Materials Science and Engineering A328 (2002) 137 146

www.elsevier.com/locate/msea

Mechanical properties of nickel silicon carbide nanocomposites

A.F. Zimmerman a, G. Palumbo b, K.T. Aust a, U. Erb a,*
a

Department of Materials Science and Engineering, Uni6ersity of Toronto, Toronto, ON, Canada, MS5 3E4
b
Integran Technologies, 1 Meridian Rd., Toronto, ON, Canada, M9W 4Z6
Received 13 March 2001; received in revised form 11 June 2001

Abstract
Nanocomposite materials consisting of a nanocrystalline Ni matrix (grain size 10 15 nm) reinforced with sub-micron size SiC
particulates (average particle size: 0.4 mm) up to 10.5 vol.% have been produced by pulse electrodeposition. Substantial
improvements in mechanical properties including hardness, yield and tensile stress were obtained for the nanocomposite material,
as compared with conventional NiSiC composites with a matrix grain size in the micrometer range. Tensile strengths up to four
times that for conventional polycrystalline Ni and two times that for conventional polycrystalline Ni SiC of comparable SiC
content was measured. The tensile and yield strengths of the nanocomposite material with SiC content less than 2 vol.% were
higher than those for pure nanocrystalline Ni of comparable grain size. For these nanocomposites an unexpected increase in
tensile ductility was also observed when compared to pure nanocrystalline nickel. At higher SiC content ( \2 vol.%) the strength
and ductility were found to decrease to the detriment of the nanocomposite. Particle clustering was considered the main cause of
this decrease. 2002 Elsevier Science B.V. All rights reserved.
Keywords: Ductility; Microhardness; Nickel silicon carbide nanocomposite; Tensile strength

1. Introduction
The mechanical properties of polycrystalline metals,
alloys and MMCs are very sensitive to their grain size.
In many cases the flow stress, at low temperature,
depends on the grain size according to the Hall Petch
relationship:
|y =|0 +kd 1/2

(1)

where |y is the yield stress, |0 is the friction stress, d is

the grain size and k is the Hall Petch coefficient (i.e.
the Hall Petch slope) [1,2]. Many models, based on the
dislocation theory, have been presented to interpret this
phenomenological equation (e.g. [3 7]). Restrictions on
dislocation generation and mobility imposed by ultrafine grain size are believed to be the dominant factor
in raising strength.
Numerous investigators have reported very high
strengths associated with Hall Petch behaviour in
nanocrystalline materials (e.g. [8 13]). For some materials, an unexpected transition from grain-size strength* Corresponding author. Fax: + 1-416-946-3316.
E-mail address: [email protected] (U. Erb).

ening to grain-size softening (i.e. inverse Hall Petch

behavior) was observed at a critical grain size [8,9,12
20]. For example, Palumbo et al. [13] reported softening
effects for a series of as-prepared nanocrystalline Ni P
alloys with grain sizes in the range from 20 down to 3
nm. The data showed an initial increase in hardness
followed by a deviation from regular Hall Petch behaviour leading to softening for the smallest grain size
range. The maximum in the hardness curve occurred
for a grain size of approximately 8 nm.
Gleiter [21] first recognized that the intercrystalline
component of a material will form a distinct microstructural constituent when a material is processed
to have a nanocrystalline structure. A detailed study by
Palumbo et al. [22] using grains modeled in the shape of
regular tetrakaidecahedrons and assuming a grain
boundary thickness of 1 nm showed that the grain
boundary volume fraction increases rapidly from a few
percent at a grain size of 100 nm to over 50% at a grain
size of 2 nm. For grain sizes below 20 nm, the triple
junction (i.e. intersection line of three or more grain
boundaries) volume fraction reaches significant values
and exhibits a greater grain size dependence than does
the grain boundary volume fraction. In fact, it has been

0921-5093/02/$ - see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S0921-5093(01)01692-6

138

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

observed that the onset of the deviation from regular

Hall Petch behaviour often occurs at grain sizes where
the triple junction volume fraction in the materials
reaches significant values, as discussed elsewhere in
more detail [13,23].
Some of the unique mechanical properties of
nanocomposite materials including increased strength
and hardness [24,25], increased wear resistance and
decreased coefficient of friction [24], can also in part be
attributed to grain size effects caused by the increased
interfacial defect volume fraction, in addition to the
strengthening effects by the reinforcing phase.
The purpose of the present study was: (1) to prepare
bulk samples of a novel nickel/silicon carbide (Ni SiC)
nanocomposite as well as pure nanocrystalline nickel
(nano-Ni) by pulse electrodeposition; (2) to characterize
their microstructure using scanning electron microscopy
(SEM), transmission electron microscopy (TEM), X-ray
diffraction (XRD) analysis and image analysis; and (3)
to determine some mechanical properties of both
nanocomposite Ni SiC and nano-Ni using Vickers
hardness and uniaxial tensile tests.

