. M3“.
_ “Ms-"...
”w .5

., g

..‘n

..:

u: .

r. . \.1
k.:.
. (

H.“ . 1.
...m a“... n

~\:.n:-.l~ . ~ -u‘I

v . ~.; ‘. ~ . ‘

.‘ , n, ‘- ‘

n V - '7“ : “A5 I» v
‘ ‘ a- . . L 'h .ufl

V V ‘ ~ u~':~-f~

:n.“

.. . .

«21.
‘I‘ “Finns-,4.
-7..- .-. ““41:
...:g . » .-.“, rd ‘1:
n..-~,~>..H\-. x ‘n
I {V a I ..Jl.
. ‘ , ‘1 .
I. . - 1 |. ~ ~\.\. -\

1‘1: x.“ v - ' - W ,. "wwugsu,

)‘ ‘ ' . - V u‘m; ""'?""‘!

> 1-- ,1' ‘31:.

r .
4r:

soar.” .z , n

.x ...«.v mfg-yak. «

m, |.:n~:,..-.-‘,.rg.

1,. ",pL-‘Iu 4“ -,

z u
«i.

d... ,
,u a.

.. ....
Ira-{w ,- .
...,.. .
.5175, ' ..
, 4...», ,,.,w...r
v”. A,” ’.3 . .
. 4 _ >

M : :1.

fn’:r"I'F
q rem-1..

 

a... “in-...: y'—

 

21% ”I? 6590‘”

 

LIBRARY 7"
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

The effect of Rolling on the Redistribution of

SiCp Clusters and its Influence on the Mechanical
Pr0perties of 10% SiCp/606l-A1 Composites
presented by

Jae-chul Lee

has been accepted towards fulfillment
of the requirements for

Master's of Science degree in Material Science

L(git/QMWW‘"W M‘

Date7/QS/8C/

 

 

 

 

O—7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

Ml CHI IGAN STATE

III I II III

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

III/IIIIIIIIIIIIIIIIIIIIIIII

00605 0557

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

DATE DUE DATE DUE DATE DUE

    
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P

MSU Is An Affirmative Action/Equal Opportunity Institution

'THE EFFECT OF ROLLING ON THE REDISTRIBUTION 0F SiCp
CLUSTERS AND ITS INFLUENCE ON THE MECHANICAL PROPERTIES OF

10 % SiCP/60619A1 COMPOSITES

by Jae-Chul Lee

A THESIS
submitted.to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Metallurgy, Mechanics and Materials Science
1989

 

O
r“
<I’
0.
d
o
9

ABSTRACT

Warm rolling in a direction perpendicular to the
direction of banding in SiCp/6061 Al-alloy composites
having a severely banded SiCp clusters causes significant
redistribution of SicP within the composites. The
mechanical properties were found to become more isotropic
with increased percent of reduction. The strength and
elongation parallel to the rolling direction were found to
increase with increasing reduction ratio up to 50-60%. On
the other hand the properties perpendicular to the rolling

direction were found to decrease slightly with increasing

_reduction ratio. Such phenomena are analyzed based on

metallographic and fractrographic examinations.
Possible mechanisms operative in the strengthening of
SiCp/Al composites are considered. Failure mechanism of

SiCp/Al composites under uniaxial tension is proposed.

Acknowledgement

I would like to take this oppertunity to sincerely thank
my advisor, Dr. K. N. Subramanian, for his constant support
and guidance during the research. Without his
encouragement and help this study would have been far from
reality.

I would also like to thank Dr. A. Hingwe and R. Vaidya
for supporting me the material used for the research.

I would like to take this opportunity to extend my most
sincere gratitude to my parents, brother, and sister for
“their sacrifice. Last but not the least, a special thanks

tn: my wife, Chae-Hong, for her endurance and assistance.

TABLE OF CONTENTS

List of Figures ....... IV
List of Tables ....... v“
1. INTRODUCTION ....... 1
2.

STRENGTHENING MECHANISM OF SiCp/Al COMPOSITES

2.1 Orowan Strengthening ....... 7
2.2 Composites Strengthening ....... 12
2.3 Thermal Strain Hardening ....... 15
2.4 Strengthening due to subgrain size ....... 13

EXPERIMENTAL PROCEDURE

3.1 Material ....... 23
3.2 Specimen Preparation ....... 23
3.3 Tensile Test A .......34
3.4 Fractrographic and Metallographic Examination --36

RESULTS and DISCUSSION
4.1 Microstructural Features ....... 37
4.2 Mechanical Properties .......49

4.3 Failure Mechanism of SiCp/Al Composites
Under Uniaxial Tension

000000.70
CONCLUSIONS

5.1 Strengthening Mechanism of SiCp/Al Composites ...35
5.2 Microstructural Features .......86
5.3 Mechanical Properties .......37

5.4 Failure Mechanism of SiCp/Al Composites

Under Uniaxial Tension .......33

REFERENCES ........9

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.
Figure 9.

Figure 10.

Figure 11.

Figure 12.

IV.

LIST OF FIGURES

Fractographs showing the dimple size in
a) unreinforced aluminium alloy
b) aluminium alloy reinforced with
10 % sic ........... 19
P
The correlation between steady-
flow stress and mean subgrain diameter
of commercial aluminium for various
deformation conditions ........... 20

Metallograph illustrating the grain size of
a) unreinforced aluminium alloy
b) aluminium alloy reinforced with

10 % SiCp ........... 22
a) Grid analysis
b) Line intercept method ........... 24

Three dimentional view of as-extruded
10 % SiCp/6061-Al composite ........... 26

Porosities of the as-extruded SiC /Al
composite viewed parallel to p
the extrusion direction ........... 27

a) Section of the extruded bar stock
cut out for specimen preparation
b) Schematic of procedure used

for making specimens ........... 3]
Dimensions of sheet tensile specimen ....... 33
Details of the grips ...... .....35

Micrographs illustrating the initiation (X)
and propagation (Y) of edge cracks

along the clustered masses of Sic

during rolling process .... ....... 33

Micrographs illustrating

the crack propagation along

the cracked particles (a)

and debonded interfaces (b)

during rolling process ........... 39

a) Particulate cracking
in 40 % cold rolled composite
b) Abscence of particulate cracking
in 60 % warm rolled composite .......... 42

V

Figure 13. a) Micrograph exhibiting the severely

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

14.

15.

18.

19.

banded SiCp clusters of

as-received composite
Micrograph exhibiting the redistribution
of sic clusters in
b) 40 E warm rolled composite
c) 50 % warm rolled composite
d) 60 % warm rolled composite
e) 70 % warm rolled composite ........... 43

Growth of pores during cold rolling
a) 70 % warm rolled

b) 70 % cold rolled ... ........ 46
The comparision of the pores size

between

a) as-received, and

b) 70 % warmed roled composite ........... 47

Load-elongation plot exhibiting
dynamic strain aging behavior ........... 5]

a) Variation of the yield strength
of T0 tempered composite
with respect to reduction ratio
b) Variation of the yield strength
of T6 tempered composite
with respect to reduction ratio ........ 56

a) Variation of the U.T.S.
of T0 tempered composite
with respect to reduction ratio
b) Variation of the U.T.S.
of T6 tempered composite
with respect to reduction ratio ...... °-58

Variation of the fracture elongation
of To and T6 tempered composites
with respect to reduction ratio ...........60

Optical micrograph illustrating
debonded interfaces .... ....... 61

Particulate cracking

a) SEM micrograph

b) optical micrograph. Note that the

matrix is squeezed into the broken
particulate in regions indicated

by arrow.

Particulate cracking in

40 % cold rolled composite

Note the elongated grains

in the rolling direction ........... 62

C

v

Figure

Figure

Figure

Figure
Figure

Figure

Figure
Figure
Figure

Figure

Figure

22.

25.

26.

VI

Grain size of

a) as—received composite (T6)

b) 30 % warm rolled composite (T6)

c) 70 % warm rollde composite (T6) ......... 64

Tensile fracture surface exhibiting
large amounts of locallized
plastic deformation ........... 67

Fracture surface of the
longitudinal specimen ........... 68

Debonded interface and
particulate cracking of
tensile fracture surface . .......... 72

Stress concentration
at the pole of the inclusion
during uniaxial loading ........... 73

Particulate cracking in
a tensile specimen ........... 76

Stress concentration
at the equator of the inclusion
during uniaxial tensile loading ........... 78

Visualization of the stress—
concentration near the cavity
under uniax1al tension ........... 79

The schematic diagram showing
the failure mechanism of SiCp/Al

composites under uniaxial tension ......... 32

SEM micrograph showing crack development
on a tensile specimen ........... 33

SEM micrograph showing void formation
due to joining of cracks ...........34

 

Table

Table

Table

Table

 

VII

LIST OF TABLES

Slected properties of the 6061-Al and Sic
used in this thesis .......... 9

Variation in yield strength (in MPa)

with respect to the reduction ratio

along the transverse and the longitudinal
directions ..........52

Variation in tensile strength (in MPa)

with respect to the reduction ratio

along the transverse and the longitudinal
directions .......... 53
Variation in fracture elongation (in %)

with respect to the reduction ratio

along the transverse and the longitudinal
directions ..... .....54

 

1. INTRODUCTION

Since the early 1960s, with the impetus of high
temperature structural applications, various kinds of metal
matrix composites have been investigated by incorporating
high strengh ceramic materials such as alumina(A1203) and
silcon carbide(SiC) whiskers, and boron carbide(B,C)
particulates into molten metals [1,2].

