic reinforcement. The
extruded composite bars were sectioned into cubes (10 mm x 10 mm x 10 mm) for the
heat treating using a silicon carbide impregnated abrasive whee] cutter and a precision
diamond saw. After the aging, the composite specimens were encapsulated in a
metallographic mount, 1% inch (3.175 mm) in diameter, for polishing. Initial grinding of
the polishing process was accomplished using an 120 grit zirconia/alumina abrasive belt

polisher. The mounted specimens were subsequently polished on a lapping wheel using

32

30 um, 6 um and lum diamond suspensions. The ﬁnal polishing step was conducted

using 0.06 um colloidal silica suspension on a vibratory polisher (Buehler Vibromet II®)

for 2 hours. Detailed description of this procedure is given in Table 5.

Table 5. Descriptions of metallographic preparation for optical and SEM samples.

 

 

step 1 Section as extruded 19 mm thick and 77 mm wide bars into 10x10x10(mm)

cubes using a silicon carbide impregnated abrasive wheel cutter and a precision
diamond saw.

step 2 Heat treatments (Solution heat treatment and artiﬁcial aging).

step 3 Metallographic mount in 11/4 inch diameter by thermoplastic Lucite.

step 4 Remove major roughness using zirconia/alumina impregnated abrasive belt
polisher (120 grit.).

step 6 8 inch lapping wheel polisher for 3 steps, 30 um diamond suspension with
'ULTRA-PAD® cloth, 6 pm diamond suspension with 'POLIMET” cloth, and 1

pm diamond suspension with ‘TEXMET 2000® cloth. Each step is taken for ~5
minutes.

step 7 Final polishing (~2 hours) by vibratory polishing machine ‘VIBROMET 2‘8
using 'MASTERMET” (0.06pm colloidal silica suspension) with 'TEXMET
2000‘ID cloth.

 

 

*Products from BUEHLER" Inc.

2.4 Electropolishing and TEM Sample Preparation

In order to obtain more details of the interface morphology, the aluminum matrix
was removed by electro-chemical dissolution with a potential of 10 DC volts and 22 Amp
current in 5°C electrolyte composed of 33% HNO3/67% methanol solution. The
(A1203)p/6061A1 composites were electro—chemically dissolved for 5 minutes and cleaned

with methanol. Care was taken to limit the exposure of SiCp/6061Al composites to a

33

moist environment due to the hydrophilic nature of the interfacial reaction products
(A14C3). Electrochemical dissolution for the SiCp/6061Al composites was allowed for ~3
minutes and followed by cleaning in acetone immediately. The reinforcement phase was
extracted from the matrix by electrochemical dissolution. The extracted particles were
observed with a Hitachi S-2500C® scanning electron microscope (SEM). The
observation of moisture effect on the interface of SiCp/6061Al composite was observed
by immersing of electro-chemically extracted SiC particles into de-ionized water in the

Electroscan 2020® environmental scanning electron microscope (ESEM).

Specimens for transmission electron microscope (TEM) examination were prepared
from peak aged (PA) 20"/o (A1203)p/6061Al and 20"lo SiCP/6061Al composites to
characterize the interface microstructure. Mechanically-thinned (~100 um) composite
specimens were electropolished by a Dimpler® machine until about 20 um thickness was
achieved. Final thinning was performed using an argon ion mill at an accelerating voltage
of 6 KV with 15 ~ 5° sample inclination to the ion beam. Thin foils were cleaned for 3
minutes with the sample inclination of 20° after a hole was obtained in the foils. TEM
examinations were conducted using Philips CM30® operated at an acceleration voltage of
200 KV. Convergent beam electron diffraction (CBED) was also performed on the

interface regions using a 50 nm electron-beam spot size.

2.5 Fracture Toughness Testing

Fracture toughness tests were conducted on the materials using Fractometer II® test
system. Half-inch width (B = 0.5 in) short bar (SB) specimens were tested according to
ASTM 1309 test procedures. A schematic diagram for the chevron notch short bar

specimen is given in Figure 7. When conducting a chevron notch SB fracture,

34

precracking is not necessary since a sharp crack initiates at the apex of the chevron notch
ligament in the specimens. A sharp crack at the notch tip can be initiated at relatively low
loads due to the high stress concentration. The fracture toughness is determined from the
peak load value which occurs when the propagation of crack attains a critical length in the
SB specimens [52]. A short bar fracture toughness specimen is shown in Figure 6b.
Fracture toughness testing was conducted on specimens in the under aged (UA), peak
aged (PA) and over aged (0A) conditions to ascertain the aging effects. Using the SEM,
the fracture morphology was assessed by examining fracture surface of PRMMC
specimens. The crack propagation path was observed from half-sectioned (transverse to
the chevron notch direction) and polished SB specimens using optical and scanning
electron microscope. Figure 6c shows a transverse sectioned SB specimen used for the

observation of the crack propagation proﬁle.

2.6 Bend Testing

Crack initiation mechanisms were observed using 4-point bend test specimens. A
double-notched bar specimen is shown in Figure 6d. The dimension of the bend
specimen is also given in Figure 8. In 4-point bend testing, one notch will fail by
catastrophic crack growth, while the companion notch will be preserved in a pre-crack
propagation deformation state. Crack initiation behavior was observed for 20%
(A1203)p/6061Al and 20% SiCp/6061Al composites in the UA, PA and 0A conditions.
Optical micrographs were taken before and after bend tests were conducted on polished

surface around the notch tip region to assess changes in crack initiation due to aging

effects.

35

 

(pr. ..

fracture toughness to st

30 um

 

crack propagation

 

 

 

 

 

Figure 6. Optical micrographs showing various specimens for mechanical testing and typical
location of hardness indents, a) hardness indentation placed well between the reinforcement
particles, b) half-inch width (B=O.5) short bar fracture toughness specimen, 0) half-sectioned
and mounted fracture toughness specimen for crack propagation observation, d) double-notched
4-point bend bar specimen for crack initiation observation.

 

 

 

 

 

 

w = 0.750 ”
4 r
A H = 0.435 ”
. sen-"5" .

 

B =0.5 ”

 

 

 

 

Figure 7. A schematic diagram showing the half-inch width short bar fracture toughness
specimen [52].

37

 

 

 

 

 

, l A 0.125
: ' 7 i . ;
0.25% < /} \ i F4). 0.25
P P

 

 

 

Figure 8. A schematic illustration for the dimension of a double notched 4-point bend test
specimen. Linear dimensions given are in inches.

Chapter III

RESULTS AND DISCUSSION
3.1 Microstructure Characterization

Three dimensional optical micrographs of the composite materials used in this study
are shown in Figure 9. Microstructural anisotropy and reinforcement clustering are noted,
partially, along the extrusion direction. The 10% SiCp composite shows the smallest
particle size distribution with an average particulate size of ~5 um, while 20"/o (A1203)p
reinforced MMC reveals the largest particle size distribution with the average particulate
size of ~18 um. The 10"/o (A1203)p composite shows the average particulate size ~ 9 mm
and 20% SiCp reinforced composite shows the average particulate size ~13 urn. The
distribution in particle size is shown in Figure 10 for the composite materials
investigated. Interfacial reaction products (appeared small dots) can be seen around the
reinforcements in (A1203)p composites (Figure 11a and 11b). Regarding the shape, SiCp
tended to be more angular with sharp comers compared to (A1203)p. Also, the frequency
of particulate clustering in SiCp reinforced composites appeared to be somewhat greater

along with extensive reaction at the particle/matrix interface (Figure 11c and 11d).

38

39

 

20V/o Al20J6061 Ai

L.

