Materials Characterization 172 (2021) 110856

Contents lists available at ScienceDirect

Materials Characterization
journal homepage: www.elsevier.com/locate/matchar

The effect of microstructure and strain rate on the 25 ◦ C and 700 ◦ C

compression deformation behavior of powder metallurgy processed
Ti-45Al-2Nb-2Mn (at.%)-0.8 TiB2 (vol%) alloy
C.J. Boehlert a, b, *, D. Hernández-Escobar b, B. Ruiz-Palenzuela a, I. Sabirov c, J. Cornide a, E.
M. Ruiz-Navas a
a
Department of Materials Science and Engineering and Chemical Engineering, University of Carlos III of Madrid, Avda. Universidad 30, 28911 Leganés, Spain
b
Department of Chemical Engineering and Materials Science, Michigan State University, 428 South Shaw Lane, 2100 Engineering Building, East Lansing, MI 48824, USA
c
IMDEA Materials Institute, C/Eric Kandel 2, 28906 Getafe, Spain

A B S T R A C T

The effect of microstructure and strain rate on the room-temperature (RT) and 700 ◦ C compression deformation behavior of a powder metallurgy processed γ-TiAl
intermetallic alloy, Ti-45Al-2Nb-2Mn (at.%)-0.8 (vol%) TiB2, was investigated. Samples were heat-treated to obtain a duplex two-phase α2+γ microstructure and two
nearly fully-lamellar α2+γ microstructures with different lamellar spacings and γ-phase volume fractions. Compression experiments were performed to a minimum
deformation of 10% true strain under strain rates of 10− 2, 10− 3, 10− 4, and 10− 5 s− 1. The compression strength, strain rate sensitivity, colony size, interlamellar
spacing, and microhardness were dependent on microstructure. The nearly fully-lamellar microstructures exhibited higher compression strengths than the duplex
microstructure for all the testing conditions. The strain rate sensitivity index (m), tended to increase with increasing temperature, and for the 700 ◦ C deformation, m
increased with increasing true strain. The apparent activation volumes, decreased with increasing true strain at 700 ◦ C. Scanning electron microscopy observations
showed that cracking preferentially occurred within the γ phase and the extent of cracking increased with increased temperature, strain, and strain rate. Overall, the
RT deformation was considered to be controlled by dislocation glide, while at elevated temperature, the likely thermally activated process controlling dislocation
glide was associated with the forest junctions acting as pinning points.

1. Introduction toughness are intimately linked to the work-hardening capacity [5]. In

the past several decades, the effect of strain rate on the deformation
Two-phase intermetallic TiAl based alloys, consisting of α2 (Ti3Al, behavior of TiAl alloys has been investigated [6–9]. For example, the
hexagonal close packed (hcp), DO19) and γ (TiAl, face centered compression strength of a TiAl-Cr-V alloy increased at 1170 ◦ C when the
tetragonal (fct), L10) intermetallic compounds, have exhibited potential strain rate changed from 10− 3 to 102 s− 1 [6], and the flow stress of Ti-
as high-temperature structural materials due to their low density and 45Al-7Nb-0.4 W(at.%) increased with the increase in strain rate (from
excellent creep, oxidation, and corrosion resistance [1–4]. The 10− 3 to 10− 1 s− 1) for temperatures ranging between 1000 ◦ C and 1200
morphology of the microstructures, which is dependent on the pro­ ◦
C [7]. γ-TiAl alloys exhibit a superplastic behavior at temperatures
cessing conditions, has a critical influence on the mechanical behavior higher than 1000 ◦ C and low deformations [10]. This behavior can be
[3,4]. In general, microstructures containing only lamellar colonies, separated into three sections as a function of temperature, as shown by
termed fully-lamellar (FL), are endowed with high strength as well as U. Fröbel and Appel [11,12]. While these works were focused on the
excellent creep resistance [3,4]. Duplex microstructures, in which high-temperature deformation behavior, few studies have systemati­
different volume fractions of equiaxed γ grains are embedded in a matrix cally investigated the mechanical behavior of TiAl alloys over a wide
of lamellar colonies, are softer and usually possess higher RT elongation- range of strain rates at room temperature (RT). In particular, Shu et al.
to-failure (εf) [3,4]. For structural components, a reliable material [13] found that the addition of 2 (at.%) Mn (where the Mn was mostly, if
design requires an understanding of the mechanical properties under not always, in solid solution) increased the RT compression strength
different loading conditions. In particular, the work hardening behavior (sometimes up to 125 MPa) for strain rates between 10− 2 s− 1 and 10− 5
is an important factor in plastic deformation. Deformability, εf, and s− 1 .

* Corresponding author at: Department of Chemical Engineering and Materials Science, Michigan State University, 428 South Shaw Lane, 2100 Engineering
Building, East Lansing, MI 48824, USA.
E-mail address: [email protected] (C.J. Boehlert).