Among many of those metal matrix composites, continuous
graphite fiber/copper, continuous graphite fiber/aluminum
alloy, SiC(whisker and particulate)/aluminum alloy,
A1203(particulate)/ magnesium, SiC(particulate)/magnesi0m
were particularly promising for structural applications.
But the cost of ceramic renforcements prevent the
composites from the practical applications inspite of their
attractive mechanical properties to the engineers and
designers.

In 1973, as the new technology for making fi-Sic whisker
by pyrolizing rice hull was developed, silicon carbide
‘whiskers could be made much cheaper, finer, and purer than
previous ones.

Since then, especially during the past ten years,
silicon carbide reinforced aluminum matrix composites have
‘been studied extensively in many areas, due to their
potentials in automotive, structural, and aeronautical

applications.

Many of these studies carried out so far deal with
.continuous fiber and discontinuous whisker reinforced
composites. Although these composites are as light as
aluminum and its alloys, they possess a significant
improvement in strength and stiffness [3-9], fatigue
resistance [10‘12], damping capacity [13],and wear
resistance [14,15], in addition to high temperature
properties [15-17]. Particularly, strength and stiffness
of SiC particulate reinforced Al-alloy matrix composites
(SiCp/Al composites) are comparable to titanium and its
alloys [18,19], and enable them to replace titanium
forgings. For example, the ultimate tensile strength and
the stiffness of silicon carbide fiber reinforced aluminum
matrix composites are much higher than those of
unreinforced titanium even at 500°C [19].

However, the main drawback of these continuously
reinforced composites are their severe anisotropy in
mechanical properties. In addition, it is difficult to
fabricate, and shape them into their final configurations.
On the other hand, SiCp/Al composites can be manufactured
relatively easily and economically by melting and casting
techniques. For example, near—net shapes of track shoes
and pistons can be produced by using squeeze casting method
[20]. The basic principle of the squeeze casting is to
’forge' a liquid metal into a closed die to reduce the
porosity due to the shrinkage and gas, and to make the

products solidify rapidly under high pressure of 50 to 100

MPa. In order to significantly increase the volume
fraction of reinforcement without flocculation, Rheocasting
(compocasting), which consists of vigorously agitating a
semisolid composites before casting [21] can be effectively
used.

Moreover, unlike polymer matrix composites and
discontinuous fiber reinforced composites, which are
-usually formed into the final shapes, metal matrix
composites containing particulate reinforcements can be
shaped into their final configurations by using the
conventional mechanical workings such as forging,
extrusion, and rolling, etc. Such components as
automobile-engine connecting rod and compressor blade can
be forged from bar stock that was extruded from a cast
billet [22].

Such mechanical workings to obtain the final shape will
alter the size, shape, and distribution of clustered
particulates. All these parameters have significant
influences on the mechanical properties of such a composite
system.

Recently, it has been found that nearly all commercially
important ceramic reinforcements including silicon carbide
have a poor wettability by molten aluminium and its alloys
[23-25]. As a result of the poor wettability of silicon
carbide by molten matrix alloys, the direct incorporation
of SiC into molten aluminum alloys causes flocculation.

P
Under such conditions, extensive clustering or

 

agglomeration of the particulates occurs in order to reduce
surface tension. This effect becomes more significant as
the particulate size becomes smaller than 40 pm [27,28] and
the difference in density between the matrix alloy and
reinforcements becomes larger [15].

Such clustering not only cause poor overall mechanical
properties and machinability [28], but also cause
anisotropy of particulate reinforced composites in their
as-manufactured state and prohibit the wider use of these
composites for practical applications. If the distribution
of the SiCp can be made more uniform, the properties of
these composites can become more isotropic.

A number of different techniques have been used to
overcome this problem. The following methods have been
found to be very effective in preventing significant
particulate clusterings or agglomerations.

1) Matrix modification by adding some alloying elements,
such as Li or Mg, has been proved to be effective in
improving the wettability of the reinforcements by the
molten matrix [14,27,29—32]. These alloying additives are
added just before incorporating the reinforcements into
molten matrix.

2) Preheating the reinforcements before introducing into
molten matrix allows their uniform distribution in the
matrix [31,33].

3) The use of particulate coatings, such as Ni or Cu

coatings on graphite or alumina particles, has been tried

out successfully and thereby improves the wettability by
the molten aluminum alloys [34].
Mechanical working of the composites can be used.as

another means for making the SiC to be uniformly

P
distributed. Rolling has been used very effectively in
hypoeutectoid steels to break up pearlite lamella and to
obtain a semi-spherodized struture. The principle behind
this process is the differece in the ductilities of the
ferrite and pearlite. During rolling, the pearlite phase
tends to break up due to its lower ductility. The SiCp/Al
composites are very similar to hypoeutectoid steels in this
point of view. The aluminum matrix has much higher
ductility as compared to the brittle SiCp clusters. Hence,
it is expected that rolling can break up the clusters of
the SiCp and thereby change the mechanical properties of
the composite system.

Very few investigations have studied the effects of
mechanical working on the composites so far [6,18,35,37].
However, various investigations have been focussed on the
charecterization of the interface [37-40], mechanical
properties at room and elevated temperatures, and on

various manufacturing processes.

The objectives of the present study are

1) to investigate the effects of warming rolling on the
size and distribution of the Sicp clusters and its

influence on the mechanical properties

2)

3)

4)

to evaluate the variation in the mechanical
properties of the SiCp/Al composites after different

amounts of deformations and heat treatments.

to identify the strengthening mechanisms operative in

SiCp/Al composites

to propose a failure mechanism of SiCp/Al composites

under uniaxial tension.

M

2. STRENGTHENING MECHANISM OF SiCp/Al COMPOSITES

The addition of ceramic reinforcement in the forms of
fibers, whiskers, and particulates into metal matrix can
lead to significant increases in yield strength and elastic
modulus as compared to those of the matrix, while there are
some negative effects on the mechanical properties such as
decreases in ductility and fracture toughness.

Although the theories for the strengthening mechanism of
the SiCp/Al composites have not yet been completely
estabilished, some plausible strengthening mechanisms, such
as Orowan theory, composite strengthening from modified
shear lag theory, thermal strain hardening due to enhanced
dislocation density, and strengthening due to smaller
subgrain and grain, can be used to explain the
strengthening of discontinuously reinforced SiC/Al
composites.

Therefore the expected yield strength of the SiCp/Al

composites can be expressed as

acy = any + AU --------- 1)
where ”cy = yield strength of the composites
”my = yield strength of the matrix
A0 = increased amount in yield strength
= A"orowan'+ A”comp-I'M disloc + A” sub gr.

A increased amount in yield strength due to

a orowan =
dislocation looping-
Aadisloc = increased amount in yield strength due to

the enhanced dislocation density

Aasub gin: increased amount in yield strength due to
the formation of subgrain
Accomp = increased amount in yield strength due to
load transfer
The principle idea for calculating A0 is the

comp
application of the continnum mechanics to the composite

strengthening. The'externally applied load is transferred
from the matrix (Al alloy) to the reinforcement (SiCp) via
the SiCp/Al interface. The strengthening depends on the
interfacial bonding strength and the efficiency of the load
transfer from the matrix to the reinforcement i.e, shape of
the reinforcement.

The material properties used in this thesis are listed

in Table.1.

 

 

Property

 

 

 

 

 

v.3 UTS 5f E V a
_1 Ref
Material (MPa) (MPa) (°/o) (GPa) (mm-°K)
TO 55 124 25-30 68.3 0.33 28x 10'5 41
BOBIAI _
T6 235 257.5 12 74 0.33 28x10 6 41
SIC 1380 — — 438 0,19 3 x 10-6 42

 

 

 

 

 

 

 

 

Table 1. Slected properties of the 6061-A1 and 81C

Note

used

: 1. T6 : Solution treatment :

in this thesis

0
529 C * 70 min.

0
R.T. aging : 24 C * 48 hours
0
Artificial aging : 200 C * 10 hours

2. cf:
3. E :

4.11:

Fracture elongation

Elastic modulus

Poisson’s ratio

5. a : Thermal expansion coefficient

 

 

 

IO

2.1 Orowan strengthening

Since most of the SiC exists in the form of particles,

it may be possible to propose the Orowan strengthening

mechanism.

The Orowan-Ashby equation [43] or looping of

precipitates is given by

A1

where Ar

t

V

P

[0.81-p0b]/[2n°(1-y)2 oDz]o1n(2r°/b) ---- 2)

increased amount in resolved shear stress due to

the precipitates

shear modulus of the matrix (28 GPa for 6061 Al
-T6 treated

Poisson’s ratio (0.33)
mean particulate radius
Burger’s vector (2.86*10-8cm)
effective interparticulate spacing [44]
(Lt/v )1/2
P
length of particulate

thickness_of particulate

volume fraction of the particulate

Generally, the increased amounts in yield strength (A0)

is expressed as

A0 =

M 0 Ar ------- 3)

where M is approximately 2.