     

extrusion

extrusion

 

<—> <—>
a) 150m b) 150nm

 

extrusion

 

4——-> {—D
C) 150 um d) 150 um

 

 

 

 

Figure 9. Three dimensional optical micrographs of PRMMCs showing partial anisotropy and
reinforcement clustering along the extruded direction, a) 10"/o(Al203),/6061Al, b) 20%
(A1203),/6061Al, c) 10% SiCP/6061Al, d) 20% SiC,/6061Al

 

 

  

 

 

 

 

 

 

 

 

 

 

 

571'! » . . 5- — 41m 39m .
3 51.3 lOv/atAl,o,)y6()61 ... 36.912 3 34.33 _ ,_
U APS: 9.23 um g u
'1‘ 45.6! . '.- 32.31: 'I 3.40 .
39.” .. . .. - 28.71! 26.63
34.25 ... .... .. ,. ., .- 34.“ 22.93
28.3 ... ..... . .............3- 313 19_- ___
ma .. ...s 123; 11.48
11.48 E .. . . _ a.” 7.60 ..
- 4.10: am
an
. as .w .
m M It:
(a)
51.81:
8 10'/o SiC/6061 «.53: c
u APS:5.58um 3
'7‘ are: u
r

3.2?!
31.“
5.91!
an:
15.5%

 

 

 

 

 

 

 

 

(C) ((1)

Figure 10. Distribution in particle size and average particle size (APS) for the composite
materials, a) 10% (Ahoy/6061,41, b) 20%(Al203)/6061Al, c) 10% SiC/6061A], d) 20%
SiC/6061A].

41

 

 

10

 

 

 

 

 

Figure 11. Optical micrographs of PRMMCs, a) 10% (Al203),/6061Al, b) 20%(Al203),/6061Al,
c) 10% SiC/6061M, d) 20% SiC/6061M.

42

3.2 Interface Characterization
3.2.1 ALOJ6061AI Interface

SEM micrographs in Figure 12 show the matrix/reinforcement interface region of
the (A1203)p/6061Al composites. Interfacial reaction product phases are clearly seen in
the (A1203)p composite (Figure 12a). The irregular interface structure between (A1203)p
and the matrix alloy microstructure indicates that A1203 was partially consumed in
chemical reaction at the interface. Wettability and bonding at the A1203/matrix interface
are improved with the formation of spine] (MgAl204) [34,45,47]. Therefore, spine] is
thought to promote interfacial bonding strength by forming strong metallic and covalent

bonds with the matrix phase.

Electrochemically extracted reinforcement particles in Figure 12b and 12c show that
MgA1204 crystals formed on the A1203 particulates. The interfacial product, MgAl204,
forms when sufﬁcient contact between alumina and molten aluminum alloy prevails [45].
Coverage of alumina particulates with spine] (MgAl204) crystals (~ 1 um) is nearly
100%. Studies suggest that MgAl204 single crystals grow on the surface of alumina

particle via the following chemical reaction [34,46]:
4 2
Mg (1) + 3 A1203 (3) —) MgAl204 (s) + 3 A] (l). (4)

Figure 12d shows the individual octahedral-shape MgAl204 crystals. Also note that
the surrounding areas near individual spine] crystals are preferentially etched. Such
regions around the spine] crystals resulted from the rejection of elemental A] during the

interfacial reaction according to Equation 4 [34,46].

43

 

Argo, particle in 6061 AI Argo, particlgain 6061 Al

13.?

 

Spinei on a AIZO3 particle

 

 

 

 

 

Figure 12. SEM micrographs showing matrix/reinforcement interface region of the
(A1203),/6061Al composites, a) irregular interface structure between (141203)p and the matrix
alloy indicating a chemical reaction by partial consumption of (Al203),,, b) electrochemically
extracted reinforcement particles showing MgAl204 crystals formed on the A1203 particulates, c)
an alumina particulate covered nearly 100% with spine! (MgAl204) crystals (~ 1 pm), d)
individual octahedral-shape MgAl204 crystals.

44

Figure 13a shows a TEM bright ﬁeld image of the interface region indicating formation
of MgAl204 at the A1203/6061A] interface. The irregular interface between the MgAl2O4
and the A1203 particle depicted in the TEM image suggests partial consumption of the
A1203 particulate as revealed earlier (Figure 12a and 12c). Figure 13b and 13c represent
the convergent beam electron diffraction patterns (CBED) of MgAl204 spine] and A1203.

The CBED of the MgAl2O4 corresponds to the [001] zone diffraction pattern of FCC,

MgAl204 spine]. The CBED of A1203 particulates also corresponds to the [411] zone

diffraction pattern of or-Al203 with corundum structure.

45

 

 

6081 Al

Al203 '
particle

<———>
1 pm

 

 

 

 

 

 

Figure 13. TEM micrographs showing the interface of the (A1203)p/6061Al composites, a) bright
ﬁeld image of interface region showing formation of MgAl204 at the Al20/6061Al interface with
irregular interface between the MgAl204 and the A1203 particle, b) convergent beam electron
diffraction pattern on the MgAl204 spine! (FCC spine! structure), c) convergent beam electron
diffraction pattern on the A1203 particle (a—Al203 corundum structure).

3.2.2 SIC16061AI interface

SEM micrographs in Figure 14 show the matrix/reinforcement interface of
SiCp/6061Al composites. The arrows in Figure 14a indicate the formation of interfacial
reaction products. The black phase and short rod-like phase (arrows in Figure 14a point
to these phases) around the SiC particles are typically observed. The morphology of the
interfacial reaction product is revealed in more detail in Figure 14b and 14c showing
extensive formation of hexagonal platelets of aluminum carbide (Al4C3), which grew at
the SiC/6061A] interface. Tire hexagonal A14C3 platelets are on average ~0.7 urn thick
and ~3 um wide. Due to SiC particulate consumption during the interfacial reaction,

Al4C3 forms as result of the following reaction [37~42,52]:
4A] (1) + 3SiC (s) -> A14C3 (s) + 38i (s). (5)

A14C3 is a brittle and hydrophilic compound that can adversely affect the mechanical
properties of metal matrix composites [37~42,52]. Extensive formation of interfacial
A14C3 may result in the degradation of mechanical and physical properties, such as the
reduction in interfacial strength, fracture strain (8f), fracture toughness, and an increase in
corrosion sensitivity [38,39]. Previous investigations show that A14C3 forms mainly due
to prolonged contact between SiC and molten aluminum alloy [35,37,40,42], however,
A14C3 growth is observed also at moderately elevated temperature, ~450°C [41]. Figure
14d represents a the magniﬁed image of the hexagonal, A14C3 phases with some second
phase particles attached to its surface. The second phase particles residing on the A14C3
surface is rejected elemental silicon [4]] which is typical interfacial reaction in

accordance with Equation 5.

47

 

SiC particles in 6061 AI "SiC particles in 60st

»~..
la.--
. ... .I 2
, .r .
LI '3

,

 

 

 

 

 

 

Figure 14. SEM micrographs showing matrix/reinforcement interface of the SiC/6061.41
composites, a) arrows indicating formation of interfacial reaction products; dark and short rod-
like phase around the SiC particles, b) electrochemically extracted SiC particles, c) an SiC
particle showing extensive formation of hexagonal platelets of aluminum carbide (Al4C3)
growing at the SiC/6061M interface, d) magniﬁed image showing the hexagonal, Al4C 3 phases
with rejected elemental silicon on its surface.

48

Formation of Al4C3 can be mitigated by increasing the silicon content of the matrix
alloy. Addition of Si to the matrix alloy make the forward reaction of Equation 5 more
difﬁcult because rejected excess elemental Si will be increased beyond its equilibrium
content [42]. Other means of reducing the formation of A14C3 consist of limiting the time
in which SiC particulates and molten A] are in contact [38,40,43]. Also, coating SiC
reinforcement with oxides, such as SiO2 and/or Ti02 will block A14C3 nucleation and

growth [44].

