Materials Science and Engineering A 497 (2008) 212–215

Contents lists available at ScienceDirect

Materials Science and Engineering A

journal homepage: www.elsevier.com/locate/msea

Effects of Changes in Notch Radius and Test Temperature on the Toughness of a

Nano-crystalline Aluminum Alloy Composite Produced via Extrusion of
Amorphous Aluminum Alloy Powders
Hala A. Hassan a,b,∗ , John. J. Lewandowski b
a
Department of Design and Production Engineering, Faculty of Engineering, Ain Shams University, Cairo, Egypt
b
Department of Materials Science and Engineering, Case Western Reserve University, Cleveland, OH, USA

a r t i c l e i n f o a b s t r a c t

Article history: Amorphous aluminum alloy powders were extruded to produce a nano-crystalline aluminum alloy
Received 3 June 2008 composite-containing nano-sized particles. Significant increases to the hardness/strength were obtained
Accepted 30 June 2008 via the processing-induced evolution of 35 vol.% of nano-sized particles in the aluminum matrix. Acoustic
measurements of the elastic constants revealed elastic modulus (E) in the range of 86–89 GPa. The fracture
Keywords: toughness of these nano-structured composite materials were evaluated at room temperature and 225 ◦ C
Nano-crystalline composites
(498 K) on fatigue-precracked samples as well as on notched samples containing different root radii. The
Toughness
effects of changes in the root radius and test temperature on the toughness are compared to the behavior
Temperature effects
of conventional aluminum alloys and metal–matrix composites (MMCs).
© 2008 Elsevier B.V. All rights reserved.

1. Introduction sized Al2 O3 particles in Al via a powder-metallurgy process doubled

the yield strength. Currently, there are several fabrication methods
Nano-crystalline metallic materials and metal–matrix com- that have been used to produce nano-structured MMCs (NMMCs),
posites (MMCs) both provide unique, but different combination including: mechanical alloying with high-energy milling [10], ball
of properties. Because of their nano-scale grain size, the milling [11], nano-sintering [12], vortex process [13], spray deposi-
nano-crystalline metallic materials typically possess high-yield tion, and laser deposition. However, another potential approach to
strengths, as predicted by the Hall–Petch relationship [1,2] produce NMMCs is via the consolidation and subsequent working
although this may not hold at the very fine grain sizes due to of amorphous metal powders [14].
changes in deformation at these very fine grain sizes. MMCs The present work was conducted in order to determine the
have been reported to have attractive physical and mechanical effects of notch radius, including fatigue precracks, on the tough-
properties such as high-specific modulus, good high-cycle fatigue ness of NMMCs prepared by processing-induced devitrification
resistance, and improved thermal stability [3–7]. Among the MMCs, of amorphous Al-alloy powders. While the standard method to
particulate-reinforced MMCs are of particular interest due to their determine, the fracture toughness requires fatigue-precracked
ease of fabrication, lower cost, and isotropic properties. specimens, these specimens can require significant preparation
Combining the two concepts of MMCs and nano-crystalline time. Also, there are some applications where well-defined radii
materials in the form of nano-structured particulate-reinforced are present and the toughness under these conditions is needed.
MMCs has the potential to provide combinations of properties not In the present work, both room temperature and high tempera-
possible with conventional structural materials. For example, it has ture (T = 225 ◦ C) were studied, while testing was conducted and
been reported that the addition of only 2 wt.% of nano-sized SiC par- analyzed as done previously [15]. Comparison to conventional
ticles improves the yield strength of as-cast aluminum alloy A356 particulate-reinforced Al MMCs also presented [15].
by 50%, significantly higher than the strength improvements pro-
vided using the same amount of micrometer-sized particles [8].
Other work [9] has shown that the dispersion of 4 vol.% nano- 2. Experimental work