2. Experimental procedures
Both nanocomposite Ni SiC and pure nanocrystalline Ni samples were prepared by pulsed current
plating [26]. The nanocomposite samples were deposited from a modified Watts bath described elsewhere
[27] containing 100 g/l and 20 g/l high purity SiC of
0.4 90.2 mm average particle diameter. A modified
Watts bath without SiC was employed in the plating of
the nanocrystalline Ni samples [26]. All plating baths
contained 5g/l of saccharin and a plating time of 34 h
duration was used. In order to obtain free-standing
samples that could be machined into suitable specimens
for the uniaxial tension tests, a Ti cathode with plating
area dimensions 40 20 mm was used. After removal
from the Ti substrate the test coupons were cut to the
dimensions (in mm), shown in Fig. 1, using electric
discharge machining (EDM). The specimens were then
ground and polished using standard metallographic
techniques. The final thickness of the tensile specimens
ranged from 100 to 300 mm.

Fig. 1. Schematic diagram of the minature plate-type tensile coupon

(dimensions in mm).

The grain size was determined using XRD in conjunction with the Scherrer method [28] and also by
direct measurement of about 250 grains on TEM
darkfield micrographs of some samples. A particle distribution analysis was utilized to calculate the vol.%
SiC in the deposits using a Globe Image Analyzer.
Samples were mounted and polished to a mirror like
finish using standard metallographic techniques including polishing with 6 and 1 mm diamond pastes and
finally with a 0.3 mm aluminum oxide slurry. Six readings were carried out and averaged for each sample.
Microhardness tests were conducted on a Vickers
hardness tester using a load of 100 g, applied for 20 s.
The tensile tests were carried out on a MTS machine
with a low capacity load cell (100 kg maximum) and a
clip-on high sensitivity extensometer with a gauge
length of 5 mm to measure elongation. The tests were
conducted under load control condition at 0.2 MPa
s 1. (i.e. equivalent to a strain rate of 5 10 5 to
10 4 s 1) to the fracture point or until the maximum
load was achieved on the load cell. Both load and
elongation signals were digitized by an Omega W-800
A/D card and recorded using an IBM PC. From the
load/extension data the yield strength (0.2% off-set),
tensile strength and elongation were determined for
both the nanocomposite NiSiC and the nanocrystalline Ni samples.
In order to determine the fracture mode and the
microstructural changes resulting from deformation,
standard metallographic and fractographic techniques
were employed.

3. Results

3.1. Sample characterization

The grain-sizes and SiC content (in vol.%) of the
nanocomposites produced in this study are summarized
in Table 1. The SiC content varied from 0.3 vol.% to a
maximum of 10.5 vol.% due to the variation in the
plating parameters used [27]. A typical microstructure,
consisting of irregular shaped particles and agglomerates of SiC embedded in a nanocrystalline Ni matrix, is
revealed by the SEM micrograph of a nanocomposite
Ni SiC electrodeposit shown in Fig. 2. The SiC concentration in this sample was 8.7 vol.%.
The grain sizes, determined by XRD, ranged from
10.3 to 15.0 nm as indicated in Table 1. A direct
measurement of 250 grain diameters from a TEM
darkfield micrograph of the nanocrystallline Ni-matrix
between the SiC particles of nanocomposite sample E8
(Fig. 3a), produced a grain size distribution with a
mean value of 18.39 6.8 nm, shown in the histogram of
Fig. 3b. In comparison, the value for the grain size
obtained from XRD line broadening [28] for the same

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

139

Table 1
Grain size, vol.% SiC and mechanical properties of nanocomposite NiSiC
Sample no. n-NiSiC

|y (MPa)

|UTS (MPa)

mf (%)

Hardness (GPa)

G.S (nm)

Vol. % SiCc

E3
E5
E15
E8
E17
E16
E2
E6a
E18
E9
E4
E19
E20
E7a
E1
E13
MMCb
NiSiC1
NiSiC2
NiSiC3
NiSiC4