3.2.30bservationolmolstureefiectoninbrfaces

Evidence of the moisture sensitivity of SiC/6061A] interface phase materials was
revealed further by the etching away of interfacial and near-interface phase materials
following immersion of metallographically-prepared specimens into de-ionized water of
48 hours (Figure 15). Figure 15a represents the unexposed specimen and Figure 15b
represents the specimen exposed to water for 48 hours. Clearly, Figure 15b shows the
dissolution of interfacial products near a cluster of SiC particulates as the pits have
formed. Using the ESEM, a more detailed assessment of the moisture effects was made
by regarding the change in the interface phase morphology of extracted particulates of
SiC and A1203 which were exposed to a water environment over several time intervals.
The ESEM images taken over an elapsed time period of 120 hours show conclusive
evidence of dissolution of the A14C3 phase (see Figure 16a,b,c). In contrast, there was
virtually no evidence of moisture induced degradation of the spine] phase (MgA1204)
which surrounds aluminum oxide particulate (see Figure 16d,e,f). Dissolution of

aluminum carbide in water is given as:

Al4C3 + 18H2O —) 4A1(OH)3 + 3CO2 + 12H2. (6)

49

 

 

 

 

 

 

Figure 15. SEM micrographs showing the moisture sensitivity of the SiC/6061A! interface, a)
metallographically prepared SiC/6061A! composite showing interfacial reaction products
(indicated by arrows) around SiC particles, b) the interfacial reaction products are dissolved
(indicated by arrows) by immersing into water for 48 hours showing pits.

50

 

 

 

 

 

 

 

 

 

Figure 16. Observation of moisture ejfect on the SiC/6061.41 and A1203/6061141 composite
interface using ESEM: a) as extracted SiC/6061Al; as electro-chemically extracted SiC particle
showing sharp edges in the hexagonal shape of A14C3 platelets, b) 65 hours moisture
environment of SiC/6061111; dull edges and cracks in swollen Al4C3 platelets indicating
dissolution of the interface reaction product, c)120 hours moisture environment of SiC/6061.41;
remarkably dissolved A14C3 platelets showing the hydrophilic property of the interface in 20%
SiC/6061111 composite. The SiC particle indicated by the arrow, however, remains the same
thorough out the moisture absorption, d) as extracted Al203/6061Al; as electro-chemically
extracted Al203 particle showing formation of MgAl204 at the interface, e) 65 hours moisture
environment of Al203/6061Al; no change on the interface, f) 120 hours moisture environment of
A1203/6061A]; no change on the interface.

51

Although the gases were not chemically analyzed, the fact that gas bubbles emerged from
the polished surface of a SiC/6061A] composite material when immersed in water
enhances the plausibility of the reaction given in Equation 6. It is interesting to note that
within 120 hours of exposure to water nearly all the faceted A14C3 crystals were severely
degraded by dissolution (see Figure 17). As found in this work, when water is in direct
contact with Al4C3, dissolution is quite rapid. It should be noted, however, that the rate
of A14C3 dissolution is mitigated somewhat for SiCp/6061Al composite itself because the

exposed surface area of A14C3 is less.

52

 

 

 

 

 

 

 

Figure 17. Magniﬁed images for observation of moisture ejfect on the SiC/6061M interface
using ESEM showing severe dissolution of hexagonal shaped Al4C3 platelets, a) as extracted
Al4C3 particles, b) 20 hours moisture absorption, c) 65 hours moisture absorption, d) 120 hours
moisture absorption (all micrographs are at the same magniﬁcation) .

53

3.3 Agng Kinetics

Figure 18 and Table 6 show average microhardness variation for several PRMMCs
and the complementary unreinforced matrix alloy as a function of aging time (a complete
compilation of the microhardness data table is given in the appendix). Accelerated aging
(corresponding to the reduced incubation time to reach the peak hardness) is observed for
all of the PRMMCs as relative to the unreinforced matrix alloy. The peak aging time is
reduced by ~l.5 hour for 10% (A1203)p/6O61Al composite and by ~6 hours for 20"/o
SiCp/6061Al composite. Accelerated aging has been widely observed in PRMMCs
[20,24~29,32], and, there is general agreement that the accelerated aging is due in part to
the mismatch in the coefﬁcient of thermal expansion (ACTE) between the matrix and the
reinforcement which results in an increase in the dislocation density and/or enhanced
diffusion. The strain ﬁeld generated by ACTE upon quenching is relaxed by the
formation of matrix dislocations. ACTE induced dislocations serve as heterogeneous
nucleation sites for strengthening precipitates. Short circuit of diffusion paths for solute
atoms are also enhanced [21]. Additionally, the elastic strain ﬁelds from residual stress
around the reinforcements can induce the enhanced diffusivity at relatively low

dislocation density and/or large particle size [25].

Composites with higher volume fraction of reinforcement (Vf) show higher
microhardness values for the solutionized and the as-quenched condition as well. Higher
strength results in composites with increasing Vf because of the enhancement in

dislocation density from reinforcement/matrix thermal expansion mismatch [32].

54

 

 

140

 

ﬁ

H (SiC)p20% / 6061 AI
M (SiC)p10% / 6061 Al

1 20 H (Ale.).20°/o I 6061 Al
M (AI20.).10°/. / 6061 Al

LH unreinforced 6061 Al

   
 
 
   

 

 

 

 

 

 

 

 

 

A
>
I
v
3
a: 100 -
C
'O
t—
to
.c
o
L.
.2
E

80

l
60 1 lglllngiL 1 1 1111111 1 1 1111111 1 L+Illll
100.0 ‘l 01.0 102.0 103.0 104.0

aging time (min.)

 

 

 

Figure 18. Microhardness variation for several PRMM Cs and the complementary unreinforced
matrix alloy as a ﬁtnction of aging time (aging temp. 180 °C). Accelerated aging is indicated by
the curve peaks shifting to the left.

55

Table 6. Average microhardness values for several PRMMCs during aging. Standard deviation
values are in parentheses. (Unit: Hv)

 

 

 

time 10"Io (A1203)P/6061Al 20'lo (A1203)p/6061Al 10'Io 20'lo
SiCp/6061Al SiCp/6061Al
1 87.1 (1.3) 104.2 (3.3) 90.6 (3.2) 95.9 (2.6)
30 87.9 (1.2) 105.2 (1.9) 91.0 (0.7) 97.0 (3.2)
100 93.6 (1.6) 108.4 (2.1) 98.9 (2.5) 100.9 (2.5)
200 95.3 (2.4) 110.6 (2.2) 106.6 (2.6) 109.2 (1.9)
300 101.0 (2.0) 111.2 (1.9) 110.0 (2.2) 118.6 (3.6)
400 104.8 (1.9) 115.4 (1.1) 121.4 (2.3) 126.6 (2.6)
550 111.2 (2.9) 117.6 (1.1) 132.0 (2.9) 137.0 (3.9)
700 113.4 (2.9) 121.2 (2.2) 133.8 (1.0) 133.0 (1.5)
850 118.0 (2.5) 119.6 (6.3) 129.8 (1.9) 131.0 (2.3)
1200 115.6 (1.9) 117.4 (2.1) 122.8 (1.5) 127.2 (3.7)
2000 109.0 (1.9) 116.2 (2.2) 117.4 (2.6) 120.6 (5.1)
3000 105.6 (4.0) 114.0 (2.5) 112.0 (2.1) 114.0 (2.4)

 

 

56

For similar reinforcement content, the SiCp reinforced MMCs exhibit signiﬁcantly
higher microhardness values than that of A1203 reinforced composites in the peak aged
condition. The hardness difference can be rationalized by noting the difference in
resultant matrix composition due to the different interfacial reactions between A1203 and
SiC reinforced composites. At the A1203/6061A] interface, magnesium is consumed and
the pure aluminum is rejected during the formation of MgA1204 according to Equation 4.
Since the major matrix strengthening precipitate in 6061Al-T6 is the [3 phase (MgZSi
precipitate) [19], the interfacial reaction results in a decrease in matrix strengthening
MgZSi precipitation near the reinforcement particles due to Mg consumption during
formation of the spine] phase. In contrast, aluminum is consumed and silicon is rejected
by the formation of A14C3 at the SiC/6061A] interface according to Equation 5. The
higher hardness of SiC/6061A] is presumed to be due to the presence of rejected Si

particles and the lack of interfacial reactions which consume matrix-hardening alloying

elements [33].