https://doi.org/10.1016/j.matchar.2020.110856
Received 14 June 2020; Received in revised form 23 December 2020; Accepted 23 December 2020
Available online 26 December 2020
1044-5803/© 2020 Elsevier Inc. All rights reserved.
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Although the majority of deformation studies on TiAl-based in­ minimum of 5% plastic deformation. After unloading, the samples were
termetallics have involved castings and ingot metallurgy (IM) [3,14,15], cooled at a cooling rate of 5 ◦ C/s. A thermocouple was spot welded in the
powder metallurgy (PM) processing routes have been more recently middle of the sample at its periphery, and the temperature was
explored for Ti-based intermetallics [16–20]. Some of the disadvantages controlled within ± 2 ◦ C of the targeted 700 ◦ C temperature throughout
associated with IM include casting defects, segregation of alloying ele­ the experiments. For some conditions, the experiments were repeated.
ments, and the need for post consolidation heat treatments [21,22]. After the deformation, the samples were sectioned along the longi­
Techniques, such as counter-gravity and centrifugal casting, have been tudinal direction and the cross section, using a slow-speed diamond
developed to overcome some of these drawbacks, though they require blade and metallographically polished using colloidal silica (60 nm) for
further heat treatment optimization [23]. PM offers several advantages the finishing polish. Secondary electron (SE) and backscattered electron
such as refining the microstructure, improving structural and chemical (BSE) SEM images, taken using either a Tescan Mira III (Brno, Czech
homogeneity, incorporating heat treatment into the consolidation cycle, Republic) Field Emission Gun (FEG) SEM, or a Philips XL30 SEM (now
and directly obtaining final product shape requiring minimal post pro­ Thermo Fisher Scientific Hillsboro, Oregon), were used to examine the
cessing treatments, thereby reducing manufacturing times. Moreover, phase distributions and their morphologies. Electron backscatter
Voisin et al. [24] has proven that PM can be successfully used for diffraction (EBSD) of the γ phase was performed using an EDAX-TSL
obtaining high quality TiAl turbine blades. (Mahwah, NJ, USA) EBSD system. Phase volume fractions, which were
The aim of this paper was to gain insight into the relationship be­ determined with an accuracy of ± 5% from approximately 20 BSE im­
tween the microstructure, strain rate, and the work-hardening behavior ages acquired at different magnifications for each microstructure, were
of PM processed Ti-45Al-2Nb-2Mn(at.%)-0.8TiB2(vol%) as the under­ measured using ImageJ software. Grain size was determined using the
standing gained will be useful for both researchers interested in mean line intercept method [26]. Energy Dispersive Spectroscopy (EDS)
processing-microstructure-property relationships and industries target­ analysis within the SEM was performed to analyze the local chemical
ing the PM version of this alloy for structural applications. Compression makeup of the microstructures.
tests were carried out to a minimum of 5% plastic strain using different For transmission electron microscopy (TEM) specimen preparation,
strain rates (10− 2, 10− 3, 10− 4 and 10− 5 s− 1) at RT and 700 ◦ C. the deformed samples were mechanically ground to a thickness of ~
Compression testing was chosen over tension because the strain rate 100 μm, followed by electropolishing at − 30 ◦ C, 40 V, for 10 s using an
sensitivity of the flow stress and the occurrence of individual deforma­ electrolyte consisting of 60% methanol, 34% butoxyethanol, and 6%
tion mechanisms can be studied over a wider range of strain rates than perchloric acid (vol%). TEM imaging was carried out in bright-field (BF)
tension, as compression does not restrict the low strain (~2–3%) plastic and high-angle annular dark field (HAADF) modes with a FEG S/TEM
deformation at high strain rates (10− 2 to 10− 3 s− 1). Prior to testing, the (Talos F200X, FEI) operated at an accelerating voltage of 200 kV. The
alloys were subjected to heat treatments to obtain different micro­ resulting TEM images were used, in conjunction with SEM images, to
structural morphologies containing different colony sizes (CS), lamellar estimate the lamellar spacings of the microstructures after deformation.
spacings, and phase volume fractions. Thus, the effects of microstruc­ The phases of the alloys were examined by X-ray diffraction (XRD)
tural parameters on both the properties and deformation will be using a Philips X’Pert-MPD System (Almelo, Netherlands) with Cu-Kα
highlighted. (λ=1.542 Å) radiation at a voltage of 40 kV and 40 mA.