1050
820
800
730
800
775
920
820
775
820
800
925
730
880
840
850

1413
1350
1150
1256
1246
1115
1431
\1163
1275
1072
1064
1298
1070
\1033
1075
955

0.70
2.20
2.50
3.40
2.00
2.80
2.10
\0.65
1.50
0.50
0.60
1.00
1.00
\0.35
0.50
0.35

6.83 90.15
6.49 9 0.29
5.81 90.25
6.13 9 0.27
6.13 90.28
6.15 9 0.22
5.99 90.29

6.54 90.35
6.27 9 0.15
6.39 90.36
6.31 9 0.33
6.31 90.22
6.35 9 0.25
6.59 90.16
6.46 9 0.25

12.4
10.6
13.7
14.2
13.7
12.2
10.3
14.2
11.7
14.0
14.2
11.7
12.8
15.0
13.7
14.5

0.3
0.4
0.6
0.7
1.2
1.4
1.8
3.0
3.2
3.4
3.5
3.9
4.0
5.9
8.7
10.5

569
588
588
539

676
735
833
726

B3
B3
B3
B3

3.15
3.90
4.02
4.41

n.a.d
n.a.
n.a.
n.a.

7.1
11.2
14.6
23.4

These samples did not fail during testing.

Conventional MMCs produced in a Watts bath [3538].
c
The vol.% SiC data was converted from wt.% values taken from the literature [3537].
d
n.a. means not available.
b

sample was 14.2 nm which is in reasonable agreement

considering that X-ray diffraction averages the grain
size over much larger volume than TEM dark field
imaging.
The grain sizes for the pure nanocrystalline Ni samples ranged from 9.9 to 40 nm (See Table 2).

3.2. Mechanical testing results

The results of mechanical testing, i.e. Vickers hardness, 0.2% off-set yield strength (|y), ultimate tensile
strength (|UTS) and elongation (mf) are given for both
nanocomposite Ni SiC and pure nanocrystalline Ni in
Table 1 and Table 2, respectively. Comparative mechanical properties for conventional polycrystalline
Ni SiC MMCs and annealed polycrystalline Ni are
also displayed.
In the following sections the results are presented in
form of scatter plots containing points labeled E for
nanocomposite materials and F for nanocrystalline Ni
without SiC additions according to Table 1 and Table 2
respectively. In order to separate the effects of SiC
content and grain size on the mechanical properties of
these materials each scatter plot is followed by a histogram comparing relevant properties. In histograms
4b (hardness) and 5b (yield stress) materials with comparable SiC content and/or grain size are compared.
Histogram 6b (tensile strength) also shows comparisons
for nanocrystalline materials with varying SiC content

and grain size, while histogram 7b (strain to failure)

compares nanocrystalline materials.
The microhardness values, for the nanocomposite
samples appearing in Table 1 range from a low of
5.819 0.25 GPa to a high of 6.839 0.15 GPa, for the
grain size range 10.315.0 nm. These hardness values
are comparable to those for pure nanocrystalline Ni
given in Table 2 which reveal a range of 4.8890.05 to
6.8990.10 GPa for a wider range in grain sizes, from
9.9 to 40 nm. Vickers hardness as a function of vol.%
SiC for the nanocomposite specimens is shown in the
scatter plot of Fig. 4a. A linear regression analysis gives

Fig. 2. SEM micrograph of a nanocomposite electrodeposit containing 8.7 vol.% SiC.

140

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

Fig. 3. (a) TEM darkfield micrograph of nanocomposite Ni SiC

sample E8. (b) Histogram of Ni grain size distribution for the same
sample E8.

a correlation coefficient of r= 0.43 indicating a slight

trend of increasing hardness with increasing vol.% SiC.
The F points on the Y-axis represent the hardness

values of the pure nanocrystalline Ni samples (i.e.