57

3.4 Fracture Behavior
3.4.1 Crack initiation

Crack initiation processes were observed by conducting 4-point bend testing on
double notched bar specimens. During loading of a 4-point bend specimen, catastrophic
fracture will occur at one of the two notches. The deformation/microfracture state of the
other notch, which exists just prior to catastrophic failure occurs, will be preserved.
Thus, optical microscope and SEM examinations performed in the vicinity of the notch
tip allow crack initiation for fracture to be assessed. Similar studies of crack initiation
events using 4-point bend specimens have been carried by Lewandowski et. al. [59].
Figure 19 shows the preserved notch tip for UA and 0A conditions of 20%
(A1203)p/6061A1 composite specimens. The microstructure of the notch tip of the UA
specimen after bend testing is depicted in Figure 19b. Extensive plastic deformation is
seen around the notch tip, at fractured particles, and at the sharp edges of the particles
which foster high stress concentration. Figure 20a shows the magniﬁed image of the
notch tip area in Figure 19b. The arrows in Figure 20a indicate the fractured particles
around the notch tip and the tendency for void initiation at sharp edges of the
reinforcements. By comparison, the preserved notch tip region of the 0A specimen
(Figure 19d) revealed no observable crack initiation and/or microfracture events. Void
growth and microcrack link-up occur by the matrix failure. No signiﬁcant damage
initiation at the preserved notch tip in the 0A condition suggests that no time lag between
microcrack and catastrophic main crack propagation. This type of crack initiation and
fracture behavior typify materials which possess low fracture toughness, tensile ductility

and strain to fracture (8f).

58

 

 

b) UA, after

 

 

      

‘ m

d) 0A, after ,.

 

 

 

 

 

 

Figure 19. Observation of crack initiation processes in 20% (A1203),/6061Al composite by
conducting 4-point bend testing on double-notched bar specimens, a) before the test (UA), b)
after the test (UA) showing extensive plastic deformation around the notch tip (a magniﬁed
view(5x) of area I is shown in Figure 20), c) before the test (0.4), d) after the test (0A), the
arrows indicating the void nucleation at the reinforcements and link up by matrix failure.

59

 

20.), Agog/6061 AI. UA 20v. SiC/6061 Al, OA

t? r
«5‘

 

 

 

 

 

Figure 20. Magnified images of notch tip area in PRMMCs from post-tested 4-point bend bar
specimens, a) magniﬁed image from Figure 19b, arrows indicating fractured particles around
the notch tip and tendency for void initiation at sharp edges of the reinforcements, b) magniﬁed
image from Figure 21d showing preferential crack propagation path through SiC, cluster
regions and SiC, clusters link up via matrix failure.

60

Aspects of crack initiation behavior at the notch tip for UA and 0A, ZOV/o
SiCp/6061Al are shown in Figure 21. Crack initiation occurs at the SiCp/matrix interface
for the UA specimen (Figure 21b). Similarly in OA condition, microfracture
preferentially occurs at the SiCp/matrix interface fracture. Also, the crack propagation
path tended to seek out SiCp cluster regions. As is common with other materials,
microcracks associated with SiCp clusters link up via matrix failure (Figure 20b and 21d).
It was quite apparent that preferential crack initiation favored the SiCp/matrix interface
and SiCp clusters regardless of the aging condition or reinforcement content. Preferential
void nucleation and microcrack initiation in the SiCp reinforced composite is seemingly
inﬂuenced by the weak bonding strength of SiCp/matrix interface due to formation of

interfacial A14C3.

61

 

notch root

 

 

icirioA, before ' ‘ » " d) 6A, after

 

 

 

 

 

Figure 21. Observation of crack initiation processes in 20% SiC/6061.41 composite by
conducting 4-point bend testing on double-notched bar specimens, a) before the test (UA), b)
after the test (UA) showing preferential crack initiation at the interface of the particles, c) before
the test (0.4), d) after the test (0.4) showing preferential crack propagation through SiC, cluster
regions, and link up of the microcracks by matrix failure (a magniﬁed view(5x) of area 11 is
shown in Figure 20).

62

3.4.2 Crack propagation

Crack propagation was observed on the sectioned and polished halves of post-tested
short bar fracture toughness specimens. Figures 22 and 23 show the SEM images for the
actual propagating crack in the PRMMCs and a corresponding illustrative drawings of the
crack path proﬁle and damage enclave consisting of fractured particles left in the crack
wake. The letters ‘X’ and ‘O’ of the illustrative drawings denote transparticle fracture
and particle interface debonding, respectively. The 20% (A1203)p/6061Al composite
(UA) in Figure 22 shows signiﬁcant amount of transparticle fracture in the crack enclave,
and in this case, particle fracture dominated the microfracture process. In contrast, for
20% SiCp/6061A1 composite (UA)in Figure 23 shows that interface debonding is the
dominant fracture mechanism. In addition, the size of the damage enclave associated
with crack propagation in SiCp reinforced composite is considerably smaller than for
A1203 reinforced composites. Since the damage enclave is associated with the extent of
energy absorption at the crack tip, a wider damage enclave around the propagating crack

should promote higher fracture toughness. Also, the size of the damage enclave is
commensurate to the size of the crack tip process zone (I. ). Thus, the fracture toughness

is expected to increase with l' as following equation [52,62]:

2

G, = Ef =a"ef‘ool'. (7)

 

Where GC is fracture energy of a composite, Kc is critical stress intensity factor, E is
elastic modulus of a composite, of is constant, a; is stress-modiﬁed fracture strain, and

00 is composite strength.

63

50h m

 

Figure 22. SEM images of the actual crack propagation in 20% (A1203),/6061Al composite (UA)
and a corresponding illustrative drawings of the crack path proﬁle and damage enclave
(indicated by dashed lines) consisting of fractured particles left in the crack wake (note that ‘X ’=
transparticle fracture, and ‘0 ’= interface debonding).

50 um

 

Figure 23. SEM images of the actual crack propagation in 20% SiC/6061M composite (UA)
and a corresponding illustrative drawings of the crack path proﬁle and damage enclave
(indicated by dashed lines) consisting fractured particles left in the crack wake (note that ‘X ’=
transparticle fracture, and ‘0’: interface debonding), a) 20"/o(A1203),/6061Al composite (UA),
b) 20% SiC/6061A! composite (UA).