1.1. Experimental procedures 2. Results

The TiAl intermetallic alloy studied in this work had a composition of 2.1. Prealloyed TiAl intermetallic alloy powder
Ti-30.80Al-4.80Nb-2.85Mn-0.25B (wt%), which is equivalent to Ti-
44.70Al-2.02Nb-2.03Mn-0.91B (at.%). The starting material was gas XRD confirmed that the primary phase present in the gas atomized
atomized prealloyed powders produced at the Helmholtz-Zentrum für powders after cooling to RT was the α-Ti3Al phase (a=2.8175 Å;
Material und Küstenforschung GmbH (Geesthacht, Germany) by elec­ c=4.6100 Å) [19]. The lack of the γ phase was consistent with the high
trode induction gas atomization (EIGA) [25]. The estimated cooling rate cooling rate that the powders experienced. The average size of the
reached during the EIGA process was ~105 ◦ C/s, which was lower than powders, which exhibited a typical spherical morphology, was 51 μm,
that considered for rapid solidification processes, but higher than that and the powder sizes measured at 10, 50, and 90% of the total powder
typical of conventional gas atomization. The powders were consolidated volume were 17.3, 51.2, and 117.8 μm, respectively [19]. Powders
by hot isostatic pressing (HIP) at 1200 ◦ C and 200 MPa during 4 h at the larger than 50 μm typically exhibited an elongated dendritic structure
Forschungszentrum Jülich (Jülich, Germany). Cylindrical specimens, likely due to the longer solidification period and slower cooling rates
nominally 15 mm in length and 10 mm in diameter, were electrodis­ they experienced compared with the smaller powders. The borides
charged machined from the as-HIP material. Some of the specimens precipitated with lacey, string-like shapes, likely due to lower solidifi­
were solution treated in a vacuum furnace (<10− 5 Torr) for 2 h at 1300 cation rates. EDS analysis confirmed the heterogeneous microstructure
◦
C followed by furnace cooling, hereafter referred to as HT1. Other of the powders, which exhibited uneven Nb and Mn concentrations [19].
specimens were vacuum sealed in quartz tubes and placed in a muffle Similar microstructures have been obtained by Yang et al. [27]. Overall,
furnace at 1300 ◦ C for 2 h followed by water quenching, where the the microstructural features were suggested to be a function of the in­
quartz tubes were broken upon entering the water. All of the water dividual powder cooling rates.
quenched samples were then heat treated in the same vacuum furnace The gas atomized powders were heated to 1000 ◦ C and XRD analysis
mentioned above; some at 850 ◦ C for 8 h followed by furnace cooling, was performed at different temperatures during the heating in order to
hereafter referred to as HT2, and others at 700 ◦ C for 8 h followed by understand the change in their crystal structure as a function of tem­
furnace cooling, hereafter referred to as HT3. The heating rate for all perature. The results showed a decrease in the intensities of the α peaks
samples heat-treated in the vacuum furnace was 10 ◦ C/min. at 600 ◦ C compared to those obtained before heating. A further increase
Uniaxial compression tests were carried out on the cylindrical in the temperature to 800 ◦ C led to a greater decrease of the α-phase
specimens using a Gleeble 3800 equipment (Dynamic Systems Inc., peaks combined with the appearance of both γ-TiAl and α2-Ti3Al peaks,
Poestenkill, New York, USA) at RT and 700 ◦ C at strain rates of 10− 5, suggesting that the metastable α phase started to transform into the α2+γ
10− 4, 10− 3, and 10− 2 s− 1. The compression axis was normal to the faces microstructure around 800 ◦ C [19]. At 1000 ◦ C, the transformation was
of the cylindrical specimens. For the elevated-temperature experiments, complete and the α2 and γ peaks’ intensities were even higher. Thus,
the specimens were heated from RT to 700 ◦ C at a heating rate of 20 ◦ C/ once the metastable powder is heated to temperatures above 800 ◦ C, γ
s, and after holding at 700 ◦ C for 10 s, the samples were compressed to a forms and increasing the temperatures leads to stabilization of the α2

2
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 1. (a) BSE SEM photomicrograph and (b) EBSD inverse pole figure and fcc unit triangle (with only the γ phase included) of the as-HIP microstructure. In (a) the
white boride-rich needles were dispersed within the α2 (bright) and γ (dark) phases.

and γ phases according to the equilibrium phase diagram [28–30].

Therefore, in order to achieve a complete transformation during the
consolidation of the previous metastable α powder, the sintering tem­
perature must be equal or greater than 1000 ◦ C.

2.2. As-HIP and as-HIP then heat-treated microstructures

The 1200 ◦ C HIP treatment transformed the metastable α phase into

the α2 and γ phases as confirmed by XRD [19]. The as-HIP material
developed a predominately γ microstructure, see Fig. 1. The grains
varied in diameter from 0.6–11.4 μm and the average grain size was
4.4±2.6 μm [19]. XRD revealed that the HT1 and HT2 microstructures
were composed of only the α2 and γ phases. The HT1 microstructure
exhibited a duplex microstructure combining lamellar colonies that
were 98±34 μm in diameter on average and equiaxed γ grains with an
average size of 4±1 μm, see Fig. 2 (a). The HT1 lamellar CS ranged from
45 to 205 μm. In contrast, the HT2 sample exhibited a refined near fully-
lamellar microstructure with an average CS of 33 μm, as observed in
Fig. 2 (b). The HT3 sample exhibited a refined near fully-lamellar
microstructure with an average CS of 23 μm, as observed in Fig. 2 (c).
The volume fractions of the γ phase in the HT1, HT2, and HT3 micro­
structures were 0.84, 0.98 and 0.52, respectively. The volume fraction of
the boride-rich phase in all the microstructures was less than 0.01. For
TiAl alloys containing 45-48Al (at.%), the equilibrium α2 and γ phase
volume fractions typically lie between 0.05–0.2 and 0.8–0.9, respec­
tively [31], and this was similar for the HT1 and HT2 microstructures.
However, the γ-phase volume fraction for the HT3 microstructure was
quite low (0.52) and this may have been a result of the lower temper­
ature aging.
Fig. 3 shows representative TEM images for the HT2 sample tested at
10− 2 s− 1 strain rate at RT. From the TEM analysis, the average lamellar
spacing independent of phase was ~ 12 ± 2 nm. It is noted that the SEM
resolution was not adequate to accurately measure the lamellae thick­
ness and therefore TEM was required. Though we did not perform TEM
on the HT3 microstructure, it is believed that the average lamellae
spacing was similar to that for the HT2 microstructure. Overall, in the
colonies of all the microstructures, the α2 lamellae had smaller widths
than those for the γ lamellae. In summary, the HT2 microstructure
exhibited the greatest volume fraction of the γ phase (0.98) and similar
CS to that for the HT3 microstructure.

2.3. Compression properties

The true stress versus true strain curves obtained from the
compression tests are presented in Fig. 4. It is noted that the absolute Fig. 2. BSE SEM photomicrographs of the microstructures (a) HT1, (b) HT2 and
values of the true stress and true strain values are plotted rather than the (b) HT3. The white boride-rich needles were dispersed within the α2 (bright)
actual values. In general, the strengths decreased with increasing strain and γ (dark) phases.