 

The evaluated increases in yield strength (A0) is only
about 2 MPa, while the observed increases in yield—strength
for T6 tempered composites is 67 MPa.

It is obvious from eq—2) that the amount of the
strengthening due to the Orowan mechanism is strongly
dependent on the interparticulate spacing. However, the

interparticulate spacing between the SiC in the composites

P
is usually too large for diSlocations to be bowed between
the SiCp during deformation. Hence, in this point of view,
it is considered that the Orowan strengthening mechanism

contributes very little to the strengthening of the SiCp/Al

composites.

 

 

2.2 Composite Strengthening

In 1952, Cox developed the shear lag model to predict
the yield strength of short fiber reinforced composites
[45]. The most important assumption in the shear lag model
is that the load transfer occurs only between the fiber and
matrix by means of shear sterss at the fiber-matrix
interface. This theory can be used successfully for
predicting the yield strength of composites having large
aspect ratio reinforcement.

However, underestimation in yield strength are expected
for the composites having small aspect ratio such as
whisker or particulate reinforced composites, because the
normal load transfer at the whisker and particulate ends
and side surfaces was ignored in the shear lag theory.

In 1986, Nardone and Prewo proposed a modified shear lag
model to explain the strengthening of discontinuously
reinforced composites having small aspect ratio
reinforcements such as SiCp and Sij [46]. Although this
theory took into account of the normal load transfer at the
whisker and particulate ends, the normal load transfer at
the side surfaces of the whisker and particulate was still
ignored.

The modified shear lag theory for particulate reinforced
composites gives the composites yield strength (ay) by the

following equation[46]:

 

I3

0y = amy .[1 + (L1+t)os/4L1]on+ amy-V m ------- 4)
where ”my = yield strength of the matrix
= 235 MPa for 6061 Al-T6
L1 = length of the particulate perpendicular to the
applied stress
t = thickness of the particulate
Vp = volume fraction of the SiCp
Vm = volume fraction of the matrix
= 1 - Vp
s = particulate shape factor = 2L/t
L = particulate length in the tensile direction.

Assuming le t gives

a =

o + e + e . ---------
y amy [1 s/2]Vp a v 5)

my m
Therefore, the increases in yield strength due to load

transfer (Aacom is

A"comp" y my

= amy o[(1+s/2)-vp+ vm- 1]

any . (s/2) .v

p. --------- 6)

According to Eq-6), once the particulate shape factor 5 is
fixed, the increases in composite yield strengthAacomp

depends only on the volume fraction of reinforcement or

vice versa. Moreover, Aa comp has no relationship with the

14

strength of the reinforcement. Appropriate values of s(=3)

and Vp(=0.1) give Aacomp:= 35.3 MPa.

On the other hand, Aa can also be expressed in terms

comp
of interparticulate spacing (D) by the following procedure:

s = particulate shape factor
= ZL/t -------------- 7)
D = effective interparticulate spacing

«Hint/VP) -------------- 3)

Solving eq-7) and eq-8) for ‘5’ gives

2 ' 2
s = (2D Vp)/t . -------------- 9)
Substitutinon of eq-9) into eq—6) gives the relationship
between A0 comp and interparticulate spacing (D)
- v t 2 D2 -----
Aacomp""my°( p/') . 10)

Therefore, for a fixed volume fraction of the
reinforcement, if the average size of the reinforcement
becomes finer, the interparticulate spacing will decrease
by eq-8), thereby the composites yield strength is expected

to increase substantially.

15

2.3 Thermal strain hardening

Arsenault and Fisher proposed a strengthening mechanism
which can be explained by an enhanced dislocation density
which is caused by the large difference (10:1) in the
thermal expansion coefficient between the Al matrix and
SiCp in the composites [47]. A high dislocation density of
log-lomcrfl'2 ‘was observed experimentally during cooling
down from the annealing temperature to the room temperature

[48]._ The increase in the shear stress(Ar) due to the

presence of dislocations can be expressed as

Ar = a’opobo/z- -------- ----- 11)
where a’ z 0.5
p = shear modulus of matrix (28 GPa for 6061 Al-T6)
-8
b = Burger’s vector (2.86*10 cm)
p = dislocation density of the matrix

Therefore, the following equation may be used for

estimating the increase in yield strength (Aa c) due to

dislo
the enhanced dislocation density:

A0 diSlO = a'fl'b'( 5']; ) ------------- 12)

where a = 1.25 for Al [49]

p = dislocation density of the matrix of the SiCp/Al

composites

 

 

p0 = dislocation density of the matrix in the absence

_6 _2
of SiCp (=10 cm for annealed Al).
Since p is much bigger than p0, eq-12) can be reduced into
A”disloc = a'“'b"/p_

= 1.25p-b'/p_ . _____________ 13)

In 1986, Arsenault and Shi developed an equation for
evaluating the dislocation density by using the model of
"the prismatic punching of dislocations" [50]. From their

analysis, the dislocation density in the matrix was found

to be
p = [B-Vp-AeJ/[(l-Vp)-b-d] ------------- 14)
where B = a geometric constant between 4 and 12
= 8 (for particulate reinforcement)
V = volume fraction of reinforcement

Ac = misfit strain due to the difference in the

thermal expansion coefficient

= AaOAT
b = Burger’s vector
d = the smallest dimension of the particulate

Substitution of eq-14) into eq-13) results in the final
expression for the increases in yield strength due to the

enhanced dislocation density.

A"disloc: 3-54wb't (Vp~Aa~AT)/((1-vp)ob-d) 1 .-- 15)

17

For a fixed volume fraction of reinforcement, from eq-15)
it is expected that Aa disloc W111 increase Significantly by
incorporating the smaller size of reinforcement. For
Vp=0.1 and d=10pm, at least a 22 MPa increase in yield
strength is expected due to the enhanced dislocation

density.

 

 

 

18
2.4 Strengthening due to subgrain

A smaller grain and subgrain size as compared to those
of the wrought Al alloy can also lead to a significant
increase in yield strength.

Since the interface between SiC and A1 matrix act as a
source of dislocation generation, small subgrains are

expected to be formed near SiC due to tangling of

P
dislocations. Arsenault and Fisher observed a high
dislocation density of lolocm' and small subgrains of 1.5-
2.5 pm in size in an SiC/Al composites during cooling from
annealing temperature in an in situ experiment with a
transmission electron microscopy (TEM) [47]. On the other
hand, the fracture surface of the wrought Al alloy reveals
relatively large subgrain of 5 pm in size [48,50]. Such
results can also be verified by examining a fracture
surface of the composites with a scanning electron
microscopy (SEM), since the dimple size almost corresponds
to the subgrain size [48,51]. Fig.1) shows the dimple
sizes of the reinforced and the unreinforced Al alloy.

A is equal to 29.4 MPa for an average subgrain size

”sub gr.
of 2.5 pm observed in Fig.2).

The Hall-Petch relation can be a possible candidater for
the strengthening of SiCp/Al composites, because the

incorporation of SiC into the molten A1 alloy and the

P
rapid solidification of the composite system obstruct grain

growth during cooling from fabrication temperature.

I9

Figure 1. Fractographs showing the dimple size in

a) unreinforced aluminium alloy
b) aluminium alloy reinforced with 10 % SiCp

 

 

 

Figure . 1

20

Figure 2. The correlation between steady flow stress and
mean subgrain diameter of commercial aluminium
for various deformation conditions

Compression [52]
Extrusion [53]
Creep [54]

Hot torsion [55]

ODD.

. gear-.7

 

 

 

 

 

20_ I I I I I I I . ,
I8— - E
.. . 2
Is— -
7314- ‘ 94.6
It)? ~ d
..Iz- - 81.1
b . __ .-
810" “ 67.6
g ' .
(I?) 8' i543
{>51 6' ~ 40.5
u_ - -
4 I- “ 27.0 I
L
2 309 — 13.5
DA Ii
' I .‘ I I I I I I

 

12.34.567.89
RECIPROCAL SUBGRAIN DIAMETER, d-I (pt-I

Figure . 2

.K‘

21

The grain sizes between the reinforced and the unreinforced
Al alloy are compared in Fig.3).

Therefore, due to above considerations, at least 89 MPa
of theoretical increases in yield strength is expected with
the incorporation of 10% SiCp. This gives the lower bound
value for a theoretical yield strength of 324 MPa without
including the strengthening due to small grain size (Hall-
Petch relation). However, the measured yield strength of
the 10%—SiCp/6061-Al composites was only 302 MPa in their
longitudinal direction.

The mismatch between the theoretical and measured yield
strength seemed to be due to the following factors :

1) Random orientation of some of the SiCp with respect

to the tensile direction

2) Pre-existing porosity (or void) in the matrix

3) Pre-existing debonding at the interface

4) Particulate free-zone that may results in big sub-
grain
5) Particulate cracking from fabrication i.e during

extrusion process

 

22

Figure 3. Metallograph illustrating the grain size of

a) unreinforced aluminium alloy
b) aluminium alloy reinforced with 10 % Sicp

 

22-A

 

Figure . 3

23

3. EXPERIMENTAL PROCEDURE

3.1 Material

6061 aluminium alloy reinforced with 10 % (by volume)
of SiCp, procured as an extruded cylinderical bar of
diameter 64.3 mm from ALCAN, was used for this study. The
matrix of the composites was characterized by using EDAX
after solutionizing.