65

The effect of aging on fracture mechanisms along the crack path is exhibited in
Figure 24 and Table 7 for the PRMMCs. 20"lo (A1203)p I6061Al composite shows that
transparticle-fracture is the dominant fracture mechanism irrespective of aging
conditions. However, transparticle fracture is slightly less for PA and 0A conditions.
The proportional increase in interfacial debonding or corresponding decrease in
transparticle fracture for the PA and 0A conditions suggests that changes in precipitation
and phase content affect the interfacial bonding strength [59]. Reduction of interfacial
bond strength due to precipitation [53] explains why no recovery in fracture toughness
occurs upon overaging PRMMCs. For 20% SiCp/6061Al composite, interface debonding
dominates the fracture process without signiﬁcant variation due to aging. Apparently, the
effect of aging on the microfracture mechanisms is masked by the inherently weak

interface as a result of A14C3 formation in the SiCp reinforced composites.

The difference in crack propagation behavior between A1203 and SiC reinforced
composites can be explained by the contrast in the interfacial reactions and the resultant
composition redistribution of matrix alloying elements. The interfacial reaction in
SiCp/6061A1 composites produces weak interfacial bonding strength combined with
higher peak hardness and lower ductility due to the alloying element redistribution. In the
SiCp/6061A1 composites, the crack can propagate without signiﬁcant damage around the
crack tip due to the low load transfer efﬁciency at the weak interface and low fracture
toughness of the matrix. However, extensive particle fracture in the crack wake occurred
for (A1203)pl6061Al composites, which suggested that the interfacial strength was

considerably higher.

 

 

 

 

 

 

Figure 24. Eﬁ’ect of aging on fracture mechanisms along the crack path for the PRMMCs. 20%
(Alloy/6061A! composite shows the transparticle fracture as dominant fracture mechanism
irrespective of aging conditions. 20% SiC/6061M composite shows the interface debonding as
dominant fracture made without signiﬁcant variation due to aging.

67

Table 7. Data for the eﬁ'ect of aging on fracture mechanisms along the crack path for the

 

 

 

 

PRMM Cs.
aging am total total through interface
conditions (mmz) particles fractured particle debonding
particle fracture
UA 0.157 269 1 15 79 36
(42.7%) (68.7%) (31.3%)
20 "lo PA 0.155 230 99 66 33
(A1203)p/6061Al (43.0%) (66.7%) (33.3%)
CA 0.151 236 101 62 39
(42.8%) (61.4%) (38.6%)
UA 0.103 362 132 37 95
(36.5%) (28.0%) (72.0%)
20 ”/0 PA 0.1 10 275 102 33 87
SiCp/6061Al (37.1%) (27.5%) (72.5%)
0A 0.1 19 274 91 26 65
(33.2%) (28.6%) (71.4%)

 

 

The SEM micrographs in Figure 25 show fracture of the crack tip area of the
PRMMCs. In Figure 25, the 0A 20% (A1203)p/6061A1 composite exhibits the extensive
particle cracking around the crack tip, while the PA 20% SiCp/6061Al composite shows
the fractured particles only at the crack path and no signiﬁcant damage around the crack
tip. Also, Figure 25b shows preferential crack propagation through SiC particulate
clustered regions and microcrack link up between the particulate clusters. Microcracks
that nucleated at SiCp clusters ahead of the crack tip grew in the composite material and
linked up with the main crack to sustain crack propagation. The SEM fractographs in
Figure 26 show typical fracture behavior of the particles in 20V/o (A1203)p/6061A1 (0A).
Figure 26a shows the morphology of the fracture surface which exhibits transparticulate
dominant fracture mechanism. Figure 26b shows cracked A1203 particles on the fracture
surface and Figure 26c shows transparticulate fractured particle with interface debonding.
The typical transparticulate fracture of an A1203 particle is shown in Figure 26d, and void
growth around the (A1203)p is quite evident. This type of microcrack nucleation by
particulate cracking tends to impede void coalescence by constraining plastic ﬂow in the
matrix, thus leading to reduced ductility and fracture toughness [53]. The SEM
fractographs in Figure 27 show typical fracture behavior of the particles in 20%
SiCp/6061A1 (OA). Figure 27a shows the morphology of the fracture surface which
exhibits interface debonding-dominant fracture mechanism. At higher magniﬁcation,
Figure 27b shows interface debonded SiC particles on the fracture surface and Figure 27c
shows the transparticulate fracture. Figure 27d shows the typical interface debonded SiCp
in the particle in 2070 SiCp/6061Al composite. Debonded SiC particles are abundantly
found on fracture surface for all aging conditions indicating that the interfacial debonding

was the dominant microfracture mechanism.

69

 

201/0 AIZO;,/6061Xl. OA' 20‘“/o SiC/6061 Al. PA

 

 

 

 

 

Figure 25. SEM micrographs showing the crack tip of the PRMMCs, a) crack tip of 20%
(Al203),/6061Al composite(OA) showing the extensive particle cracking around the crack tip, b)
20% SiC/6061M composite (PA) showing preferential crack propagation through SiC
particulate clustered region and microcrack link up between the particulate clusters.

70

 

gaffe-5156963643411. 6A

a”“‘
cracked Al 0
particles

 

\

ﬂ},

Intertaco (101)011de li‘iﬂQD'li'lIHll'th
( L L’ J l

\‘ fractured A'IHLO‘.’

\

"-' (rang jamculale
. heritage

. ; - <—————>
\ﬁ-yum _, 10 um

 

 

 

 

 

Figure 26. SEMfractography of 20% (Ahoy/6061111 composite(OA), a) morphology of fracture
surface, b) cracked A1203 particles, c) both transparticulate fractured and interface debonded
A1203 particle, d) transparticulate fractured A1203 particle.

71

 

20‘-‘/o SiC/6061 Al. 011

Y; Ameﬁa’Ce debonded SIC
particles

 

transparticulate

fractured SIC interface debonded

. SIC particle

\ \- r‘? 3‘“ ""13? \
ll \

 

 

 

 

 

Figure 27. SEM fractography of ZOV/o SiC/6061A! composite(OA), a) morphology of fracture
surface, b) interface debonded SiC particles, c) transparticulate fractured SiC particle, d)
interface debonded SiC particle.

72

3.5 Fracture Toughness

Figure 28 and Table 8 show the comparison of the fracture toughness (Kg) relative to
aging time. None of the composite materials exhibited recovery of fracture toughness
with aging. The fracture toughness of PRMMCs, unlike unreinforced matrix alloys,
typically do not recover on overaging [50,53,56,59]. Manoharan and Lewandowski [53]
observed that the toughness decrease between UA and 0A composite materials was due
to the transition in the local fracture mode from SiC particle cracking to SiC/matrix near
interface debonding. Increased interface debonding in OA condition was due to the
precipitation at SiCp/matrix interface that reduced the interfacial bond strength [53].
Klimowicz and Vecchio [56] suggested that the lack of recovery of fracture toughness of
over aged MMCs was due to the strain localization in matrix ligaments resulting from
fracture of the reinforcement particles. The ligaments showed premature fracture in OA
condition as secondary microvoids readily formed within the matrix ligaments due to the

presence of large second-phase precipitates.

Figure 28 illustrates that the fracture toughness decreases signiﬁcantly as the
reinforcement content and aging time (i.e. increased hardness and strength) increase for
both (A1203)p and SiCp reinforced composites. Fracture toughness for crack initiation
decreased linearly with the increase in the volume fraction of the reinforcement [53]. For
the UA condition, A1203 reinforced PRMMCs show considerably higher fracture
toughness values than that of SiC reinforced composites. However, the difference of the
fracture toughness values between A1203 and SiC reinforced composites is noticeably less
in the PA and 0A conditions. This difference in fracture toughness behavior as a result

of aging can be rationalized as follows. The low fracture toughness of SiC reinforced

73

composites in UA condition is due to the weak interfacial bonding strength resulting in
low load transfer efﬁciency. Considerably stronger interfacial bonding strength in A1203
reinforced composites in the UA condition results in more efﬁcient load transfer in the
ductile matrix, thus contributing to higher fracture toughness. With increasing aging
time, at the PA and 0A conditions, the fracture toughness of the A1203 reinforced
composites decreases due to the reduction of the matrix ductility and weakening of the
interface. For SiC reinforced composites, aging effects on fracture toughness are reduced

by the weak interfacial bonding strength resulting in smaller variation of the fracture

toughness upon aging.