2.1. Materials
∗ Corresponding author at: Department of Design and Production Engineering,
Faculty of Engineering, Ain Shams University, Cairo, Egypt. Atomized Al89 Ni3 Gd7 Fe1 powders were canned in an Al tube
E-mail address: [email protected] (H.A. Hassan). and extruded producing an extruded billet 15.9 mm in diameter

0921-5093/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2008.06.048
 H.A. Hassan, John.J. Lewandowski / Materials Science and Engineering A 497 (2008) 212–215 213

with a 2 mm-thickness Al layer on the outside. An extrusion ratio

of 20:1 was used at 673 K. Hot extrusion of the amorphous pow-
ders results in the formation of an ultra-fine structure consisting
of roughly 35 vol% of nano-structured intermetallic particles (e.g.
100 nm-thick), in an aluminum matrix [27]. Elastic constants were
measured using acoustic techniques and revealed Young’s modulus
(E) in the range 86–89 GPa.

2.2. Specimen preparation

As-extruded bars were machined using electro-discharge

machining (EDM) into rectangular beams with a cross-section of
5 mm × 10 mm eventual for three-point bend (3PB) testing. Tough-
ness tests were conducted in three-point bending for both notched
and precracked samples. The notches were produced either by
grinding (i.e.  = 450 ␮m) or by using a low speed well vertical
diamond impregnated wire saw (i.e.  = 100 ␮m). The fatigue pre- Fig. 1. The effect of changes in notch root radius on the fracture toughness (KQ ) of
the nano-structured aluminum composite at room temperature (closed symbols)
cracks were placed in general accordance with ASTM standard
and 225 ◦ C (498 K) (open symbols).
E647-2000 [16]. All the notches and precracks were placed to a
depth of a/w = 0.5 according to the standards [17].
plots the data. The toughness at room temperature increased on
2.3. Testing procedure going from a fatigue precrack to 450 ␮m notch root radius. A sim-
ilar trend, but less in magnitude was exhibited at 225 ◦ C (498 K).
The toughness tests were carried out on a 20 Kip MTS closed Fig. 2 summarizes the effects of changes in test temperature on
loop servohydraulic machine using a MTS 458.20 controller, with the toughness values obtained for fatigue-precracked and notched
FTA control software. Fatigue precrack length was measured using samples. Increasing the test temperature increased the toughness
5 mm metallic foil KRAK© (KG-A05)-gages, monitored by a Frac- for both the fatigue-precracked and notched samples containing
tomat model 1288 crack measurement system. During fatigue the 100 ␮m notch. In contrast, increasing the test temperature
precracking, crack growth was measured at a frequency of 20 Hz, reduced the notch toughness for samples with the 450 ␮m notch.
load ratio of 0.1 under decreasing K conditions (K-control), Typical high-resolution SEM fractography for fatigue-
using an automated load-shedding technique. The fracture tough- precracked samples tested at room temperature are given in
ness tests were carried out under displacement rate control of Fig. 3. Samples containing 450 ␮m notches exhibited similar frac-
0.5 mm/min. The toughness tests were monitored and recorded tography. The fracture surfaces contained dimpled fracture with
by LABVIEW program. The high-temperature tests were conducted dimple sizes on the order of 0.49 ± 0.18 ␮m for fatigue-precracked
inside an ATS model controlled temperature cabinet with a tem- samples and 0.42 ± 0.10 ␮m for samples with the 450 ␮m notches.
perature control to within ±1 ◦ C using a MTS 409.8 temperature Examination of the dimples at high magnification revealed that
controller. After reaching the test temperature inside the chamber, each dimple contained multiple nuclei/particles (Fig. 3).
the specimens were held for 30 min to equilibrate prior to starting
the test. Load and temperature are monitored in addition to crack 4. Discussion
length via KRAK gages.
An increase in toughness with increase in notch radius has been
2.4. Fracture surface analysis documented on many materials systems [18–20], including MMCs
[15]. The fatigue-precracked samples tested at both room temper-
Fractured samples were examined using Philips XL30 or Hitachi ature and 225 ◦ C (498 K) can be considered as valid plane strain
S-4500 high-resolution SEM both operated at 15 KV.