3
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

2 1
Fig. 3. (a) BF and (b) HAADF TEM images of the microstructure of the HT2 sample after deformation to ~15% true strain during compression testing at 10− s−
strain rate at RT.

rate. However, this was not the case for the RT deformed samples which The n values calculated from the Hollomon and modified Ludwik ap­
underwent the HT2 heat treatment, where the strength was not depen­ proaches are given in Table 1. The n values obtained from the Hollomon
dent on the strain rate and a near 0 value of the strain rate sensitivity equation ranged between 0.076 and 0.248, whereas those obtained from
index, m, was measured. The strengths decreased with increasing tem­ the modified Ludwik equation ranged between 0.502 and 0.849. The n
perature. This decrease in strength was more dramatic for the near fully- values obtained from the modified Ludwik approach were always larger
lamellar microstructures (i.e. HT2 and HT3 samples). The heat-treated than those obtained from the Hollomon approach. Overall, there was no
HT3 microstructure exhibited a larger drop in true flow stress at 700 clear dependence of the n values on the strain rate, for any of the two
◦
C compared with the HT2 microstructure. approaches. However, the n values obtained from the modified Ludwik
The strengths of the HT2 and HT3 samples (with near fully-lamellar approach tended to increase with increasing strain rate for samples
microstructures) were significantly greater than those for the HT1 tested at 700 ◦ C, and they were generally lower than those for samples
samples (with duplex microstructures) under similar conditions. For the tested at RT.
700 ◦ C deformation, the true flow stresses of the HT2 microstructures
were at least 100 MPa higher than those of the HT3 microstructures and
at least 115 MPa higher than those of the heat-treated HT1 duplex mi­ 2.5. Strain rate sensitivity
crostructures, see Table 1.
The true stress, true strain values obtained from the compression
tests were used to calculate m. To determine this parameter, several
2.4. Strain hardening exponent methods can be used, such as determining the slope of the log(σ)-log(ε̇)
from instantaneous values, changing the deformation rate, or by per­
The strain hardening exponent (n) is an important parameter that forming stress-relaxation tests [11–13,34–36].
reflects the work-hardening effect of the material during plastic defor­ In the current work, the m values were calculated for a constant
mation. It is readily apparent in Fig. 4 (a-f) that the true stress-strain strain and test temperature using the following equation: [37,38].
curves of the tested materials exhibited work-hardening. The n values [ ]
were determined by two different approaches. The Hollomon relation­ ∂ ln(σ )
m= (3)
ship, which is the conventional and most widely used approach to ∂ ln(ε̇) ε,T
evaluate the flow stress of a material in the plastic deformation region, is
given by: where σ is the true flow stress and ε̇ is the corresponding strain rate. The
m values were determined from the slope of the ln(σ)-ln(ε̇) curve formed
σ = K εn (1)
by the true flow stress values at the following true strain values: 5%, 6%,
7%, 8%, 9%, 10% and 11%, as observed in Fig. 5. In this way, the m
where K is the strength coefficient, σ is the true stress, and ε is the true
values were obtained as a function of the true strain and test tempera­
strain. Only the plastic regime is used in the curve fitting to determine n.
ture, and they are tabulated in Table 2. The m values for the HT1 samples
To better quantify n, Afrin et at. [32] proposed the relationship in Eq. (2)
ranged between 0.008 and 0.015 at RT and between 0.010 and 0.039 at
below, which excludes the elastic portion of the true stress-strain curve:
700 ◦ C. The m values for the HT2 samples ranged between − 0.032 and
( ) ( )n
σ − σ y = K ε − εy (2) − 0.012 at RT and between 0.025 and 0.054 at 700 ◦ C. The m values for
the HT3 samples ranged between − 0.028 and − 0.003 at RT and between
where σ y is the yield stress and εy is the strain at yielding. As this was 0.005 and 0.032 at 700 ◦ C.
derived from the Ludwik equation [33], it will be referred as the Fig. 6 (a, b and c) illustrates the evolution of m values as a function of
modified Ludwik approach hereafter. The yield stress and the strain at true strain for different heat-treated samples and testing temperatures. It
yielding were determined from the conventional 0.2% strain offset of the was noted that for HT2 and HT3 samples, the m values obtained at 700
true stress-strain curves and used as inputs in the curve fitting of Eq. (2). ◦
C were always larger than those at RT, which was also true for HT1

4
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 4. True stress vs. true strain curves as a function of strain rate for samples (a and b) HT1, (c and d) HT2 and (e and f) HT3, obtained from compression tests at RT
and 700 ◦ C. Note that the stress and strain are portrayed as positive values in these plots.

samples above 7% true strain. In particular, the m values of HT1 samples ( )

∂ lnε̇ MkT
tested at 700 ◦ C increased gradually with increasing strain, however, V * = MkT = (4)
they decreased slightly with increasing strain for samples tested at RT.
∂σ mσ
The trend of increasing m values at increasing strain held true for HT2
where M is the Taylor factor (equal to 3.06 for random textures), k is the
and HT3 samples at both testing temperatures.
Boltzmann’s constant, T is the absolute temperature, σ is the true flow
stress, ε̇ is the strain rate, and m is the strain rate sensitivity. The V*
2.6. Apparent activation volume
values were determined in the range of 5–11% true strain, which be­
longs to the plastic regime, for each of the four strain rates tested, and
The apparent activation volume, V*, of each of the microstructures
their average values are provided in Table 2, assuming a Burger’s vector,
was determined by the isostrain method using the following equation:
b, of 0.283 nm [39]. Note that V* values from HT2 and HT3 samples
[37].
tested at RT were negative because of the corresponding negative m

5
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Table 1
Summary of the true yield stresses/strains, and strain hardening exponents (n) (based on Hollomon and modified Ludwik approaches) obtained from the compression
data.
Heat treatment Test temperature Strain rate (s− 1) True yield stress (MPa) True yield strain n