 

74

 

 

underaged

 

(AL-0.). 10% r 6061 Al

(Ai.o.). 20% / 6061 Al

    
 

3O _. (SiC).10%/6061 Al
l (8.0). 2092. 6061 AI
. peak aged
. over aged
25 —

fracture toughness (MPa*m"2)

N
O
l

 

 

 

aging time (min.)

 

 

 

Figure 28. Comparison of fracture toughness (Kq) variation for the PRMMCs as a ﬁmction of
aging time (three specimens were tested to obtain each datum point).

 

Table 8. Fracture toughness(Kq) values for the various PRMMCs as a function of aging

75

conditions. (Unit: MPasl-rr—t , unit of values in parentheses: KsiJi; )

 

 

 

 

 

 

 

 

 

 

 

UA PA 0A

31.36 (28.53) 22.83 (20.77) 22.60 (20.56)

10'lo (A1203)p/6061A1 32.40 (29.48) 25.77 (23.45) 23.35 (21.25)

32.70 (29.75) 23.15 (21.06) 21.98 (20.00)

average 29.25 (32.15) 21.76 (23.91) 20.61 (22.65)
standard deviation 0.64 (0.70) 1.47 (1.62) 0.63 (0.69)

27.45 (24.98) 20.52 (18.67) 19.99 (18.19)

20'lo (A1103)p/6061A1 27.04 (24.60) 21.66 (19.71) 21.67 (19.72)

26.62 (24.22) 22.62 (20.58) 20.30 (18.47)

average 24.60 (27.04) 19.65 (21.60) 18.79 (20.65)
standard deviation 0.38 (0.42) 0.96 (1.05) 0.81 (0.90)

28.41 (25.85) 25.98 (23.64) 23.56 (21.44)

10'lo SiC,,l6061A1 27.88 (25.37) 23.04 (20.96) 23.98 (21.82)

29.15 (26.52) 25.04 (22.78) 22.41 (20.39)

average 25.91 (28.48) 22.46 (24.68) 21.22 (23.32)
standard deviation 0.58 (0.63) 1.37 (1.50) 0.74 (0.81)

23.99 (21.83) 20.73 (18.86) 19.35 (17.61)

20% SiCp/6061Al 21.35 (19.43) 21.56 (19.62) 20.41 (18.57)

........ 23.99 (21.83) 19.35 (17.61) 19.35 (17.61)

average 21.03 (23.11) 18.70 (20.55) 17.93 (19.71)
standard deviation 1.39 (1.52) 1.01 (1.12) 0.55 (0.61)

 

 

ChapterW

CONCLUSIONS

Accelerated aging is observed for all of the particulate reinforced metal matrix
composites. Higher hardness values were observed with higher Vf regardless of the
type, however, SiCp MMCs exhibited signiﬁcantly higher hardness than A1203

composites in the peak aged condition.

The major interfacial reaction products found were MgA1204 for (A1203)p/6061Al
composites and A14C3 for SiCp/6061A1 composites. Coverage of alumina particulates
with octahedral-shaped spine] (MgA1204) crystals (~ 1 pm) is nearly 100%. A
signiﬁcant amount of A14C3 was found at the reinforcement/matrix interface. The
morphology of the reaction product, Ath, was a hexagonal platelet shape resulting
in decrease of interfacial bonding strength. The instability of the A14C3 phase was
clearly exhibited as dissolution occurred when these crystals are in direct contact with
water. ESEM images showed that the disintegration of A14C3 occurred in less than

120 h when exposed to a moisture environment.

Preferential crack initiation occurred at the SiCp/matrix interface and near SiCp
clusters regardless of the aging condition or reinforcement content. Preferential
interfacial fracture of the SiCp composite is promoted by the weak SiCp/matrix
interface bonding strength due to A14C3 formation. In contrast, transparticle fracture

was the dominant fracture mechanism for (A1203)p/6061A1 composites irrespective of

76

77

aging conditions, as the MgA1204/(A1203)p interface is strong in comparison to the

SiCplAl4C3 interface.

The fracture toughness decreased with aging time, and no recovery of the fracture
toughness was observed with over aging. (A1203)p MMCs exhibited higher fracture

toughness than SiCp MMCs, when Vf remained constant.

Post-test examination of fracture toughness specimens revealed extensive fracture in
the wake of the propagating crack for (A1203)p/6061A1 composites. However,
substantially less particle cracking was observed in the crack wake for SiCp/6061Al
composites, as interfacial debonding was dominant. The damage enclave associated
with crack propagation and related to the ‘process zone’ in the (A1203)pl6061Al
composite was ~3 times as large as the damage enclave in SiCp/6061A1 composite

material. The composite material exhibiting a larger enclave revealed higher fracture

toughness.

9.

78

LIST OF REFERENCES

. M. M. Schwartz, Composite Materials Hand book, Heightstown, NJ, McGraw-Hill,

(1984) p. 670.
D. R. Flinn, The New Materials society vol. 2, Bureau of Mines (1990) p. 6-1

M. Hunt, “Aluminum Composites Come of Age,” Materials Engineering, Vol. 106
(1989) p. 37.

P. Rohatgi, “Advances in Cast MMCs,” Advanced Materials & Processes, Vol. 136
(1990) p. 39.

S. Caron & J. Masounav, “A literature Review of Fabrication Techniques of
Particulate Reinforced Composites, Fabrication of Particulate Reinforced Metal
Matrix Composites,” Eds. J. Masounave & F. G. Hamel, AMS Conference
Proceedings, Montréal, Québec, Canada (1990) p. 79.

Hideo Ohtsu, “Aluminum Alloys Matrix Composites Using Particle Dispersion,
Design & Manufacturing of Advanced Composites,” ASM conference proceedings
(1989) p. 187.

M. G. McKimpsom & T. E. Scott, “Processing and Properties of Metal Matrix
Composites Containing Discontinuous Reinforcement,” Materials Science and
Engineering, Vol. A107 (1989) p. 93

B. F. Quigley, G. J. Abbaschin, R. Wunderlin & R. Mehrabian, “A Method for
Fabrication of Aluminum-Alumina Composites,” Metallurgical Transactions A, Vol.
13A (1982) p. 93.

F. A. Girot, J. M. Quenisset & R. Naslain, “Discontinuously-Reinforced Aluminum
Matrix Composites,” Composite Science and Technology, Vol. 30 (1987) p. 155.

10. J. R. Davis, ASM Specialty Handbook; Aluminum and Aluminum alloys, Materials

Park, OH: ASM International (1993) p. 160.

l 1. T. J. Reinhart, Engineered Materials Handbook, ASM, Vol. 1 (1987) p. 904.

12. Duralcan® company, Data Manual, San Diego, CA: Duralcan® (1992).

13. A. L. Geiger & M. Jackson, “Low-Expansion MMCs Boost Avionics,” Advanced

Materials & Properties, Vol. 137 (1989) p. 23.

14. T. J. Reinhart, Engineered Materials Handbook, ASM, Vol. 4 (1987) p. 974

 

79

15. P. Rohatgi, “Cast Aluminum-Matrix Composites for Automotive Applications,”

Journal of Materials, Vol. 43 (1991) p. 10.