The line intercept method and the grid analysis [56]
were used to determine the volume fraction of the SiCP in
the SiCp/Al composites. The schematic drawings illustrating
these methods are shown in Fig.4-a) and Fig.4-b). The
following relation was assumed for the line intercept

method and the grid analysis [57].

 

where Vf

= volume fraction of Sicp

.............. l6)

<Af> = average area fraction
= ratio of the area which is covered by Sicp to

the total area tested

<Lf> = average line fraction

= ratio of the length of the line which is

intercepted by SiCp to the total length of the

test line

 

 

24

Figure 4. a) Grid analysis
b) Line intercept method

 

 

 

 

 

Figure . 4

 

 

25
<Pf> = average point fraction
= ratio of the number of points which lie in the
siC to the total number of points in the

P
grid.

In order to investigate the effect of warm rolling on
the distribution of SiCp, the composites with a severely
banded structure of SiCp clusters in the direction of
extrusion were used for the tests. A three dimentional
view of the as-extruded composites is shown in fig.5).

The amount of the pre-existing porosity (Fig-6) was
determined as 1-2 % from the grid analysis.

The average particulate size was found to be 10 pm with

the help of the line intercept method.
d = L/(n.M) --------- 17)

where d = the average particulate size
L = the length of the test line
n = the number of intersections with Sicp and test
line which is drawn in the micrograph concerned,
and

M = the magnification.

26

Figure 5. Three dimentional View of as-extruded
10 % SiCp/606l-Al composite

 

26—A

 

Figure . 5

27

Figure 6. Porosities of the as-extruded SiCp/Al composite

viewed parallel to the extrusion direction

27—A

 

Figure . 6

 

 

28

3.2 Specimen Preparation

The stock material was cut out from a cylindrical bar
which was obtained in the as-extruded (as-manufactured)
state, and rolled unidirectionally into different
percentages of reduction in a direction perpendicular to
the extruded direction, namely, transverse direction.

The reduction ratio (R) can be defined as the reduction

in thickness from the assumption of plane strain condition.

R= [ (Ao -Af )/A0 1x100 [as]
= [ ( to - tf )/to ] x 100 [95] ---—----.. 18)

where A0 = initial cross sectional area of the specimen
Af = final cross sectional area of the specimen
t0 = initial thickness of the specimen, and

tf = final thickness of the specimen.

In order to maximize matrix flow and therefore minimize
the particulate cracking and the possible interface
debonding, warm rolling was carried out at a temperature
below 400°C. At the same time, a high reduction ratio, but
less than 10% of reduction per each pass, was used to get
homogeneous matrix flow, which was found to be effective in
separating the SiCp clusters into separate particulates.

After each pass of rolling, the specimens were annealed

0
for 5- 7 min. at 415 C in order to get rid of the

 

 

 

29

dislocations which were induced due to rolling process.
This was checked by measuring the hardness, i.e, regardless
of the reduction ratio the value of HRBZO was obtained
after annealing.

The rolled sheets which were reduced by different
percentages of reduction, were cut into parallel
(longitudinal) and perpendicular (transverse) to the
extruded directions with a diamond sawing machine. The
schematic drawing is shown in Fig.7).

The sheet tensile specimens were made by hands with
files. The specimens were polished with abrasive papers
and rotating laps in order to remove the contact surface
with rolls during rolling process. The final shape and
dimensions of the tensile specimen is shown in Fig.8).

Two kinds of heat treatments were used before testing
the specimens, i.e, full annealing (T0) and artificial
aging (T6). Especially for the full annealed specimen, the
test was carried out within 30 min in order to prevent the
precipitation from the matrix even at room temperature.

Yhe heat treatment conditions used can be described as

following:

1) Full annealing ( To temper )
-In order to get the softest codition, the specimens
were heated at 415°C for 2 hours then quenched into
cold water.

2) Artificial aging ( T6 Temper )

 

 

 

3O

0
-solution treatment : 530 C for 70 min.
(cold water quenching)
0
-room temperature aging : 24 C for 48 hours

0
—artificial aging : 200 C for 10 hours

 

31

Figure 7. a) Section of the extruded bar stock
out out for specimen preparation

 

 

 

 

 

Extrusion direction
(Longitional direction)

Rolling.direction
(Transverse direction)

Figure.7-a)

32

Figure 7. b) Schematic of procedure used for making
specimens

 

I

 

%

Figure.7-b)

 

 

Cutting slices
——-———->

 

33

Figure 8. Dimensions of sheet tensile specimen

W = Width of the specimen ( 5 mm )
L = Gauge length ( 17 mm )
T = THickness ( 0.7 mm )

 

34

3.3 Tensile Test

The basic mechanical properties after various
treatments of the specimen were measured using an Instron
machine operated at a constant crosshead speed of 0.1
cm/min. The tests were carried out in a laboratory air
enviornment at room temperature.

In order to prevent the slipping of the specimen from
the grip, slices of file with small grooves were attached

to the grips as in Fig.9).

35

Figure 9. Details of the grips

35-A

 

Senew- hate (on

darnping Apeobnen

 

 

 

F4191)

 

 

 

 

 

 

Figure . 9

 

36

3.4 Fractographic and Metallographic Examination

The fracture surface of tensile specimens were examined
by scanning electron microscope (SEM). In addition,
metallographically polished surface of a mounted specimen
was prepared to examine the cracking of SiCp which can be
caused due to compressive and tensile force.

The redistribution of SiCp clusters due to rolling were
examined with optical microscope.

The standard method of surface preparation consisted of
successive surface removal by 240, 400, and 600 grit Sic
abrasive paper, first and second polishing on rotating laps
with 5 pm and 1 pm alumina powder, and finally first and
second stage polishing on rotating laps with 1 pm and 0.25
pm diamond powders.

The Kellers Reagant was used for slight etching. Dilute

HCl was used for deep etching of specimen surface.

 

37

4. RESULTS AND DISCUSSION

4.1 Microstructural Features

One of the advantages of the discontinuously reinforced

metal matrix composites is their formability into the net

 

shapes by conventional mechanical working or by plastic
deformation methods.

Both cold and warm rolling were attempted on as-extruded
(as-received status) composites along a direction
perpendicular to the extrusion direction, until edge cracks
were observed on the specimens. Although controlled cold
rolling was carried out (i.e 1-2 % reduction in thickness
per each pass), edge cracks and surface scuffings were
formed before 40 % reduction was reached. Such edge cracks
and surface scuffings appeared to be formed mainly due to
the large clusters of Sicp which usually have micro-voids
inside the clusters. The initiation and propagation of the
edge cracks resulting from cold rolling can be observed in
Fig.10). Once these cracks are formed, they tend to
connect the SiCp clusters and grow larger. Some of the
cracks were observed to propagate along the broken SiC and
the initially debonded interfaces as in Fig.11).

During cold rolling, the hardness of the composites
increases very rapidly. For example, 30 % of reduction

gives rise to an increase in hardness from HRBzo (T0

38

Figure 10. Micrograph illustrating the initiation (X)
and propagation (Y) of edge cracks

along the clustered masses of SiC during
rolling process p

38—A

 

Figure.10

39

Figure 11. Micrographs illustrating the crack propagation
along the cracked particles (a) and
debonded interfaces (b) during rolling process

 

Figure.11

 

.I.
..r

o
s

 

 

.
.
- ..
, ,

 

'. ‘ . ‘n -r
‘ 7‘
J ... ‘
d
.. ...
“ CR
4
K
..
1k.
.
\
r.

 

 

40

tempered state) to HRBSS. Such an increase in hardness
makes further rolling impossible without an additional
stress relief annealing.

The rolled sheets of SiCp/Al composites were cut
parallel to the rolling direction, and examined
metallographically on the polished sections. Significant
particulate crackings were observed from the polished
sections of the cold rolled SiCp/Al composites. There is
strong tendency for the crack planes to be parallel to the

direction of compression (i.e rolling pressure) and to be

perpendicular to the rolling direction(i.e feed direction).

At the same time, it was found that the average size of
the SiCp that breaks is generally larger than the average
size of all SiCp in the composites. Such observations
correspond to the results of Gurland [58] who had studied
the fracture of cementite particles in a spherodized 1.05
carbon steel, deformed under various loading conditions.
He observed that the cementite particles, that cracked
under uniaxial compression, were bigger than the average
particle size, and that the particle cracking occuredalong
a direction parallel to the compressive force. Such
formations of particulate cracking can cause decrease in
strength and ductility of the composites. The substantial
increase in rolling pressure during cold rolling, as
compared to that of warm rolling, was found to be
responsible for such crackings of the SiCpin the

composites. This comparison is shown in Fig.12).

o
6

 

41

On the other hand, in case of warm rolling, the specimen
could be rolled down to as much as 85 % of reduction
without any edge crackings or surface scuffings. .Although
relatively large reduction ratio per each pass of rolling
(about 6-7 %) was used, the first edge cracks made their
appearence at about 85 % reduction. The results indicate
that SiCp/Al composites possess an excellent warm
formability. Another advantage of warm rolling is that
particulate crackings and interface debondings can be
minimized (Fig.12), since the flow stress decreases
monotonically with increasing temperature for both the
reinforced and unreinforced material [36].