containing no SiC). A comparison of the microhardness
of a conventional Ni metal matrix composite containing 7.1 vol.% SiC, a nanocomposite NiSiC sample
with 8.7 vol.% SiC, polycrystalline annealed Ni and a
nanocrystalline Ni sample is given in the histogram of
Fig. 4b.
The yield strength values for the nanocomposite materials, ranging from a low of 730 MPa to a high of
1050 MPa (Table 1), are plotted as a function of vol.%
SiC in Fig. 5a. Linear regression analysis of the data
(i.e. a correlation coefficient of r= 0.06) demonstrates that SiC content has little effect on the yield
strength, however, the range of yield strength values for
the nanocrystalline nickel samples, 630851 MPa
(Table 2), is somewhat lower than that for nanocomposite NiSiC. For many of these specimens the yield
strength and tensile strength are identical or show very
little difference (e.g. for samples F2, F3, and F9 see
Table 2). This indicates that the point of fracture is
nearly the same as the yield point which implies that
pure nanocrystalline Ni is relatively brittle, possessing
little plastic deformation. A comparison of yield
strength values for a nanocrystalline Ni sample, a
nanocomposite NiSiC sample with a SiC content of
8.7 vol.%, polycrystalline annealed Ni, and conventional Ni MMC containing 7.1 vol.% SiC is displayed
in the histogram of Fig. 5b.
The ultimate tensile strength of the nanocomposite
Ni SiC samples ranges from 955 to 1431 MPa (Table
1). The strength values are plotted as a function of
vol.% SiC shown in Fig. 6a. Linear regression analysis
of the plotted data (i.e. a correlation coefficient of
r= 0.52) reveals a moderate to fairly strong inverse
relationship between tensile strength and vol.% SiC.
The ultimate strength values of the nanocomposites are
also substantially higher than the values for nanocrystalline Ni samples of comparable grain sizes. The values

Table 2
Grain size and mechanical properties of nanocrystalline nickel
Sample no. nano-Ni

|y (MPa)

|UTS (MPa)

mf (%)

Hardness (GPa)

G.S (nm)

F3
F4
F5b
F8
F2
F9
F7
F1c
Niand

851
830
296
830
630
641
770
700
177

851
1056
296
881
661
641
968
\1370
192

NDa
0.60
ND
0.26
0.22
ND
0.50
2.35
50

6.14 9 0.24
6.22 90.23
6.67 90.07
5.65 90.35
6.66 90.24
6.89 90.10
4.88 9 0.05
5.54 90.12
1.48

9.9
10.9
11.7
13.3
13.7
13.8
15.0
40
100 mm

No elongation at the level of sensitivity of the MTS machine.

The sample contained surface defects
c
The sample did not fail.
d
Conventional polycrystalline nickel annealed at 540 C for 2 h [45].
b

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

141

tively and conventional Ni MMC with a SiC content of

7.1 vol.%.
The elongation to failure, mf, values for many of the
nanocomposite NiSiC samples are relatively high
compared with the values for nanocrystalline Ni. The
lower the concentration of SiC in the samples the
greater the strain to failure. This inverse relationship
between elongation and SiC content is evident in the
scatter plot, shown in Fig. 7a. Linear regression analysis (i.e. a correlation coefficient of r= 0.57) indicates
a moderate to fairly strong inverse relationship between
mf and vol.% SiC. The histogram, shown in Fig. 7b,
compares the mf values of two nanocrystalline Ni samples of grain sizes 40 and 13.7 nm, respectively, and

Fig. 4. (a) Scatter plot of microhardness versus vol.% SiC in the

deposits. (b) Histogram comparing microhardness of poly- and
nanocrystalline materials. 1Taken from Refs. [35 38]. 2Taken from
Ref. [45].

are compared in the same scatter plot of Fig. 6a. The

ultimate tensile strength for pure nanocrystalline Ni,
shown on the plot, ranges from 641 to 1056 MPa for
grain sizes 515 nm (Table 2). One pure nanocrystalline Ni sample, F1, of a larger grain size (40 nm)
demonstrates a strength comparable to the strongest
nanocomposite samples (]1370 MPa) indicating a
greater plastic flow in the matrix due to a more developed dislocation mechanism at larger grain sizes. The
histogram displayed in Fig. 6b, compares the tensile
strength of polycrystalline annealed Ni, two nanocrystalline Ni samples, two nanocomposite Ni SiC samples
containing 1.8 vol.% SiC and 8.7 vol.% SiC, respec-

Fig. 5. (a) Scatter plot of yield stress versus vol.% SiC. (b) Histogram
comparing yield stress of poly- and nanocrystalline materials. 1Taken
from Refs. [35 38]. 2Taken from Ref. [45].

142

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

variation in grain size from 15.0 to 13.8 nm, for samples F7 and F9, the microhardness values increased
significantly from 4.8890.5 to 6.899 0.10 GPa, respectively. Sample F8, with a similar grain size of 13.3 nm,
also revealed a significantly lower hardness value of
5.659 0.35 GPa. These anomalies in grain sizes and
microhardness values are likely due to variations in the
crystallographic textures exhibited by the electrodeposits. XRD analysis of the nanocrystalline Ni samples
indicated a preferred (111) crystallographic orientation
for the harder samples, including F9, while a predominate (200) orientation was shown for the softer samples, including F7 and F8. This effect of texture on the
hardness of nanocrystalline Ni materials has been reported earlier by other researchers [8,30].