3. Results

Table 1 summarizes the values for toughness (KQ and KIC ) at

catastrophic fracture as a function of the notch root radius for
all the specimens tested at RT and 225 ◦ C (498 K) while Fig. 1

Table 1
Effects of notch radius and test temperature on the fracture toughness KIC and KQ

Test temperature (◦ C) (K) Materials KIC and KQ (MPa m1/2 )

Notch root radius 

0 ␮m 100 ␮m 450 ␮m

32-B 9 26.8 40
25 (298 K) 40-B 5.6 27.6 45
41-B 9 30 49

32-B 16.5 31 N/A

Fig. 2. The effect of changes in test temperature on the fracture toughness (KQ )
225 (498 K) 40-B 12.5 29.7 34
obtained for fatigue-precracked (closed symbols) and notched samples, for notched
41-B 23.3 33 39.5
radii 100 ␮m (open symbols), and 450 ␮m (half-filled symbols).
214 H.A. Hassan, John.J. Lewandowski / Materials Science and Engineering A 497 (2008) 212–215

Table 2
Summary of the plastic zone sizes at the initiation of stable crack growth Ki and at the overload KIC , and the critical thickness for each sample according to ASTM E399 [17]
for tests at 225 ◦ C (498 K)

ID Ki (MPa m1/2 ) rp (ε) (mm) rp () (mm) KIC (MPa m1/2 ) rp (ε) (mm) rp () (mm) Bcrit (mm)

32-5B 9 0.014 0.043 16.5 0.05 0.143 2.3

40-6B 11 0.021 0.064 12.5 0.03 0.082 1.3
41-6B 10.7 0.020 0.060 23.3 0.10 0.286 4.5

ent volume fraction and/or size of reinforcement for composites

containing up to 20 vol.% reinforcement and particle sizes ranging
from 3 to 50 ␮m. In this model all particles are considered to crack
or decohere ahead of the major crack tip and at low strain. The
regions of plastic flow are limited to a volume of width “d” which
corresponds to interparticle spacing, where ı = JIC / f . For a homoge-
nous arrangement of particles, the interparticle spacing () can be
estimated using the following relation [25]:

 = 0.77dv−1/2

where d is the average particle size and v is the volume faction of

the reinforcement.
The present materials contain roughly 35 vol.% of fine (e.g.
100 nm-thick) intermetallics that have been shown to decrease the
grain size while increasing the modulus to 86–89 GPa. If the parti-
cles in the nano-structured alloy are uniformly distributed in the
Fig. 3. SEM image for (41-2B) nano-structured Al composite fatigue-precracked at
aluminum matrix, an interparticle spacing of roughly 0.13 ␮m is
RT. calculated. Assuming that the Rice and Johnson model [24] can
be used in such a situation indicates that the critical crack open-
fracture toughness, KIC , as the calculated plastic zone sizes (e.g. ing displacement calculated as CTOD = 0.49 K2 / y E, should be on
rp = 70 ␮m < B/50, where B = sample thickness) are well within the the order of the interparticle spacing for fracture link-up to occur
ASTM standards and the sample thickness exceed that required for (i.e.  = ı). The predicted ıcrit for each fatigue-precracked sample
a valid KIC (Table 2). Samples tested at room temperature all failed is provided in Table 3, using data for yield strength consistent
without stable crack growth in contrast to samples tested at 225 ◦ C with previous work on similar materials [27]. Also included in
(498 K) that exhibited stable crack growth. Table 2 summarizes both Table 3 are calculated critical crack opening displacements for each
the K at fracture initiation, Ki , as well as KIC for these cases. of the fatigue-precracked samples. The predicted ıcrit is below
The trend of these results are similar to previous work [15] on the experimental results for CTOD obtained presently. The dis-
MB78/SiC/20p MMC containing 15 ␮m average size SiCp in both agreement between the experimental data and that predicted by
underaged (UA) and overaged (OA) conditions, where a linear rela- the Rice and Johnson model [24] likely results from differences
tion between KQ and 1/2 was reported for both cases until reaching in the effective particle spacing for the nucleation of cavities
a limiting value of notch root radius. Fig. 4 summarizes the present ahead of the fatigue precrack tip. The Rice and Johnson model
data along with previous work [15] on particulate-reinforced MMCs [24] assumes uniformly distributed particles with the require-
containing 15 vol.% of 15 ␮m average size SiC particulates. In those ments that each particle participates in the fracture process and
early works on particulate-reinforced MMCs [15,21–23], the Rice creates one dimple on the fracture surface. This approach works
and Johnson model [24] of ductile fracture toughness was utilized reasonably well when applied to a variety of particle-containing
to rationalize the differences in toughness obtained with differ- materials when the volume fraction is low [26] as well as to
some particulate-reinforced MMCs containing less than 20 vol.%
of ␮m sized particulates [15,23], as each particle/reinforcement
particle in those studies typically were associated with one dim-
ple on the fracture surface. However, in the present case, SEM
analyses reveal that multiple particles were present in each dim-
ple on the fracture surface (cf. Fig. 3), suggesting that clusters of
nano-sized particles each nucleate a dimple, thereby increasing
the effective spacing of cavity-nucleating features (i.e. clustered
nano-particles). If one assumes that four to seven particles are
present in each dimple, consistent with fracture surface observa-
tions, the effective interparticle spacing () increases from 0.13 ␮m
to 0.5–0.9 ␮m, somewhat closer to the experimental data obtained
presently (Table 3). Fracture surface observations suggest that the
nano-sized particles debond from the matrix as no evidence of
cracked particles was found.
Increasing the test temperature was shown to increase the
toughness for the fatigue-precracked samples and those contain-
Fig. 4. Comparison between the effect of notch root radius on the fracture (KQ ) of ing 100 ␮m notches, although samples containing the 450 ␮m
the nano-structured aluminum composite and previous work [15] on MMC at RT. notches showed a reduced notch toughness on going from room
 H.A. Hassan, John.J. Lewandowski / Materials Science and Engineering A 497 (2008) 212–215 215

Table 3
Summary of the calculated and measured parameters from SEM fractography

ID Conditions KIC (MPa m1/2 ) CTOD (ı) (␮m) ıcrit = * (␮m) Dimple size (␮m) No. of particles/dimple