Both cold and warm rolling were found to cause a
significant change in the distribution and shape of the
SiCp clusters. The most apparent difference in metallo-
graphic features between the as-extruded and the rolled
composites is the presence of banded structure of SiCp
clusters. The microstructural features of the as-extruded
and the warmed rolled composites with different reduction
ratios are shown in Fig.13).

There appear to be no significant differences in the
microstructural features of the distribution of SiCp
clusters between the cold and the warm rolled composites;
however, it is evident from Fig.14) and Fig.15) that the
cold rolled specimens exhibited the presence of larger
voids and debonded interfaces as compared to the warm

rolled specimens at the same reduction ratio (Fig.14).

 

 

42

Figure 12. a) Particulate cracking in 40 % cold rolled
composite
b) Abscence of particulate cracking
in 60 % warm rolled composite

 

 

42—A

 

Figure. 12

 

 

43

Figure 13. Micrograph exhibiting'
a) the severely banded SiCp clusters
of as-received composite
b) the redistribution of SiC clusters

P
in 40 % warm rolled composite

43-A-

111 g
‘4‘, ‘3'

73".”4r47 ~'
I 3". " ‘ ~ 2
. t‘ t. _

A
.5'

 

 

Figure 13.

44

(continued)

Micrographs exhibiting the redistribution
of SiCp clusters in

c) 50 % warm rolled composite
d) 60 % warm rolled composite

 

44-A

”if". “g“
1155'

 

Figure . 13 (continued)

45

Figure 13. (continued)

Micrograph exhibiting the redistribution
of SiCp clusters in

e) 70 % warm rolled composite

 

 

45-A

 

Figure . 13 (continued)

46

Figure 14. Growth of pores during cold rolling

a) 70 % warm rolled
b) 70 % cold rolled

46-A

 

Figure. 14

 

47

Figure 15. The comparision of the pores size between

a) as-received, and
b) 70 % warmed roled composite

 

47-A

 

Figure . 15

 

48

Moreover, these voids grow larger with increasing reduction
ratio, as can be observer in Fig.15). Such features can b
explained on the basis of enhanced plastic flow during warm
rolling.

Hence, in order to facilitate matrix flow and to
minimize either fracture of the SiCp or debonding at the
interface, plastic forming at higher temperature is
recommended. Nevertheless, particulate cracking and
interface debonding are observed even in the as-extruded
SiCp/Al composites, although they are not significant as

compared to those formed during cold working process.

 

49

4.2 Mechanical properties

Some mechanical properties of the SiCp/Al composites in
their longitudinal and transverse directions, were measured
at room temperature in the as-received and the warmed
rolled status, both after T0 and T6 heat treatments.

Significant changes in mechanical properties compared
with those of unreinforced Al-alloy, were observed for the
6061-Al alloy reinforced with 10 % of SiCp. For example,
compared with the wrought alloy, the SiCp/Al composites
revealed higher yield and ultimate strength, but lower
fracture elongation. Also, compared with the T0 tempered
SiCp/Al composites, the T6 tempered SiCp/Al composites were
much higher in yield and ultimate strength but
substantially lower in fracture elongation.

From the tensile testing of the SiCp/Al composites in
the longitudinal and transverse directions, it was observed
that as-extruded composites exibited anisotropic mechanical
properties, such as yield strength, ultimate strength, and
fracture elongation, with respect to their longitudinal and
transverse directions. It can be surmized from the Fig.10)
that the microstructural inhomogeneity of the as-extruded
composites may attribute to such anisotropic mechanical
properties. On the contrary, the microstructure of the warm
rolled composites becomes much more uniform (homogeneous)
with increasing reduction ratio, which may result in more

isotropic mechanical properties.

 

 

50

During tensile testing of the T0 tempered SiCp/606l-Al
composites with a cross-head speed of 0.1 cm/min, a strong
dynamic strain hardening was observed just beyond the yield
point of the composites. The interaction between the
impurity atoms and the dislocations is reported to cause
such an aging phenomenon during deformation [59]. It is
also reported that dynamic strain aging tends to occur over
a wide range of temperature,which depends on the strain
rate [60]. The typical plot of load vs. elongation
obtained with the Instron testing machine is shown in
Fig.16). However, such strain aging phenomenon did not
occur in T6 tempered composites, since oversaturated solute
atoms are precipitated out during artificial aging

0
procedure (T6) at 200 C for 10 hours.

Although a significant redistribution of the SiCp
clusters was achieved with increasing reduction ratio of
warm rolling, the principal effects of the warm rolling on
the mechanical properties are to decrease the yield and
tensile strength as well as the fracture elongation in the
longitudinal direction. Some mechanical properties
obtained at room temperature are presented in Tables 2, 3,
and 4.

From the tables, it is evident that the as-extruded (as-
received) composites have the largest ductility but the
lowest strength, and the T6 tempered composites have the

highest strength but the lowest ductility.

_I-gdf..- .

 

 

 

 

51

Figure 16. Load-elongation plot exhibiting
dynamic strain aging behavior

LOAD

 

 

ELONGATION

Figure. 16

52

 

 

 

 

 

 

 

 

 

 

 

 

 

Specimen; T O T 6
R W L T L T
0 124.2 115.7 302.0 269.5
20 — - - -
30 _ - 288.7 281.6
40 — - - -
50 119.3 119.0 285.1 282.1
60 119.1 119.04 283.2 283.0

 

Table 2. Variation in yield strength (in MPa)

with respect to the reduction ratio
along the transverse and the longitudinal
directions

 

 

53

 

 

 

 

 

 

 

 

 

 

 

Specimen T O T 6
R (°/o) L T L T
0 231.1 201.2 368.6 323.0
20 _ .. _ _
3O - - 356.2 339.6
40 — - - —
50 222.2 222.4 351.5 342.0
60 220.6 1' 222.1 350.0 347.2
70 214.3 207.0 — -
80 209.0 207.0 — _

 

 

 

 

 

Table 3. Variation in tensile strength (in MPa)

 

with respect to the reduction ratio
along the transverse and the longitudinal
directions

 

54

 

Table 4. Variation in fracture elongation (in %)
with respect to the reduction ratio
along the transverse and the longitudinal
directions

...‘Iru k1"...:".: '

 

 

 

'56

Figure 17. a) Variation of the yield strength
‘ of T0 tempered composite
with respect to reduction ratio

YIELD STRENGTH [MP0]

320—
310—

I
300—
290—
280—
2701
260~

250

J

 

56—A

 

0 Transverse (T6)
0 Longitudinal (T6)
I I r fir f I I I I I r I 'l

10 20 3O 4O 50 60 7O
REDUCTION RATIO [z]

Figure.17-a)

 

 

57

Figure 17. b) Variation of the yield strength
of T6 tempered composite
with respect to reduction ratio

YIELD STRENGTH [MP6]

57—A

 

 

140]
130—
ikl\\
120‘ NT
9%” """ "”
l 10—
GF-O Transverse (TO)
o—e Longitudial (TO)
100 I . I ' I ' l . I . I . l
O 10 20 30 40 50 50 7O

REDUCTION RATIO [z]

Figure. 17-b)

58

Figure 18. a) Variation of the U.T.S. of T0 tempered
composite with respect to reduction ratio

 

U.T.S [MP6]

2005

 

240~

2301

220—

210—1

190-

 

180

58-A

 

0 Transverse (T0)
0 Longitudinal (TO)
I l I I I

lrr‘l'lrl'l

IO 20 30 4O 50 60 7O 80

REDUCTION RATIO [2;]

Figure. 18 a)

59

Figure 18. b) Variation of the U.T.S. of T6 tempered
composite with respect to reduction ratio

 

U.T.S [MP6]

380—
370;,
360;
350;
340;

330—

V

c
3209

310—

 

59—A

o transverse (T6)

 

 

300

0 Longitudinal (T6)
T I l I I l I I I r 1 r 1

30 4O 50 60 7O 80
REDUCTION RATIO [2;]

Figure. 18 b)

60

Figure 19. Variation of the fracture elongation
of T0 and T6 tempered composites
with respect to reduction ratio

FRACTURE ELONGATION [73]

—L
—L
l

C)
. l l

 

 

60-A

 

 

 

 

 

2_ 0 Transverse [T6]
- I Longitudinal [T6]
1 -4 0 Transverse (TO)
' 0 Longitudinal (TO)
0 I T ' I T T ' I *T I ' I ' T ‘
O 10 20 3O 4O 50 60 7O 80

REDUCTION RATIO [7,]

Figure.19

61

Figure 20. Optical micrograph
illustrating debonded interfaces

 

61—A

 

Figure. 20

62

Figure 21. Particulate cracking

a) SEM micrograph

b) optical micrograph.
Note that the matrix is squeezed into the
broken particulate in regions indicated
by arrow. I

 

62—A

- ‘ Rolling Direction . "
.,..: ' ‘ - I ' - r,

 

Figure. 21

 

Fin n'-- -

 

 

 

 

63

Figure 21. (continued)

c) Particulate cracking in 40 % cold rolled

composite
Note the elongated grains in the rolling

direction

 

63—A

 

Figure . 2 1 (continued)

64

Figure 22. Grain size of

a) as—received composite (T6)
b) 30 % warm rolled composite (T6)

 

64-A

     

I
IJ'K/w K 0

tr . . h'g ./ ‘.y
4m thing “3\ I25 ImL| a

 

Figure.22

 

 

65

Figure.22. (continued)

c) 70 % warm rollde composite (T6)

65—A

 

Figure. 22

66

Examination of the typical fracture surface of the
SiCp/Al composites, shown in Fig.23), reveals a large
amount of localized plastic flow in the matrix surrounding
each Sicp. Particulate pull-out and interface debondings
were rarely observed from the fracture surfaces of the
tensile specimens. Such a result indicates that the

bonding strength between sic and Al matrix is very strong.