Fig. 6. (a) Scatter plot of tensile strength versus vol.% SiC for
nanocrystalline materials. (b) Histogram comparing tensile strength
of poly- and nanocrystalline materials. 1Taken from Ref. [45]. 2Taken
from Refs. [35 38].

three nanocomposite Ni SiC samples containing 8.7,

1.2 and 0.7 vol.% SiC, at relatively constant grain size
(13.7 14.2nm) respectively. Refs [35 37] state that the
elongation to failure was less than 3%.

4. Discussion

4.1. Microhardness
Table 2 indicates a substantial variation in microhardness values for some pure nanocrystalline Ni samples with similar grain sizes. For example, for a

Fig. 7. (a) Scatter plot of elongation versus vol.% SiC. (b) Histogram
comparing the elongation of nanocrystalline materials.

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

Fig. 8. Hardness versus the inverse square root of grain size for
nanocrystalline Ni.

The histogram of Fig. 4b illustrates the effects of

adding SiC particulate and decreasing grain size on the
hardness of polycrystalline Ni. The addition of about 7
vol.% of hard SiC ceramic particulate to polycrystalline
nickel results in a large increase (\2 ) in the hardness value (i.e. from 1.48 to 3.15 GPa). Thus, SiC
particles have a major hardening effect on a polycrystalline nickel matrix. When the grain size of pure nickel
is decreased to the nanocrystalline range (14 nm)
another major increase (\4 ) in the hardness values
is observed (i.e. from 1.48 to 6.66 GPa). Therefore,
decreasing the grain size to the nanometer range has a
greater hardening effect (\ 2 ) than adding a reinforcing phase to conventional polycrystalline Ni. However, the addition of a comparable amount SiC (8.7
vol.%) to a nanocrystalline Ni matrix has very little
effect on the hardness. The hardness values for
nanocrystalline Ni and nanocomposite Ni SiC are similar. Therefore, it can be concluded that the high hardness values for the nanocomposite are due mainly to
the hard nanocrystalline nickel matrix, most likely because a strengthening mechanism based on dislocation
particle interaction is not operative at such small grain
sizes.
Dislocation models are meaningful only within certain limitations at very small grain sizes. For example,
the original dislocation model for the Hall Petch relation was based on the concept that grain boundaries act
as barriers to dislocation motion, thus forming dislocation pile-ups at grain boundaries. Pile-up models be-

143

come questionable when the pile-up length is of the

order of 10100 nm (i.e. the nanocrystalline range)
since the number of dislocations in a pile-up is rapidly
decreasing with decreasing grain size [29]. Similarly, the
classical Orowan type mechanism is unlikely to operate
in the nanocomposites studied here since the reinforcing
particles are at least one order of magnitude larger than
the average matrix grain size; this resulting in a structure in which one hard particle is surrounded by many
differently oriented grains in the nickel matrix. The
high hardness of the nanocomposites produced in the
present study is therefore primarily due to the intrinsic
hardness of nanocrystalline Ni matrix itself [9] rather
than the second phase reinforcement.
A Hall Petch plot (i.e. hardness as a function of
d 1/2) for pure nanocrystalline Ni is shown in Fig. 8.
As previously observed in other studies on electrodeposited nanocrystals [8,9,13,30,31], a deviation from
HallPetch behaviour (i.e. softening effect) is observed
for the smallest grain sizes. A number of theories based
on different factors have been advanced to explain this
inverse HallPetch relationship. These include triple
junction effects [13], diffusional creep [12], texture effects [8,30], grain boundary source models [32], decrease
in interfacial excess volume [33] and bow out of
FrankRead sources [34].

4.2. Yield strength

The histogram of Fig. 5b indicates that the yield
strength of polycrystalline nickel is greatly improved
either by adding SiC reinforcement or by reducing the
grain size. Adding SiC to nanocrystalline Ni does not
have the same significant effect on increasing the yield
strength as compared to adding SiC to conventional
polycrystalline Ni. For example, as shown in Fig. 5b, a
nanocomposite containing approximately 8.7 vol.% SiC
has a yield strength only slightly higher than that of a
nanocrystalline Ni sample of comparable grain sizes
( 15 nm): compare 840 and 770 MPa, respectively.
Adding approximately 7 vol.% SiC to polycrystalline
Ni increases the yield strength more than three times
(i.e. from 177 to 569 MPa). This can be expected since
the onset of composite yielding for a MMC is governed
by the onset of matrix yielding which, for nanocrystalline materials, is not under the control of a dislocation mechanism since the grain size is too small, as
discussed in Section 4.1. A HallPetch plot for yield
strength as a function of 1/d 1/2 indicated that the
softening effect is not as apparent as it is in Fig. 8 for
hardness versus 1/d 1/2.