Hollomon Modified Ludwik

2
HT1 RT 10− 475 0.036 0.203 0.770
3
10− 550 0.050 0.172 0.808
4
10− 390 0.029 0.248 0.752
5
10− 490 0.045 0.212 0.849
2
700 ◦ C 10− 468 0.045 0.173 0.784
3
10− 422 0.044 0.177 0.722
4
10− 395 0.038 0.151 0.665
5
10− 372 0.028 0.094 0.502
2
HT2 RT 10− 655 0.051 0.153 0.749
3
10− 860 0.077 0.115 0.710
4
10− 797 0.060 0.140 0.761
5
10− 774 0.052 0.153 0.748
2
700 ◦ C 10− 657 0.053 0.141 0.739
3
10− 607 0.052 0.155 0.763
4
10− 628 0.064 0.076 0.611
5
10− 487 0.043 0.077 0.535
2
HT3 RT 10− 730 0.072 0.156 0.803
3
10− 805 0.083 0.140 0.817
4
10− 785 0.077 0.139 0.814
5
10− 465 0.036 0.241 0.820
2
700 ◦ C 10− 475 0.044 0.204 0.791
3
10− 465 0.043 0.201 0.796
4
10− 415 0.032 0.203 0.771
5
10− 387 0.036 0.148 0.696

values, and thus, they were not included in Table 2 as they lack physical 3. Discussion
meaning. Fig. 7 shows the evolution of the average V* values, calculated
from 700 ◦ C tests for all strain rates, with increasing true strain, and the 3.1. Compression strength as function of temperature and strain rate
error bars represent their respective standard deviations.
The V* values, which ranged between 12.5 and 195.92 b3 for samples The results indicate that the near fully-lamellar microstructures were
tested at 700 ◦ C, decreased with increasing true strain, as observed in significantly stronger than the duplex microstructure, consistent with
Fig. 7. These observations are similar to those from Kashyap et al. [34] the literature [3,4,19]. The deformation temperature and strain rate had
who performed tensile tests at temperatures from 100 to 350 ◦ C, which a significant effect on the true flow stress under all of the conditions
represents 0.4–0.65Tm, for strain rates from 10− 4 to 10− 1 s− 1 for an tested. In general, the true flow stress increased with increasing strain
ultrafine grained Al alloy. The 700 ◦ C compression experiments were rate and decreasing test temperature, as similarly observed for a high
estimated to be at 0.55Tm, which is within the normalized temperature strength low alloy steel at strain rates from 0.1 to 10 s− 1 between 800
range of those for Kashyap et al. [34]. The V* values at RT tended to and 1100 ◦ C [36]. This is because the dynamic softening dominates
slightly increase with increasing true flow stress. Thus, there was a work hardening at higher temperatures [42,43]. Subsequently, the flow
different dependence at lower temperatures compared with elevated stress decreases with increased temperature and decreased strain rate
temperatures. until a balance with work hardening is achieved, and this explains the
dynamic softening. The reason that the true flow stress values decrease
with increasing deformation temperature is because higher tempera­
2.7. Compression cracking behavior tures and lower strain rates provide a longer time for energy accumu­
lation and higher dislocation mobilities [44].
In order to investigate the deformation behavior, BSE SEM photo­ Deformation at high strain rates/low temperatures promotes the
micrographs of samples tested at RT and 700 ◦ C (for all strain rates after accumulation of dislocations within ultra-fine grains resulting in
reaching at least 11% true strain) are provided in Figs. 8, 9 and 10 for increased strain hardening [45] and the enhancement of the strain rate
HT1, HT2 and HT3 samples, respectively. From Figs. 8 and 9, it is sensitivity to the flow stress [46]. It has been suggested that increased
apparent that the HT1 and HT2 samples tested at 700 ◦ C resulted in strain rate sensitivity to the flow stress increases uniform elongation by
more cracking at a given strain level compared with those tested at RT. delaying macroscopic localization of plastic deformation [46,47].
In general, the cracking tended to be confined within the γ phase more Increased strain rate sensitivity has been attributed to the occurrence of
than the α2 phase, and this was especially true for the duplex micro­ grain boundary sliding and grain boundary diffusion in nanostructured
structure HT1, as observed in Fig. 8. Overall, the extent of cracking metals [48,49], where both mechanisms can be active in ultrafine
increased with increased temperature, strain, and strain rate. It is also grained and nanostructured metals during plastic deformation, even at
noteworthy that in the HT1 samples tested at 700 ◦ C, the boride phase RT [48,49]. The higher m values observed at 700 ◦ C in the current work
exhibited more cracking compared with that at RT, see Fig. 8. Similar may be a result of sliding along the fine lamellae where grain boundary
boride phase cracking has also been observed in conventional α+β Ti diffusion may have played a significant role as will be discussed later.
alloys deformed at both RT and 455 ◦ C [40,41].

6
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 5. Evolution of true stress as a function of strain rate, in the range of 5–11% true strain, used to obtain the indicated strain rate sensitivity values, m, for samples
(a and b) HT1, (c and d) HT2 and (e and f) HT3 obtained from compression tests at RT and 700 ◦ C, respectively.