P
According to the investigation on the measurement of the
interfacial bonding strength of the SiCp/6061 Al -
composites, the lower bound value of the interfacial
bonding strength for the composites was determined as 1690
MPa [61], which is 30 times higher than the yield strength
of T0 treated ( fully-annealed ) 6061 Al alloy and 7
times higher than the yield strength of T6 treated
(artificially aged) 6061 Al alloy.

Some particulate cracking, which is formed perpendicular
to the direction of rolling, was observed from the fracture
surface of tensile specimen prepared by cutting along
longitudinal direction [Fig.24].

On the other hand, the strength and the elongation in
the transverse direction were observed to increase after
early stage of warm rolling (up to 50 - 60 % of reduction )
-and decrease afterward.

Both the redistribution of the Sicp clusters and the
formation of the preferential crack initiation sites (such

as interface debonding, particulate cracking, and porosity)

can be attributed to the changes in the mechanical

67

Figure 23. Tensile fracture surface exhibiting
large amounts of locallized
plastic deformation

67—A

 

Figure . 23

Figure 24.

68

Fracture surface of the
longitudinal specimen

Particulate crackings are formed
perpendicular to the rolling direction.

Note : "A" particulate has no debonded
interface.
"B" particulate has debonded interface
which seems to be formed due to rolling.

68-A

 

Figure . 24

69

properties in the transverse direction. The redistribution
of the Sicp clusters, which results in more uniform
microstructure, appeared to be the main cause for the
improvement in the strength and the elongation. But the
formation of the preferential crack initiation sites is
considered to be the dominant factor for the decrease in

such mechanical properties.

 

70

4.3 Failure mechanism of the SiCp/Al composites

under uniaxial tension

The failure mechanism of the SiCp/Al composites under a
uniaxial tensile load is analyzed in this section.

From the results of the tensile test, it is clear that
the addition of moderate amount of SiCp ( usually less than
30 % by volume) into Al matrix can result in significant
increase in the strength, and presumably the elastic
modulus, of the composites.

However, the addition of SiCp into Al-alloy results in
substantial decrease in the ductility, and consequently the
fracture toughness of the composites, a main drawback for
the wide use of most metal matrix composites reinforced
with ceramic reinforcements, even at low volume fraction
levels.

Recently, several studies have been carried out to
improve the ductility and the fracture toughness of ceramic
reinforced metal matrix composites. Nevertheless, large
differences in the properties of the matrix and
reinforcement cause large differences between reinforced
and unreinforced alloys. Such big discrepancies in the
ductility and the fracture toughness are considered to be a
consequence of the unique failure mechanism of the
discontinuously reinforced metal matrix composites. Such a
behavior is exactly opposite with the failure mechanism of

ceramic matrix composites.

 

 

71

However, significant studies have not been carried out
so far on the failure mechanism of the SiCp/Al composite
system, or discontinuously reinforced metal matrix
compositets.

In 1986, Nutt and Duva [62] have carried out TEM studies
on SiC whisker reinforced Al alloy. They observed that the
void nucleated at the corner of the whisker ends and grew
towards the centers of the whisker ends. They did not
obseerved the void formation along the long sides of the
whisker/matrix interface, which is parallel to the tensile
direction. _

From the above experimental results, void nucleation at
the Sij/Al interface (i.e interface debonding) was
proposed as the failure mechanism of the Sij/Al composites
and as a reason for the low ductility and toughness in such
composites [62,63].

In this present experiment, some debonded SiCp/Al
interface was observed at the side-surface of the specimen
located just below the tensile fracture surface. Debonded
interfaces, which are formed perpendicular to the tensile
direction, can be seen in Fig.25). This micrograph was
taken at the side-surface of the SiCp/Al tensile specimen
just below the fracture surface. The direction of
propagation of void due to debonded SiCp/Al interface is
the same as that of the debonded Sij/Al interface observed

by Nutt and Duva.

72

Figure 25. Debonded interface and particulate cracking
of tensile fracture surface

72—A

 

 

'Figure.25

 

73

Figure 26. Stress concentration
at the pole of the inclusion
during uniaxial loading

74

Such a phenomenon of interface debonding can be
explained with the help of a simple model of "a rigid
spherical inclusion embedded in an elastic solid", shown in

Fig.26). If a tensile stress (a ) is applied on the

aPP
system, the radial tension (arr i.e decohesion stress) at
the "pole" of the spherical inclusion (i.e at the point A
and A’) becomes intensified to a value of

a = [ 2/(1+,,) + 1/(4-5,,) ]-a a ------ 19)

rr pp

in the direction of the externally applied tension (0

app)
[64]. Substitution of v = 0.33 for the Al alloy gives

arr = 1.93-0 app

“ 2'” app
in the direction of the applied tensile loading. Such
stress concentration (radial tension) at the pole of the
rigid inclusion (SiCp) can attribute to the debonding of
the inclusion-matrix interface.

Although the void nucleation mechanism is an evident
operating mechanism for the failure of the SiCp/Al
composites, this mechanism seems to be insufficient to
explain such a large difference in the ductility between
the reinforced and the unreinforced Al-alloy, since the
interface debonding due to void nucleation is a very rare
event as compared to the cracking of SiCp. The latter can
be observed from the metallographis and fractographis

observations provided in Figs.31) and 32).

 

75

Another failure mechanism was suggested in 1987 by You

et a1 [65] who conducted an SEM examination on the tensile

 

fracture surface of the SiCp/Al composites. In you's
study, the numbers of the cracked particulates and the
debonded interfaces were counted at the fracture surface.
From the analysis, the number of the cracked particulate
was found to be twice more than the number of the debonded
interface.

At the same time, intensive plastic deformation in the
matrix between SiCp was also observed from the side-surface
of the tensile specimen located just below the fracture
surface. According to the above observation, the matrix
failure was proposed as the dominant failure mechanism of
the SiCp/Al composites, i.e the cracking of Sicp or
debonding of SiCp/Al interface were attributed to the
matrix failure.

It is evident from Fig.27) that the crackings of Sicp
always preceed the matrix failure, although severe plastic
deformation can be observed in the matrix near the region
of particulate cracking. Therefore, the matrix failure
mechanism does not seem to be plausible for the failure of
the SiCp/Al composites.

In order to explain why the crackings of the SiCp
preceed the matrix failure, the model of "a spherical
inclusion embedded in the elastic material", which is under
uniaxial tensile loading , can be introduced again. If an

uniaxial tensile load (a ) is applied on such a composite

app

 

76

Figure 27. Particulate cracking in a tensile specimen

Note : Arrow mark indicates the initiation of
crack propagation into matrix.

 

76—A

 

«osmzo o:mn:o:

Figure.27

77

system, the tensile hoop stress (000) is induced at the
"equator" (i.e BCDB') in the same direction with the
applied tension. The stess distribution around the
inclusion is shown in Fig.28). The hoop stress (”96) at

the equator is given by Goodier [64].

a = [ (27-15u)/(14-10v) ]-aa

90 """"" 20)

PP
Substitution of v = 0.33 for the Al alloy gives the tensile
hoop stress at the equator as

006 = 2.1 aapp
in the direction of the applied tension. However,
according to the Saint—Venant’s principle, the change in
the stress distribution is negligible at a distance larger
than the radius of the inclusion (or SiCp). Therefore, it
is evident from Eq.l9) and 29) that once the external loal
is applied on such a composite system, an inclusion
embedded in a matrix acts as a stress raiser.

The Fig.29) shows photoelastic fringes developed around
the circular cavity, which enable the visualization of the
stress concentration near the cavity. It can be seen that
the colored fringes disappear very rapidly as one moves far
from the edge of the cavity. This pattern is in good
agreement with the Goodier’s solution. This indicates that
if elastic deformation is assumed for the SiCp and A1-

matrix during the initial stage of tensile loading, the

tensile hoop stress near the equatorial surface region of

 

78

Figure 28. Stress concentration
at the equator of the inclusion
during uniaxial tensile loading

 

 

 

 

 

 

 

 

 

 

 

 

 

 

79

Figure 29. Visualization of the stress concentration
near the cavity under uniaxial tension

Note that the stress state in-between the

cavities (X) is higher than that of equatorial
region indicated by "Y".