16. John E. Hatch, “Aluminum-properties and Physical Metallurgy,” ASM, (1984) p. 134

17. Metals Handbook, 10th edition, ASM International, Vol. 2 (1990) p. 102

18. W. F. Smith, “The Effect of Reversion of Treatments on Precipitation Mechanisms in

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

an Al-1.35 at. pct Mg2Si Alloy,” Metallurgical Transactions, Vol. 4 (1973) p. 2435.

H. J. Rack & R. W. Krenzer, “Thermomechanical Treatment of High Purity
6061A1uminum,” Metallurgical Transactions A, Vol. 8A (1977) p. 335

T. G. Nieh & R. F. Karlak, “Aging Characteristics of B4C-Reinforced 6061-
Aluminum,” Scripta Metallurgica, Vol. 18 (1984) p. 25.

M. Vogelsang, R. J. Arsenault & R. M. Fisher, “An In-situ HVEM Study of
Dislocation Generation at Al/SiC Interface in Metal Matrix Composites,”
Metallurgical Transactions A, Vol. 17A (1986) p. 379.

R. J. Aresenault & N. Shi, “Dislocation Generation Due to Difference between the
coefﬁcients of Thermal Expansion,” Materials Science and Engineering, Vol. 81
(1986) p.175.

T. Christman & S. Suresh, “Microstructural Development in an Aluminum Alloy-SiC
Whisker Composite,” Acta Metallurgica, Vol. 36 (1988) p. 1691.

J. M. Papazian, “Effects of SiC Whiskers and Particles on Precipitation in Aluminum
Matrix Composites,” Metallurgical Transactions A, Vol. 19A (1988) p. 2945.

I. Dutta & D. L. Bourell, “Theoretical and Experimental Study of Aluminum Alloy
6061-SiC Metal Matrix Composite to Identify the Operative Mechanism for
Accelerated Aging,” Materials Science and Engineering, Vol. A112 (1989) p. 67.

J. J. Stephens, J. P. Lucas & F. M. Hosking, “Cast Al - 7 Si Composites: Effect of
Particle type and Size on Mechanical Properties,” Scripta Metallurgica, Vol. 22
(1988) p. 1307.

S. Dionne, M. R. Krishnadev, & R. Bouchard, “Effects of Heat Treatment on the
Microstructure and Tensile Properties of a PIM SiC Particulate / 7091 Al MMC,”
Metal & Ceramic Matrix Composites: Processing, Modeling & Mechanical Behavior,
Eds. R. B. Bhagat, A. H. Clauer, P. Kumer & A. M. Ritter, TMS (1990).

E. Hunt, P. D. Pitcher & P. J. Gregson, “Precipitation Reactions in 8090 SiC
Particulate Reinforced MMC,” Scripta Metallurgica, Vol. 24 (1990) p. 937.

80

29. K. K. Chawla, A. H. Esmaeili, A. K. Datye & A. K. Vasudevan, “Effect of
Homogeneous Precipitation on Aging Behavior of SiCp I A1 2014 Composite,”
Scripta Metallurgica., Vol. 25 (1991) p. 1315.

30. P. Appendino, C. Badini, F. Marino & A. Tomasi, “6061 Aluminum alloy-SiC
particulate composite: a comparison between aging behavior in T4 and T6
treatments,” Materials Science and Engineering, Vol. A135 ( 1991) p. 275

31. H. L. Lee, W. H. Lu & S. L. Chan, “Effect of Cold Rolling on the Aging Kinetics of
A1203 / 6061A] Composite by Differential Scanning Calorimetric Technique,” Scripta
Metallurgica, Vol. 25 (1991) p. 2165

32. I. Dutta, S. M. Allen & J. L. Haﬂey, “Effect of Reinforcement on the Aging Response
of Cast 6061Al-A1203 Particulate Composite,” Metallurgical Transactions A, Vol.
22A (1991) p. 2553.

33. L. Salvo, M. Suéry, J. G. Legoux & G. L'Esérance, “Inﬂuence of Particle Oxidation
on Age-Hardening Behavior of As-Fabricated and Remelted SiC reinforced A1-1%
Mg Alloys,” Materials Science and Engineering, Vol. A135 (1991) p. 129.

34. J. C. Lee, K. N. Subramanian & Y. Kim, “The Interface in A1203 particulate-
reinforced aluminum alloy composite and its role on the tensile properties,” Journal
of Materials Science, Vol. 29 (1994) p. 1983.

35. J. P. Lucas, J. J. Stephens & F. A. Greulich, “The Effect of Reinforcement Stability
on Composition Redistribution in Cast Aluminum Metal Matrix Composites,”
Materials Science and Engineering, Vol. A131 (1991) p. 221.

36. S. Suresh, A. Mortensen & A. Needleman, Fundamentals of Metal Matrix
Composites, Stoneham, MA: Butterworth-Heinemann, (1993) p. 251.

37. T. A. Chemyshova & A. V. Rebrov, “Interaction Kinetics of Boron Carbide and
Silicon Carbide with Liquid Aluminum,” Journal of the Less-Common Metals, Vol.
117 (1986) p. 203.

38. D. J. Lloyd, H. Lagace, A. McLeod & P. L. Morris, “Microstructural Aspects of
Aluminum-Silicon Carbide Particulate Composites Produced by a Casting Method,”
Materials Science and Engineering, Vol. A107 (1989) p. 73.

39. G. Gonzalez, L. Salvo, M. Suéry & G. L'Esérance, “Interfacial Reactions In Al-Mg
Metal Matrix Composites Reinforced with (Sn/Sb) Oxide Coated SiC Particles,”
Scripta Metallurgica., Vol. 33 (1995) p. 1969.

81

40. H. Ribes, M. Suéry, G. L'Esérance & J. G. Legoux, “Microscopic Examination of the
Interface Region in 6061-AUSiC Composites Reinforced with As-Received and
Oxidized SiC Particles,” Metallurgical Transactions A, Vol. 21A (1990) p. 2489.

41. J. C. Lee, J. 1. Lee & H. 1. Lee, “Observation of Interfacial Reaction in SiCp / Al
composites and their characterizations,” In press.

42. T. Iseki, T. Karneda & T. Maruyama, “Interfacial Reactions between SiC and
Aluminum during Joining,” Journal of Materials Science, Vol. 19 (1984) p. 1962.

43. Ray Y, Lin, Sunil G. Warrier, Shyi K. Wu & Jein Y. Wang, “Infrared Aluminum
Matrix Composites and Carbide Formation,” Proceeding of ICCM-I 0, Whistler, BC,
Canada, August 1995, Vol II : Metal Matrix Composites, p. II-271.

44. L. N. Thanh & M. Suéry, “Inﬂuence of Oxide Coating on Chemical Stability of SiC
Particles in Liquid Aluminum,” Scripta Metallurgica., Vol. 25 (1991) p. 2781.

45. A. Munitz, M. Metzger & R. Mehrabian, “The Interface Phase in Al-Mg/A1203
Composites,” Metallurgical Transactions A, Vol. 10A (1979) p. 1491.

46. J. C. Lee, G. H. Kim, J. 1. Lee & H. 1. Lee, “Characterization of Interfacial Reaction in
(A1203)p/6061A1 Composite,” In press.

47. R. Molins, J. K. Bartout and Y. Bienvenu, “Microstructural and Analytical
Characterization of A1203-(A1-Mg) Composite Interfaces,” Materials Science and
Engineering, Vol. A135 (1991) p. 111.

48. B. F. Quigely, G. J. Abbaschina, R. Wunderlin & R. Mehrabian, “A Method for
Fabrication of Aluminum-Alumina Composites,” Metallurgical Transactions A, Vol.
13A (1982) p. 93

49. C. P. You, A. W. Thompson & I. M. Bernstein, “Proposed Failure Mechanism in a
Discontinuously Reinforced Aluminum Alloy,” Scripta Metallurgica., Vol. 21 (1987)
p. 181..