4.3. Ultimate tensile strength and elongation to failure

An anomaly stands out on the scatter plot shown in
Fig. 6a: nanocomposite samples of very low volume

144

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

fractions of SiC reinforcement exhibit relatively high

tensile strengths. For example, nanocomposite samples
E2, E3, and E5 contain less than 2 vol.% SiC, yet,
exhibit some of the highest tensile strengths: 1431, 1413
and 1350 MPa, respectively.
Particle agglomeration can play a major role in the
tensile properties of composite materials [39 44]. Agglomerates of SiC particles were quite abundant in the

Fig. 9. (a) A schematic of the microstructure of a conventional

NiSiC MMC showing SiC particles pinning dislocations. (b) A
schematic of the microstructure of Nano-Ni SiC showing the presence of microvoids in SiC agglomerates.

Fig. 10. SEM micrograph of a fracture surface of nano-NiSiC sample

E4.

materials produced in this study, as shown for example

in the SEM micrograph of sample E1 with a SiC
content of 8.7 vol.% (Fig. 2). Thus localized ductile or
soft regions can occur in a hard, relatively brittle
nanocrystalline Ni matrix, due to enhanced void formation between neighbouring particles [40,44], which can
effectively increase the porosity of an otherwise fairly
dense material. This can lead to major effects on the
mechanical and tensile properties of a nanocomposite.
A pair of schematic drawings depicting two different
microstructures, shown in Fig. 9(a) and (b) explain this
reasoning. Fig. 9(a) depicts mobile dislocations being
pinned by SiC particles and SiC agglomerates in a
conventional polycrystalline NiSiC matrix leading to
a form of dispersion strengthening. Dislocation pile-ups
at grain boundaries also occur. This leads to a large
increase in strength over conventional polycrystalline
Ni (see Fig. 6b). Fig. 9b indicates a microstructure
consisting of particles and agglomerates of SiC in a
matrix of nanocrystalline Ni. The individual submicron
size particles are much larger than the nano-sized grains
of Ni. The presence of micropores between adhering
particles of an agglomerate is also depicted. As the
concentration of SiC in the nano-Ni matrix increases
damage accumulates and increases the porosity, thus
weakening the nanocomposite. Therefore, small volume
fractions of SiC are more effective at increasing the
strength of nanocomposite SiC than larger ones, which
are actually detrimental to composite strength. The
scatter plot of Fig. 6a as well as the histogram of Fig.
6b also indicate that the tensile strength values of the
nanocomposites for low SiC content (B 2 vol.%) are
higher than those for pure nanocrystalline Ni of comparable grain size. The same histogram also shows that
the strength of the nanocomposites are approximately
two times greater than the strength of conventional
metal matrix composites with similar SiC content. This
increase in strength is again mainly due to the decrease
in grain size of the Ni matrix (i.e. from the microregime to the nano-regime).
The scatter plot depicting ductility or strain to failure
(mf) as a function of SiC content is similar to the plot of
tensile strength as a function of SiC content; compare
Fig. 6a and Fig. 7a. Both graphs indicate inverse relationships between elongation or strength and SiC content and reveal similar correlation coefficients (0.57
and 0.52, respectively). Fig. 7a is also similar to
previously published ductility versus vol. fraction
plots of conventional MMCs exhibiting ductile fracture
[43,44]. A ductile fracture mechanism operative in the
nanocomposite NiSiC samples with low SiC concentrations is also supported by examining the fracture
surfaces of nanocomposite samples (e.g. for sample E4,
shown in the SEM micrograph of Fig. 10). A typical
fibrous cup and cone morphology containing SiC particles is revealed.

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

Fig. 11. SEM micrograph of fracture surface of nano-Ni sample F2.