7
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Table 2
Summary of strain rate sensitivity exponents (m) and apparent activation volumes (V*) determined from compression experimental data. Note that V* is expressed in
terms of the Burger’s vector (b), which is assumed to be 0.283 nm [39].
True Strain (%) m V* x b3

HT1 HT2 HT3 HT1 HT2 HT3

RT 700 ◦ C RT 700 ◦ C RT 700 ◦ C RT 700 ◦ C RT* 700 ◦ C RT* 700 ◦ C

5 0.015 0.010 − 0.032 0.025 − 0.028 0.005 69.01 122.68 na 39.87 na 195.92
6 0.014 0.013 − 0.025 0.031 − 0.019 0.011 62.66 80.69 na 28.10 na 85.73
7 0.014 0.021 − 0.02 0.036 − 0.013 0.015 61.34 48.05 na 22.00 na 55.07
8 0.011 0.025 − 0.016 0.039 − 0.009 0.018 69.30 39.09 na 18.92 na 44.74
9 0.009 0.030 − 0.014 0.046 − 0.006 0.022 87.26 31.11 na 15.66 na 33.63
10 0.009 0.036 − 0.013 0.051 − 0.004 0.028 80.06 25.36 na 13.73 na 26.06
11 0.008 0.039 − 0.012 0.054 − 0.003 0.032 82.98 22.65 na 12.50 na 21.97
*
na: not applicable.

3.2. Strain rate sensitivity and strain hardening exponent obtained from isothermal compression of Ti-6Al-4 V varied from ~ 0.10
to 0.40 for testing temperatures from 830 to 1030 ◦ C [8], which are
On the one hand, m is typically used to accurately predict the comparable to the m values obtained in this work at 700 ◦ C.
deformation behavior of materials during loading [50]. On the other The factors that have been shown to have the greatest influence on
hand, n can be used to rationalize the amount of uniform plastic strain the m values are grain size, phase distribution, impurity content, su­
which the material can withstand before failure [8]. Since the elastic persaturation of vacancies, and deformation temperature [12]. The
regime of the tested samples extends up to about 5% true strain, the m different responses of the same alloy at the two different temperatures
and n values were only considered meaningful above that value. In the examined in the current work are expected to be a result of the micro­
current work, a difference in the m values was observed for each of the structural differences. The duplex microstructure exhibited less differ­
microstructures at the two different temperatures investigated, where ences between the m values measured at RT and 700 ◦ C. Thus, it is
higher m values were exhibited at 700 ◦ C as compared to RT. This in­ expected that the larger fraction of equiaxed γ grains exhibited by the
dicates a higher sensitivity to the deformation speed at elevated tem­ duplex microstructure played a role in this behavior. Although larger
peratures, as has been observed in other structural polycrystalline alloys differences in the m values at the different temperatures were exhibited
[34]. by the near fully-lamellar microstructures, the difference in m values for
The n values determined from the Hollomon and modified Ludwik, the HT3 microstructure was smaller than that for the HT2 microstruc­
confirmed the consistently higher values of the latter, in agreement with ture, and this may have been a result of the lower volume fraction of the
a previous work [32]. In general, the n values retrieved from the γ phase exhibited by this microstructure. Thus, the softer γ phase may be
modified Ludwik approach decreased with increasing deformation more sensitive to temperature changes than the harder α2 phase.
temperature, as observed in Fig. 6 (d, e and f). The n value results from a The HT2 microstructure exhibited the greatest difference in the m
competition between work hardening and thermal softening, such that a values at the two different testing temperatures. Particular noteworthy
decrease in softening (i.e. relative increase in work-hardening) leads to is that the m values at RT were close to 0 in the HT1 sample, and
an increase in the n value [51]. This phenomenon has been observed for negative in the HT2 and HT3 samples. Negative m indicates a softening
Ti-6Al-4 V [8], Ti60 [51] and Ti17 alloy [52]. Materials that exhibit a mechanism influenced by dynamic strain aging [56], which has been
relative decrease in work hardening with increased temperature tend to reported earlier in Ti alloys undergoing stress-induced martensitic
exhibit a decrease in n with increased temperature [53,54]. It is also transformation during deformation [57]. For two-phase TiAl in­
evident from Fig. 6 (d, e and f) that, for samples tested at 700 ◦ C, the n termetallics, at RT the dislocation velocity has been found to be
values increased with increasing strain rate, and this is consistent with controlled by a combination of lattice friction, jog dragging, and local­
the trend observed in Ti60 compression tested at 960 ◦ C from 10− 3 to ized pinning [12]. The glide resistance provided by these mechanisms
10− 1 s− 1 strain rates [51]. gradually decreases with increasing temperature. Deformation at tem­
The m values measured at RT in the current work were similar to peratures above 1173 ◦ C is dominated by dislocation glide and climb,
those measured at RT by others. For example, Shu et al. [13] measured m with the relative contributions depending on temperature and strain rate
values ranging between 0.009 and 0.024 at RT (for strains of 0.2%, [58]. Deformation in the 450–477 ◦ C range for TiAl intermetallics is
1.0%, and 2.0%) for both TiAl and TiAl-2Mn(at.%), where the Mn was characterized by discontinuous yielding, negative strain rate sensitivity,
mostly, if not fully, in solid solution. Like that observed in Shu et al. and strain aging effects. These phenomena are usually associated with
[13], the m values of the HT1 sample decreased with increasing plastic dislocation locking according to the Portevin-Le Chatelier effect [59]
strain at RT (see Fig. 6d). However, the m values of the HT2 and HT3 and this has been assumed to be the case for TiAl [60]. Thus, low m
samples tested at RT increased gradually with increasing plastic strain values (and even negative m values) are normally associated with the
(see Fig. 7), which might be due to a transition from dislocation- locking of dislocations due to the Portevin-Le Chatelier effect and
dominated plasticity to a combination of grain boundary diffusion and discontinuous yielding [12].
sliding, as proposed elsewhere [55]. Sabirov et al. [35] studied the RT Although this has been observed in TiAl intermetallics between 723
compression behavior of ultrafine grain Al alloy Al6082 and the m and 1023 ◦ C, the mechanisms are not completely understood due to the
values decreased from 0.026 for a strain rate of 10− 5 s− 1 to 0.013 for a microstructural complexity in such alloys. m values near 0 have also
strain rate of 10− 2 s− 1. been observed in other alloys. For example, Dudamell et al. [61] found
The m values determined at 700 ◦ C in the present work for all the low m values (and even negative m values) at 0.1 strain for Mg-1Mn-1Nd
heat treated samples increased with increasing deformation temperature (wt%) in 200 ◦ C compression tests for strain rates between 5⋅10− 4 s− 1
and plastic strain (see Fig. 6), this is consistent with the literature. For and 5⋅10− 2 s− 1, and this was related to dynamic strain aging and
example, the m values were ~ 0.13 at 100 ◦ C, and ~ 0.23 at 250 ◦ C for deformation-induced twinning. When dislocations move with a certain
ultrafine grained Al-Mg-Si [34]. During isothermal compression of the velocity associated with a certain strain rate, if the solute diffuses at a
Ti60 alloy at 980 ◦ C, the reported m values ranged between ~ 0.15 and similar velocity, then the solutes can attach to dislocations and align
0.5 for strain rates between 10− 3 and 10 s− 1. Similarly, the m values along them causing dislocation drag on a gliding dislocation. This will