41-2B Precracked, RT (298 K) 9 0.7 0.13 0.49 ± 0.18 2.5 ± 1.5

41-6B Precracked, 225 ◦ C (498 K) 23.3 5.4 0.13 0.80 ± 0.2 4.0 ± 3.0

temperature to 225 ◦ C (498 K). In the case of increasing tough- edges the support of a Fulbright Scholarship for her work at CWRU
ness obtained on the fatigue-precracked samples tested at room and the assistance of Dr. D. Li for SEM.
temperature and 225 ◦ C (498 K), fracture surface observations
of the fatigue-precracked toughness tests conducted at 225 ◦ C References
(498 K) showed 0.8 ± 0.2 ␮m dimple sizes with up to seven par-
[1] E.O. Hall, Proc. Phys. Soc. Lond. Sec. B 64 (1951) 747–753.
ticles in each dimple (Table 3). Although the calculation of CTOD [2] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25–28.
(ı) at Kmax = KIC is 5.4 ␮m, when the initiation toughness of [3] T.W. Clyne, P.J. Withers, An Introduction to Metal Matrix Composites, Cam-
10.7 MPa m1/2 is used to calculate CTOD (ı) this value reduces bridge University Press, New York, 1995.
[4] S.C. Tjong, Z.Y. Ma, Mater. Sci. Eng. R 29 (2000) 49–113.
to 1.15 ␮m. Interestingly, the present samples do not undergo a
[5] D.S. Liu, M. Manoharan, J.J. Lewandowski, Metall. Trans. A 20 (11) (1989)
drop in toughness at intermediate test temperatures like that often 2409–2417.
exhibited in dispersion strengthened aluminum alloys or other [6] A.R. Vaidya, J.J. Lewandowski, Mater. Sci. Eng. A 220 (1/2) (1996) 85–92.
nano-crystalline systems, [27], at least at the loading rates presently [7] D.S. Liu, J.J. Lewandowski, Metall. Trans. A 24 (3) (1993) 609–615.
[8] Y. Yang, J. Lan, X. Li, Mater. Sci. Eng. A 380 (2004) 378–383.
used. [9] Y.C. Kang, S.L. Chan, Mater. Chem. Phys. 85 (2004) 438–443.
[10] X.F. Chen, E.G. Baburai, F.H. Froes, A. Vassel, Proc. Adv. Mater. Processes (1997)
5. Conclusions 185–192.
[11] S.L. Urtiga Filho, R. Rodriguez, J.C. Earthman, E.J. Lavernia, Mater. Sci. Forum
416–418 (2003) 213–218.
Processing-induced devitrification of amorphous aluminum [12] J.R. Groza, Int. J. Powder Metall. 35 (1999) 59–66.
alloy powder has been used to produce nano-crystalline composite [13] K. Akio, O. Atsushi, K. Toshiro, T. Hiroyuki, J. Jpn. Inst. Light Metal 49 (1999)
149–154.
materials with increased hardness, elastic modulus and tough- [14] D.V. Louzguine-Luzgin, A. Inoue, J. Nanosci. Nanotechnol. 5 (7) (2005)
ness. The effects of changes in notch radius and test temperature 999–1014.
on the toughness were generally consistent with previous work [15] M. Manoharan, J.J. Lewandowski, Int. J. Fract. 40 (1989) R31–R34.
[16] ASTM E647, Standard Test Method for Measurement of Fatigue Crack Growth
on particulate-reinforced MMCs where increases in notch radius Rates, ASTM, Philadelphia, PA, 2000.
produced increased values for toughness. However, fracture sur- [17] ASTM E399-2000, standard test method for plane strain fracture toughness of
face features revealed the participation of the multiple nano-sized metallic materials, American Society for Testing and Materials, Philadelphia,
PA, 2000.
particles in the formation of one dimple on the fracture surface. [18] J.J. Lewandowski, Mater. Trans. JIM 42 (4) (2001) 633–637.
Predications of the critical crack opening displacement for fracture [19] J.J. Lewandowski, M. Shazly, A.S. Nouri, Scripta Mater. 54 (3) (2006) 337–341.
were consistent with such observations. [20] J.F. Knott, Fundamentals of Fracture Mechanics, Butterworths, London, 1973.
[21] S.V. Kamat, J.P. Hirth, R. Mehrabian, Acta Metall. 37 (1989) 2395–2402.
[22] M. Manoharan, J.J. Lewandowski, Acta Metall. 38 (1990) 489–496.
Acknowledgements [23] M. Manoharan, J.J. Lewandowski, Mater. Sci. Eng. A 150 (1992) 179–186.
[24] J.R. Rice, M.A. Johnson, in: M.F. Kanninen, H. Liholt, O.B. Pedersen (Eds.), Inelastic
Behavior of Solids, McGraw-Hill, New York, NY, 1969, p. 641.
The authors thank DARPA/SAM program for providing financial
[25] G. LeRoy, J.D. Embury, G. Edwards, M.F. Ashby, Acta Metall. 29 (1981) 1509–1522.
support and acknowledge interactions with Drs. Tom Watson and [26] W.M. Garrison Jr., N.R. Moody, J. Phys. Chem. Solids 48 (11) (1987) 1035–1074.
Martin Blackurn of Pratt and Whitney Company. HAH also acknowl- [27] X.L. Shi, R.S. Mishra, T.J. Watson, Scripta Mater. 25 (2005) 887–891.