Many of the pure nanocrystalline Ni samples, listed

in Table 2, have very low strains to failure. In some
cases there is negligible elongation (e.g. samples F3 and
F9), indicating that many of these samples fail by a
brittle fracture mechanism as observed in the fracture
surfaces [e.g. for nanocrystalline Ni sample F2 (mf =
0.22%), shown in Fig. 11]. Here, the micrograph reveals
cleavage-type facets (e.g. parallel plateau and ledge
morphology, flat river-patterns, etc.) typical of a brittle
fracture mechanism.
In the histogram of elongation for nanocrystalline
materials, shown in Fig. 7b, pure nanocrystalline Ni
(grain size of 14 nm) has the lowest mf value (0.22%),
close to that for a nanocomposite sample containing
one the highest SiC contents (i.e. 8.7 vol.%). The
nanocrystalline sample with the largest grain size (i.e.
40 nm for sample F1) revealed a substantially higher mf
value (i.e. 2.35%), comparable to values for nano-Ni
SiC samples of lower SiC content. At 40 nm grain size,
nanocrystalline Ni has a relatively higher value of elongation, compared to nano-Ni of smaller grain sizes (e.g.
13.7 nm), since a dislocation mechanism may be operative at the larger grain size. This also explains the
increase in tensile strength of the larger grain size
nano-Ni over the smaller grain size one. However, the mf
value is still relatively low compared to the ductility of
conventional polycrystalline nickel, i.e. 50% elongation
for annealed Ni [45] (Table 2).

5. Summary
Nanocomposite materials consisting of submicron
size SiC particulate embedded in a nanocrystalline
nickel matrix with grain sizes as small as 10 nm and SiC
contents up to about 10.5 vol.% were produced from a
modified Watts bath.
Microhardness values of the nanocomposite were
more than twice those for conventional polycrystalline

145

MMCs with similar SiC content and greater than four

times the hardness of conventional polycrystalline Ni.
The hardness values were similar to the values for pure
nanocrystalline Ni of comparable grain size.
Nanocomposite NiSiC demonstrated a tensile
strength greater than four times that of conventional
polycrystalline nickel and about two times that for
conventional polycrystalline MMCs of comparable SiC
content. The tensile strengths for the nanocomposite
samples with low SiC content (B 2%) were also higher
than for nanocrystalline Ni of comparable grain size.
Yield strength values greater than five times that for
annealed Ni and about 1.5 times that for conventional
polycrystalline MMC of comparable SiC content were
measured for nanocomposite NiSiC. Most pure
nanocrystalline Ni samples were brittle, exhibiting a
yield points close to their fracture points.
Nanocomposite NiSiC revealed an unexpected increase in ductility over that of pure nanocrystalline Ni
of comparable grain size, especially at lower SiC contents (B 2 vol.%). The ductility of nanocomposites of
lower SiC contents was comparable to the ductility of
conventional MMCs.
Fractography results revealed a ductile fracture
mechanism possibly via a void coalescence process operative in nanocomposite-NiSiC and a brittle cleavage-type mechanism in pure nanocrystalline Ni.

Acknowledgements
The authors wish to thank Dr N. Wang, Z. Wang
and V. Krstic for their contributions to this work.
Financial support for this work was provided by the
Natural Sciences and Engineering Research Council of
Canada and Ontario Power Generation. One of the
authors (A.F.Z.) would like to acknowledge financial
support by the School of Graduate Studies and Research at Queens University.

References
[1] O.E. Hall, Proc. Phys. Soc. London B64 (1951) 747.
[2] N.J. Petch, Iron and Steel Inst. 25 (1953) 174.
[3] J.D. Eshelby, F.C. Frank, F.R.N. Nabarro, Phil. Mag. 42 (1951)
351.
[4] A.H. Cottrell, Trans Metall. Soc. AIME 212 (1958) 192.
[5] A. Lasalmonie, J.L. Strudel, J. Mat. Sci. 21 (1986) 1837.
[6] R.W. Armstrong, in: T.N. Baker (Ed.), Yield, Flow and Fracture of Polycrystals, Applied Sci. Pub, Barking, UK, 1983, p. 1.
[7] J.C.M. Li, Y.T. Chou, Met. Trans. 1 (1970) 1145.
[8] G. McMahon, U. Erb, Microstr. Sci. 17 (1989) 447.
[9] A.M. El-Sherik, U. Erb, G. Palumbo, K.T. Aust, Scripta Metall.
Mater. 27 (1992) 1185.
[10] G.D. Hughes, S.D. Smith, C.S. Pande, H.R. Johnson, R.W.
Armstrong, Scripta Metall. 20 (1986) 93.