8
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 6. Evolution of the strain rate sensitivity values, m, as a function of the true strain for (a) HT1, (b) HT2 and (c) HT3 samples and evolution of the strain
hardening exponents, n, determined from the modified Ludwik relationship, as a function of the strain rate for (d) HT1, (e) HT2 and (f) HT3 samples during
compression testing at RT and 700 ◦ C.

cause hardening. However, when the strain rate is fast enough to allow a 3.3. Apparent activation volume
gliding dislocation to jump into another dislocation, then this will result
in less hardening because there will be less dislocation drag as the The activation volume (V*) is widely used to determine the possible
dislocation that jumped will jump out of the solutes that are dragging it. plastic deformation mechanisms [62]. Like the m values, the V* values
Thus, with increased strain rates, softening and serrations may occur, were only considered meaningful within the regime of plastic defor­
which could lead to low or even negative m values. Therefore, this may mation, which begins at approximately 5% true strain for all samples
explain the negative m values obtained in the HT2 and HT3 samples tested in the current work. In Shu et al. [13], the average V* values of the
deformed at RT. TiAl and TiAl-2Mn(at.%) microstructures were 22.5 b3 and 24.6 b3,
respectively. The V* calculations they used were based on a constant
that was approximately half that used in the current work. Incorporating
their data into Eq. (4), their revised V* values (~ 45–50 b3) fall within

9
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

1 Q
ln = ln(A) + ln(σ) − (7)
m RT
In the current work, Q values were calculated for a constant strain
rate, by applying the partial derivative of Eq. (7) as follows:
⎡ ⎤ ⎡ ⎤
R ⎢ ln(σ ) ⎥
⎢∂ ( R ⎢ ln(σ2 /σ1 ) ⎥
Q= ) ⎥ε̇ = ⎢ ( )⎥⎦ε̇ (8)
m ⎣ 1 ⎦ m ⎣ 1 1
∂ T T2
− T1