 

79-A

 

Figure.29

80

the Sicp becomes at least tWice higher than that of the
matrix far from the Sicp, while the stress in the matrix is

the same as the applied tension (0 ). Such stress

aPP
concentration (tensile hoop stress) at the equator of the

Sicp gives rise to the cracking of the SiCp.

has a crack in it, the constraint for the plastic

Once the SiCp

deformation of the Al-matrix will disappear. Then Al-
matrix near the SiCp can now easily undergo plastic
deformation causing the opening-up of the cracked planes in
the SiCp as can be seen in Fig.27). This situation can be
considered to be same as the Al-alloy which has penny-
shaped flaw (or crack) in it. If the external tensile
stress is applied further, the penny—shaped crack will
propagate into the Al matrix due to the stress
concontration at the crack tip. This makes the matrix fail
easily and thus the ductility of the composites will
decrease substantially.

Based on the present study, the failure mechanism of
SiCp/Al composites can be summarized as follows:

Once a tensile load is applied externally on the
composite system, hoop stress and radial stress are
generated on the Sicp in direction of the applied tension.
Such stress concentration may cause cracking or interface
debonding. As more load is_applied, the crack and debonded
interface, which are formed already, are easily opened—up
due to the plastic deformation of the matrix. New crack

will be developed in the matrix at the tip of the opened-up

 

l—__":“‘:—'Z£T“ . l ,, _1 , w r, ,,

crack, propagate into matrix, be connected with nearby
cracks, and finally form a large void. The whole steps for
the failure mechanism is drawn schematically in Fig.30).
The SEM investigation was carried out on the side-
surface of the fractured tensile specimen to examine the
proposed failure mechanism. Figs.31) and 32) are the
micrographs of the side-surface of the tensile specimen
located below the fracture surface. Extensive particulate
cracking and debonded interface, formed under the applied
uniaxial tension, can be seen in this tensile specimen.
The cconnection of the near-by cracks into a large crack is
illustrated in Fig.31-c). The void formation can be
-observed in Fig.32-a). Therefore, particulate cracking in
addition to interface debonding can be proposed as a major

contributor to the failure of SiCp/Al composites.

 

82

Figure 30. The schematic diagram showing
the failure mechanism of Sic /Al composites
under uniaxial tension p

a)

13)

CI

d)

f)
9)
g’)
h)

3')

Generation of tensile hoop stress (”00)
at the equator of SiCp
Formation of crack plane in the SiCp (Fig.27)

Opening-up of crack plane due to
plastic flow of Al-matrix (Fig.27)

Equivalent diagram of Fig.30-c)

Crack propagation inti Al-matrix due to
stress concentration build up at the crack
tip (Fig.27)

Generation of tensile radial stress (a
at the pole of Sicp

Formation of debonded interface

rr)

Opening-up of debonded interface
Equivalent diagram of Fig.30-g)
Crack propagation into matrix
Joining of cracks (Fig.31-c)

Void formation (Fig.32)

 

 

 

 

 

 

 

 

 

 

 

 

Particulate cracking
”app
it
I Gig
_)
l©l
a I)
ll
”app
— Z
c’ c

 

 

 

 

 

 

 

 

 

 

82—A

 

__-_—._—--. ----—-_- —————-

 

Interface debondingl

 

 

 

 

 

”app
1]

 

T

 

e ¢QT

 

 

U
“899

 

O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure.30

 

83

Figure 31. SEM micrograph showing crack development
on a tensile specimen (Etched with dilute HCl)

 

84

Figure 32. SEM micrograph showing void formation
due to joining of cracks (arrow mark)

 

 

 

84-A

 

 

Figure. 32

 

85
5 . CONCLUSIONS
5.1 Strengthening Mechanism of SiCp/Al composites

1. Some plausible strengthening mechanisms, such as
Orowan strengthening, composite strengthening using the
modified shear lag theory, thermal strain hardening due to
enhanced dislocation density, and strengthening due to
subgrain and smaller grain, contribute to the enhanced
yield strength of SiCp/Al composites.

2. Composite strengthening, thermal strain hardening,
and strengthening due to subgrain were found to be major
contributors to the strengthening of SiCp/Al composites.

3. On the other hand, it is considered that the effect
of the Orowan strengthening is too small as compared to the
substantial increase in the yield strength observed in

such composites.

 

86
5.2 Microstructural Features

1. As-extruded composites (10 % SiCp/6061-Al)_exhibited
severely banded structure of SiCp clusters. Significant
redistribution of the SiCp clusters , which resulted in the
uniform distribution of such clusters, was achieved with
increasing reduction ratio by both warm and cold rolling.
Banded structure of SiCp clusters totally disappeared
beyond 60 % reduction.

2. The cold rolled composites exhibited the presence of
larger voids and debonded interfaces as compared to those
of the warm rolled composites. Such voids grew larger with
increasing reduction ratio.

3. In case of warm rolling, the composites could be
rolled down to as much as 85 % of reduction without forming
any edge crackings or surface scuffings. On the other
hand, in case of cold rolling, edge cracks and surface
scuffings were formed before 40 % reduction.

4. Both particulate cracking and interface debonding
were observed after rolling; Such features were more
predominant in case of cold rolling rather than in warm
rolling. There is strong tendency for the crack planes in
the particulate to be formed parallel to the direction of
rolling pressure, and perpendicular to the rolling

direction.

87
5.3 Mechanical Properties

1. As-extruded composites exhibited anisotropic
mechanical properties in their longitudinal and transverse
direction. Such anisotropic mechanical properties result
from the microstructural inhomogeneity of the as-extruded
composites. Mechanical properties become more isotropic
with increasing reduction ratio; at about 50-60 % reduction
longitudinal and transverse directions possess the same
properties

2. Although a significant redistribution of SiCp
clusters was achieved after warm rolling, the principal
effect of the warm rolling on the mechanical properties was
to decrease the yield and tensile strength as well as the
fracture elongation in the longitudinal direction with
increasing reduction ratio. However, such properties in
the transverse direction were found to increase with
increasing reduction ratio up to 60 % and decrease
afterwards.

3. The decrease in the mechanical properties in the
longitudinal direction can be attributed to the growth of
pores, the interface debonding, the particulate cracking,
and the grain growth of the matrix during stress relief
annealing. On the other hand, the improvement in the
mechanical properties in the transverse direction may
be due to the redistribution of the Sicp clusters, which

results in more uniform microstructure.

5.4 Failure mechanism of the SiCp/Al composites

under uniaxial tension

1. Interface dedonding due to void formation at the
composites interface, which was proposed as a failure
mechanism of Sij/Al composites by Nutt and Duva, was also
observed near the pole of SiCp. Such an interface
debonding is considered to be formed due to the radial
tension induced at the pole of Sicp; the magnitude of this
radial tension is almost twice that of the applied tension.

2. The matrix failure mechanism proposed by You gt_a;
does not seem to be plausible for the failure of the
SiCp/Al composites, since the crackings of SiCp were found
to always preceed the matrix failure.

3. Tensile hoop stress induced at the equator of Sicp
appears to be responsible for the cracking of Sicp. The
magnitude of the hoop stress was found to be at least twice
as large than that in the matrix.

4. Once cracking of SiCp and debonded interface are
formed by the applied tensiOn, such cracks can easily be
opened up, connected with each other, and at last make a
large void under further applied stress.

5. Particulate cracking under uniaxial tension appears
to be a more dominant failure mechanism of the SiCp/Al
composites rather than the void nucleation mechanism at the
interface, since the number of cracked Sicp was observed to

be more than that of debonded interface.

 

 

10.

11.

89

6. REFERENCES

I.Ahmad, V.P.Greco and J.M.Barranco "Reinforcement of
nickel with some high strength filament", J. Composite
Materials, 1, 19 (1967)

E. G. Wolff "Hydrodynamic Alignment of discontinuous

Fibers in a Matrix" Fiber Science and Technology,
Elsevier publishing co. Essex. England (1969)

N. Tsangarakis, B.O.Andrews and C.Cavallaro
"Mechanical properties of some silicon carbide
reinforced aluminium composite", J. Composite
Materials, g1, 481-492 (1987)

T.G.Nieh and D.J.Chellman "Modulus measurements in
discontinuous reinforced aluminium composites"
Scripta Metallurgica, lg, 925-928 (1984)

R.J.Arsenault and S.B.Wu "A comparison of PM vs. melted
SiC/Al composites", Scripta Metallurgica, 2;, 767-772
(1988)

R.J.Arsenault "The strength of aluminium alloy 6061 by
fibre and platelet silicon carbide", Material Science
and Engineering, g3, 171-181 (1981)

Y.Flom and R.J.Arsenault "Interfacial bond strength in
an aluminium alloy 6061-Sic composites", J. Metals,;§,
31-34 (7.1986)

V.C.Nardone "Assessment of models used to predict the
strength of discontinuous silicon carbide reinforced
aluminium alloys", Scripta Metallurgica, 21, 1313-1318
(1987)

R. H. Jones, C. A. Lavender and M. T. Smith "Yield strength
-fracture toughness relationship in metal matrix
composites", Scripta Metallurgica, 2;, 1565— 1570
(1987)

W.A.Longsdon and P.K.Liaw "Tensile, fracture toughness
and fatigue crack growth rate properties of silicon
carbide whisker and particulate reinforced aluminium
metal matrix composites", Engineering Fracture
Mechanics, 21, No.5, 737-751(1986)