50. D. J. Lloyd, “Aspects of Fracture in Particulate Reinforced Metal Matrix
Composites,” Acta Metallurgica., Vol. 39 (1991) p. 59.

51. D. L. Davison, “Fracture Characteristics of Al-4 Pct Mg Mechanically Alloyed with
SiC,” Metallurgical Transactions A, Vol. 18A (1987) p. 2115.

52. J. P. Lucas, P. K. Liaw, J. J. Stephens & J. Nunes, “The Effect of Particulate Fracture
in Determining Fracture Toughness of Metal Matrix Composites,” Proceedings of

53.

82

Morris E. Fine Symposium, Eds. P. K. Liaw, J. R. Weertman, H. L. Marcus & J. S.
Santner, TMS (1991) p. 171.

M. Manoharan & J. J. Lewandowski, “Crack Initiation and Growth Toughness of an
Aluminum Metal-Matrix Composite,” Acta Metallurgica., Vol. 38 (1990) p. 489.

54. M. Manoharan & J. J. Lewandowski, “In-situ Deformation Studies of an Aluminum

55.

Metal-Marix Composite in a Scanning Electron Microscope,” Scripta Metallurgica.,
Vol. 23 (1989) p. 1801.

S. V. Nair, J. K. Tien & R. C. Bates, “SiC Reinforced Aluminum Metal Matrix
Composites,” International Metals Reviews, Vol. 30 (1985) p. 275.

56. T. F. Klimowicz & K. S. Vecchio, “The Inﬂuence of Aging Condition on the Fracture

57.

58.

59.

61.

62.

Toughness of Alumina-Reinforced Aluminum Composites,” Fundamental
Relationships Between Microstructure & Mechanical Properties of Metal-Matrix
Composites, Eds. P. K. Liaw & M. N. Gungor, TMS (1990) p. 255.

P. M. Mummery, B. Derby & C. B. Scruby, “Acoustic Emission from Particulate-
Reinforced Metal Matrix Composites,” Acta Metallurgica., Vol. 41 (1993) p. 1431.

J. Yang, C. Cady, M. S. Hu, F. Zok, R. Mehrabian & A. G. Evans, “Effect of Damage
on the Flow Strength and Ductility of Ductile Al Alloy Reinforced with SiC
Particulates,” Acta Metallurgica., Vol. 38 (1990) p. 2613.

J. J. Lewandowski, C. Liu & W. H. Hunt Jr., Powder Metallurgy Composites, Eds. M.
Kumar, A. Ritter & K. Vedula, TMS-AIME, Warrendale, PA (1988) p. 117.

.W. A. Logsdon & P. K. Liaw, “Tensile, Fracture Toughness and Fatigue Crack

Growth Rate Properties of Silicon Carbide Whisker and Particulate Reinforced
Aluminum Metal Matrix Composites,” Engineering Fracture Mechanics, Vol. 24
(1986) p. 737.

Y. Flom & R. J. Arsenault, “Effect of Particle Size on Fracture Toughness of SiC/Al
Composite Material,” Acta Metallurgica., Vol. 37 (1989) p. 2413.

RD. Ritchie & A. W. Thompson, “On Macroscopic and Microscopic Analyses for
Crack Initiation and Crack Growth Toughness in Ductile Alloys,” Metallurgical
Transactions A, Vol. 16A (1985) p. 233.

 

83

APPENDIX

Microhardness values for several PRMMCs during aging. (Unit: Hv)

 

’AR. ‘srr. 30 100 200 300 400 550 700 850 1200 2000

3000

 

 

 

 

 

 

 

 

43.1 89.2 88.4 93.9 93.3 98.9 106.0111.0114.0120.0115.0107.0 102.0

10 V/o(A1203)p 41.4 86.4 88.4 96.2 93.3 102.0 102.0 109.0 111.0 115.0 117.0 111.0 106.0

/6061A1 43.1 86.1 86.9 92.4 94.5 100.0 107.0 108.0 113.0 118.0 118.0 108.0 101.0

45.9 86.3 86.4 92.4 98.9 104.0 104.0 113.0 118.0 116.0 113.0 108.0 110.0

45.4 87.6 89.4 93.3 96.6 100.0 105.0 115.0 111.0 121.0 115.0 111.0 109.0

Average 43.8 87.1 87.9 93.6 95.3 101.0 104.8 111.2 113.4 118.0 115.6 109.0 105.6
.St. Dev. 1.9 1.3 1.2 1.6 2.4 2.0 1.9 2.9 2.9 2.5 1.9 1.9 4.0

70.7 98.9 108.0 110.0 114.0 111.0 115.0 117.0 120.0 126.0 116.0 115.0 115.0

20 v/o(A1203)p 72.9 106.0 104.0 107.0 111.0 109.0 116.0 118.0 120.0 122.0 119.0 119.0 114.0

/6061Al 70.7 103.0 106.0 108.0 110.0 112.0 117.0 119.0 125.0 120.0 117.0 114.0 110.0

65.0 107.0 103.0 111.0 108.0 114.0 114.0 116.0 120.0 121.0 120.0 118.0 117.0

68.6 106.0 105.0 106.0 110.0 110.0 115.0 118.0 121.0 109.0 115.0 115.0 114.0

Average 69.6 104.2 105.2 108.4 110.6 111.2 115.4 117.6 121.2 119.6 117.4 116.2 114.0
St. Dev. 3.0 3.3 1.9 2.1 2.2 1.9 1.1 1.1 2.2 6.3 2.1 2.2 2.5

53.6 90.7 89.2 100.0 106.0 109.0 117.0 131.0 134.0 130.0 119.0 120.0 112.0

10 ‘70 49.0 89.4 92.0 101.0 107.0 114.0 122.0 133.0 132.0 128.0 122.0 115.0 114.0

SiCp/6061Al 50.0 90.7 89.6 98.8 104.0 108.0 121.0 131.0 137.0 132.0 123.0 117.0 111.0

52.5 90.7 89.2 94.5 106.0 109.0 125.0 133.0 133.0 130.0 124.0 119.0 106.0

57.2 91.3 94.9 100.0 110.0 110.0 122.0 132.0 133.0 129.0 126.0 116.0 117.0

Average 52.5 90.6 91.0 98.9 106.6 110.0 121.4 132.0 133.8 129.8 122.8 117.4 112.0
St. Dev. 3.2 0.7 2.5 2.6 2.2 2.3 2.9 1.0 1.9 1.5 2.6 2.1 4.1

64.8 96.6 101.0 103.0 110.0 127.0 123.0 138.0 132.0 126.0 125.0 113.0 113.0

20 "/o 57.9 92.4 92.8 98.0 111.0 123.0 126.0 142.0 129.0 130.0 127.0 118.0 109.0

SiCp/6061Al 66.1 94.0 98.0 104.0 106.0 120.0 130.0 137.0 132.0 127.0 131.0 123.0 113.0

67.7 98.4 94.9 99.3 109.0 118.0 128.0 131.0 132.0 125.0 121.0 123.0 110.0

67.5 98.0 98.4 100.0 110.0 125.0 126.0 137.0 133.0 130.0 128.0 126.0 115.0

Average 64.8 95.9 97.0 100.9 109.2 122.6 126.6 137.0 131.6 127.6 126.4 120.6 112.0
St. Dev. 4.0 2.6 3.2 2.5 1.9 3.6 2.6 3.9 1.5 2.3 3.7 5.1 2.4

 

*A.R.—) As Received, S.T.—) Solution Treated, St. Dev. —) Standard Deviation.

 

"lllllllillllllli“