146

A.F. Zimmerman et al. / Materials Science and Engineering A328 (2002) 137146

[11] G.W. Nieman, J.R. Weertman, R.W. Siegel, Scripta Metall.

Mater. 24 (1990) 145.
[12] A.H. Chokshi, A. Rosen, J. Karch, H. Gleiter, Scripta Metall. 23
(1989) 1679.
[13] G. Palumbo, U. Erb, K.T. Aust, Scripta Metall. Mater. 24
(1990) 2347.
[14] K. Lu, M.L. Sui, Scripta Metall. Mater. 28 (1993) 1465.
[15] R.O. Scattergood, C.C. Koch, Scripta Metall. Mater. 27 (1992)
1195.
[16] N. Wang, Z. Wang, K.T. Aust, U. Erb, Mat. Sci. Eng. A237
(1997) 150.
[17] T.G. Nieh, J. Wadsworth, Scripta Metall. Mater. 25 (1991) 955.
[18] K. Lu, W.D. Wei, J.T. Wang, Scripta Metall. Mater. 24 (1990)
2319.
[19] T. Christman, M. Jain, Scripta Metall. Mater. 25 (1991) 767.
[20] V.B. Rabukhin, Phys. Met. Metalloved. 43 (1995) 3027.
[21] H. Gleiter, in: N. Hansen, et al. (Eds.), Second Riso Int. Symp.
Met. and Mater. Sci., Riso Nat. Lab. Denmark, 1981, p. 15.
[22] G. Palumbo, S.J. Thorpe, K.T. Aust, Scripta Met. Mater. 24
(1990) 1347.
[23] K.T. Aust, U. Erb, G. Palumbo, Interfacial Effects in Nanocrystalline Materials, in: C. Suryanarayana, J. Singh, F.H. Froes
(Eds.), Processing and Properties of Nanocrystalline Materials,
The Minerals, Metals and Materials Society, Warrendale, PA,
1996.
[24] Y. Ding, D.O. Northwood, A.T. Alpas, Surf. and Coat. Tech. 62
(1993) 448.
[25] Nieman G.W., Weertman J.R., Siegel R. Win, in: David C., Van
Aken, and Gary S. Was (Eds.), Microcomposites and
Nanophase Materials, The Mineral, Metals and Materials Society, Warrendale, PA, 1991, 15.

[26] El-Sherik A.M., Erb U., Production of Nanocrystalline Metals,

US Patent 5,433,797 (1995).
[27] Zimmerman A.F., Clark D.G., Aust K.T., Erb U., Mat. Letters,
submitted for publication.
[28] B.D. Cullity, Elements of X-ray Diffraction, 2nd, Addison
Wesley Publishing Co. Inc., London, 1978.
[29] Y.T. Chou, J. Appl. Physics 38 (1967) 2080.
[30] C. Cheung, F. Djuanda, U. Erb, G. Palumbo, Nanstr. Mat. 5
(1995) 513.
[31] C. Cheung, G. Palumbo, U. Erb, Scripta Metall. Mater. 31
(1994) 735.
[32] S. Li, L. Sun, Z. Wang, Nanostr. Mat. 2 (1993) 653.
[33] K. Lu, M.L. Sui, Scripta Metall. Mater. 28 (1993) 1465.
[34] J. Lian, B. Baudelet, Nanostr. Mat. 2 (1993) 653.
[35] M. Viswanathan, K.S.G. Das, Metal Finishing Feb (1972) 83.
[36] M.N. Joshi, M. Totlani, J. Electrochem. Soc. India 28 (1) (1979)
35.
[37] E. Broszeit, G. Heinke, H. Wiegand, Metall. 25 (1971) 470.
[38] W.H. Safranek, The Properties of Electrodeposited Metals and
Alloys, 2, AEFS, Florida, 1986, p. 365.
[39] K.E. Puttick, Phil. Mag. 4 (1959) 964.
[40] J.R. Fisher, J. Gurland, Metal Sci. 15 (1981) 185.
[41] G.L. Roy, J.D. Embury, G. Edwards, M.F. Ashby, Acta Metall.
29 (1981) 1509.
[42] A.S. Argon, Im J. Safoglu, Metall. Trans. 6A (1975) 825.
[43] A.R. Rosenfield, Met. Rev. 121 (1968) 29.
[44] T.W. Clyne, P.J. Withers, An Introduction to Metal Matrix
Composites, Cambridge University Press, Cambridge, 1993, pp.
256 263.
[45] ASM Metals Handbook vol. 2, ASM International, Metals Park,
OH, 1993.

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.