T.Christman and S.Suresh "Effects of Sic reinforced and
aging treatment on fatigue crack growth in an Al-SiC
composites", Material Science and Engineering, 19;,
211-216 (1988)

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

90

J.K.Shang, W.Yu and R.O.Ritchie "Role of silicon
carbide particles in fatigue crack growth in Sic
particulate reinforced aluminium alloy composites"
Material Science and Engineering, 102, 181-192 (1988)

 

H.J.Heine "Cast aluminium metal metrix composites are
here", Foundary management & technology, 116, 25-30
(7.1988)

F.M.Hosking, F.F.Portillo, R.Wunder1in and R.Mehrabian
"Composites of aluminium alloy : fabrication wear
behavior", J. Material Science, 11, 477-498 (1982)

F.A.Girot,J.M.Quenisset and R.Naslain "Discontinuously
reinforced aluminium matrix composites", Composite
Science and Technology, 39, 155—184 (1987)

V.C. Nardone and J.R. Strife "Analysis of the creep
behavior of silicon carbide reinforced 2124 Al (T4)"
Metallurgical transaction A, 18A, 109-114 (1987)

K.S.Ravichandran "Advanced aerospace Al alloy"
J. Metals, 32 28-32 (1987)

F.Dolowy "Increasing focus on silicon carbide
reinforced aluminium composites", Light Metal Age,
7-14, (6.1986)

4,

D.Hughes "Textron unit makes reinforced titanium,
aluminium parts", Aviation week and space technology,
129, 91-95, (11.1988)

M.A.H.Howes "Ceramic reinforced MMC fabricated by
squeeze casting", J. Metals 38, 28-29, (3.1986)

F.A.Girot, L.Albingre, J.M.Quenisset and R.Naslain
"Rheocasting Al matrix composites", J. Metals, 18-21,
(11,1987)

T.R.Pritchett "Advenced technology aluminium materials
for aerospace application", Light metal age, 44, 10-14
(10.1986)

A.Mortensen, J.A.Cornie, and M.C.Flemings
"Solidification prossing of metal matrix composites"
J. Metals, 12-19 (1988)

T.Choh and T.Oki "Wettability of SiC to aluminium and
aluminium alloys", The Institute of Metals, 378-385
(1987)

V.Laurent,D.Chatain and N.Eustathopoulus "Wettability
of SiC to aluminium and Al-Si alloys", J. Material
Science, 2;, 244-250 (1987)

26.

91

A.Banergi and P.Rohatgi "Cast aluminium alloy

containing dispersions of TiO2 and ZrO2 particles"

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

J. Material Science, 21, 335-342 (1982)

S.Deonath, R.Bhatt and P.Rohatgi "Preparation of cast
aluminium alloy - mica particle composites"
J. Material Science, 25, 1241-1251 (1980)

G.Mott and P.K.Liaw "Correlation of mechanical and
ultrasonic properties of Al-SiC metal matrix
composites", Metallurgical transaction A, 19A, 2233-
2246 (1988)

 

B.C.Pai, S.Pay, K.V.Prabhakar and P.K.Rohatgi
"Fabrication of aluminium — alumina (magnesia)
particulate composites in foundaries using magnesium
addition to the metals", Material Science and
Engineering, 24, 31-44 (1976)

P.K.Rohatgi, B.C.Pai and S.C.Panda "Preparation of cast
aluminium - silica particulate compositesfl, J. Material
Science, 15, 2277-2283, (1979)

M.K.Surrapa and P.K.Rohatgi "Preparation and properties
of cast aluminium — ceramic composites", J. Material
Science, 26, 983-993, (1981)

T.W.Clyne, M.G.Bader, G.R.Cappleman and P.A.Hubert "The
use of 6-alumina fibre for metal matrix composites",
J. Material Science, 29, 85-96, (1985)

B.P.Krishman, M.K.Surrpa, and P.K.Rohatgi "The UPAL
process : a direct method of preparing cast aluminium
alloy - graphite particle composites", J. Material
Science, 26, 1209-1216, (1981)

R.Mehrabin, R.G.Riek,and M.C.F1emings "

Metallurgical Transaction, 5, 1899-1905, (1974)
T.G.Nieh and R.F.Karlak "Hot-rolled silicon carbide -
aluminium composites", J. Material Science Letter, 2,
119-122 (1983)

J.R.Picken, T.J.Langan, R.O.England, and M.Liebson "A
study of the hot working behavior of sic - Al alloy
composites and their matrix alloys by hot tortion
testing", Metallurgical Transaction A, 122, 303-312
(1987)

K.Kannikeswaran and R.Y.Lin "Trace element effects on
Al-SiC interface", J. Metals, 22, 17-19 (9.1987)

R.J.Arsenault "Interface in metal matrix composites"
Scripta Metallurgica, 22, 1131-1134 (1984)

 

 

 

 

 

39.

40.

41.

43.

43.

44.

45.

46.

47.

48.

49.

50.

51.

92

T.Iseki, T.Rameda, and T.Murayama "Interfacial
reactions between Sic and alum1n1um during joining"
J. Material Science, 22, 1692-1698 (1984)

S.G.Fishman "Interface in composites", J. Metals, 22,
26-27 (3.1986)

ASM Metals Handbook, 9th ed, 2, "Properties and
selection : Nonferrous alloys and pure Metals", 24-42,
(1979)

CRC Handbook of Materials Science,
CRC Press Inc. (1975)

2, 378-380,

J.H.Beatty and G.J.Shiflet "Orowan strengthening by
M02C fibers and needle interphase precipitates in Fe-C-
Mo dual-phase steel", Metallurgical Transaction A,
222, 1677-1620 (1988)

P.M.Kelley "The quantitative relationship between
microstructure and properties in two-phase alloys"
Int. Met. Reviews. 22, 31-36, (1973)

H.L.Cox "The elasticity and strength of paper and other
fibrous materials", Br. J. Appl. Phys. 2, 72—79,(1952)
V.C.Nardone and K.M.Prewo "On the strength of
discontinuous silicon carbide reinforced aluminium
composites", Scripta Metallurgica 22, 43-48 (1986)

R.J.Arsenault and P.H.Fisher "Microstructure of fiber
and particulate SiC in 6061 Al composites"
Scripta Metallurgica 21, 67-71 (1983)

M.Vogelsang, R.J.Arsenault and P.M.Fisher "An in situ
HEVM study of dislocation generation at Al/SiC
interface in metal matrix composites", Metallurgical
Transaction A, 212, 379-389 (1986)

N.Hansen "The effect of grain size and strain on the
tensile flow stress of aluminium at room temperature"
Acta Metallurgica 22, 863-869, (1977)

R.J.Arsenault and N.Shi "Dislocation generation due to
difference between the coefficients of thermal
expansion", Materials Science and Engineering 22, 175-
187 (1986)

V.C.Nardone and J.R.Strife "Analysis of the creep
behavior of silicon carbide whisker reinforced 2124 Al
(T4)", Metallurgical Transaction A, 18A, 109-114
(1987)

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

93

H.J.MaQueen and J.J.Jonas "Recovery and
recrystallization during high temperature deformation"
Plastic deformation of materials, Academic press, New
York (1975)

H.J.MaQueen, W.A.Wong and J.J.Jonas "Deformation of
aluminium at high temperature and strain rates"
Can. J. Phys. 22, 1225-1235 (1967)

W.A.Wood and W.A.Rachinger
J. Inst. Metals 12, 237-253 (1949)

M.W.Farag, C.M.Sellars and W.J.Tegart, Deformation

under hot working conditions, 6Q-67, Iron steel Inst.
London 8‘

R.T.DeHoff "Problems solving using quantitative
stereology", Applied Metallography editted by V.Voort,
Van Nostrand Reinhold Co. New York (1986)

R.T.DeHoff and F.N.Rhines, Quantitative Micrography,
McGraw-Hill, New York (1968)

J.Gurland "Observation on the fracture of cementite
particles in a spherodized 1.05 % C steel deformed at
room temperature", ACTA Metallurgica, 22, 735-741,
(1972)

J.D.Lubahn "Simultaneous aging and deformation in
metals", Trans AIME, 185, 702-708, (1949)

R.E.Reed-Hill Physical Metallurgy Principle, 2nd ed,
347-351, D.Van Nostrand Co. New York (1973)

Y.Flom and R.J.Arsenault "Interfacial bond strength in
an aluminium alloy 6061-Sic composites", Material
Science and Engineering, 11, 191-197, (1986)

S.R.Nutt and J.M.Duva "A failure mechanism in Al-SiC
composites", Scripta Metallurgica, 22, 1055-1058,
(1986)

S.R.Nutt and A.Need1eman "Void nucleation at fiber ends
in Al-SiC composites", Scripta Metallurgica, 22, 705-
710, (1987)

J.N.Goodier "Concentration of stress around spherical
and cylindrical inclusions and flows", J. Applied
Mechanics, 22, A-39-43, (1933)

C.P.You, A.W.Thompson and I.M.Bernstein "Proposed
failure mechanism in a discontinuously reinforced
aluminium alloy", Scripta Metallurgica, 22, 181-185,
(1987)

 

 

    

T

    

ll lll

‘IIIilIIIIIi IIlIl IIIIIIIIIIIIIIIIIIIII I II

2 3006050557