where σ2 and σ1 are the values of the true flow stress corresponding to
the testing temperatures T2 and T1, respectively. The Q values were
determined using T1=700 ◦ C, T2=RT=25 ◦ C, their respective true flow
stress values at 10% true strain, and the calculated m values. The
computed Q values ranged between 19–38 kJ/mol, 6–36 kJ/mol and
16–47 kJ/mol for HT1, HT2 and HT3, respectively. For the three HT
conditions, there is a consistent trend of increasing Q with decreasing ε̇,
Fig. 7. Evolution of the activation volume (V*) in the range of 5–11% true which agrees with that obtained from compression tests performed at
strain for HT1, HT2 and HT3 samples during compression testing at 700 ◦ C. 850–1150 ◦ C in Ti-15-3 at 10− 3-10 s− 1 strain rates [67]. No significant
Data points and error bars represent, respectively, the average V* values and differences were observed in the Q values obtained at different flow
their corresponding standard deviations for tests performed at 10− 5 s− 1, 10− 4 stress values within the range of 5–10% true strain.
s− 1, 10− 3 s− 1 and 10− 2 s− 1 strain rates. Note that V* is expressed in terms of the For TiA15 (Ti-6Al-2Zr-1Mo-1 V(wt%)), Yasmeen et al. [68]
Burger’s vector (b), which is assumed to be 0.283 nm [39].
measured the Q values for tensile deformation between 880 and 920 ◦ C
and strain rates of 10− 2 s− 1 to 5⋅10− 4 s− 1. In the early stages of defor­
the range of V* values reported here for samples tested at 700 ◦ C for mation, they measured Q=256 kJ/mol (at a true strain of 0.25) and
0.07–0.08 true strains, as observed in Table 2 and Fig. 7. According to Q=276 kJ/mol (at true strain of 1). In the current work, the Q values
the literature [63,64], such low V* values are often associated with ranged between 6 and 47 kJ/mol, and there was no clear relationship
dislocation glide. between deformation or microstructure and the Q values. These values,
An interesting feature of plastic deformation is the decrease of the which are roughly half of the self-diffusion activation energy values for
activation volume with increasing true stress which occurred for the both α-Ti (204 kJ/mol) and β-Ti (161 kJ/mol) [69], do not lie in the
700 ◦ C deformed samples. This behavior, particularly the inverse pro­ range of activation energy values for dislocation creep of α-Ti (200–360
portionality of the activation volume to stress, is consistent with the kJ/mol) [70]. Thus, the activation energy values more represented grain
assumption that the thermally activated process controlling dislocation boundary diffusion processes suggesting that the lamellar boundaries
glide is associated with the forest junctions acting as pinning points may be playing an important role in the deformation behavior. It has
[37]. The magnitude of the activation volume (i.e. between 0 and 200 been shown that the dislocations can move by glide or climb, depending
b3, see Table 2) is semi-quantitatively consistent with the forest junction on the temperature range, and that some dynamic strain aging occurs at
mechanism. Assuming a V* value of 50 b3, the spacing, L, between the intermediates temperatures [71]. In lamellar alloys, the interaction at
pinning points estimated on the basis of the formula [65]. interfaces is found to play a determining role [71]. As mentioned pre­
2
( )( )
L d viously, the higher m values measured during the 700 ◦ C deformation
V * = b3 (5) may be a result of sliding along the fine lamellae, where grain boundary
3 b b
diffusion may have played a significant role, and this was supported by
is 400b, if the stress-dependent width of the junction, d, is estimated to the low Q values, which are more representative of grain boundary
be 0.1875b. The typical spacing between the forest junctions, ρ-1/2–10− 7 diffusion.
m ~ 400b; is quantitatively similar. Thus, the overcoming of the forest
junctions by gliding dislocations can be considered as the most plausible
thermally activated glide mechanism. It is noted that full thermal acti­ 3.5. Deformation observations
vation analysis, namely, evaluation of the activation energy for the
mechanism governing dislocation glide as well as the dynamic recovery The cracking, which preferentially occurred within the equiaxed γ
mechanism, which controls steady-state deformation, generally is a phase, could be due to the buildup of stress concentrations caused by
useful tool to gain further insight into the possible deformation mech­ dislocation pileup, both within the grains and at the grain boundaries
anisms [66]. and lamellar phase interfaces. The balance between work hardening and
dynamic softening led to lower stresses achieved at elevated tempera­
3.4. Activation energy tures, as explained earlier, and the buildup of the stress concentrations
could have lead to the cracking, which occurred to a lesser extent at RT.
To calculate the activation energy, the relationship between the flow The γ phase in particular exhibited a greater extent of cracking than the
stress (σ), ε̇, and T during hot deformation can be expressed using the lamellar colonies. This is not unexpected as the γ phase is softer than the
following relationship: α2 phase and therefore would be expected to crack prior to the α2 phase.
( ) In addition, the limited deformation exhibited by α2 Ti3Al-based alloys,
Q which can be attributed to the lack of independent slip systems, has been
ε̇ = A σ (1/m) exp − (6)
RT well-documented in the literature [72]. Hence, the von Mises criterion is
not satisfied. Thus, the deformation of α2+γ TiAl alloys is considered to
where ε̇ is the strain rate, σ is the true flow stress, m is the strain rate mainly be carried by the γ phase, and this could explain the preferential
sensitivity, A is a material constant, R is the universal gas constant, T is cracking exhibited within this phase. In addition, because most of the
the absolute temperature and Q is the activation energy. Taking natural deformation is considered to be carried by the γ phase, the activation
logarithm on both sides of Eq. (6), it can be expressed as: volume analysis used is this work, with only the γ phase being consid­
ered, is justified.

10
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 8. BSE SEM photomicrographs of the cross-section of samples HT1 after deformation to ~12% true strain at (a and b) 10–5 s− 1, (c and d) 10–4 s− 1, (e and f) 10–3
s− 1 and (g and h) 10–2 s-1 strain rates during compression testing at RT and 700 ◦ C, respectively. Note the white boride-rich needles dispersed within the α2 (bright)
and γ (dark) phases. Cracking preferentially occurred in the γ-phase in (a and b). Note the cracking of the boride-rich needles in (d). Cracking preferentially occurred
in the γ-phase in (e, f, g and h), and more cracking was exhibited after the 700 ◦ C deformation.

3.6. Summary and conclusions the microstructures evaluated. For all the strain rates and temperatures
examined, the near fully-lamellar microstructures exhibited higher
Duplex and near fully-lamellar Ti-45Al-2Nb-2Mn(at.%)-0.8TiB2(vol strengths than the duplex microstructure.
%) microstructures, obtained through different heat treatments, were The strain rate sensitivity exponent tended to increase with
subjected to compression tests at strain rates ranging between 10− 2 s− 1 increasing temperature. For the 700 ◦ C deformation, m increased with
to 10− 5 s− 1 at RT and 700 ◦ C. The strain rate sensitivity of the flow stress increasing true strain. SEM observations showed that cracking prefer­
was estimated and the associated deformation mechanisms were dis­ entially occurred within the γ phase and the extent of cracking increased
cussed. With increased strain rate, the flow stress increased for each of with increasing temperature, strain, and strain rate. The activation

11
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 9. BSE SEM photomicrographs of the cross-section of samples HT2 after deformation to ~13–15% true strain at (a and b) 10–5 s− 1, (c and d) 10–4 s− 1, (e and f)
10–3 s− 1 and (g and h) 10–2 s− 1 strain rates during compression testing at RT and 700 ◦ C, respectively. Note the white boride-rich needles dispersed within the α2
(bright) and γ (dark) phases. More cracking was exhibited after the 700 ◦ C deformation.

12
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

Fig. 10. BSE SEM photomicrographs of the cross-section of samples HT3 after deformation to ~15–18% true strain at (a and b) 10–5 s− 1, (c and d) 10–4 s− 1, (e and f)
10–3 s− 1 and (g and h) 10–2 s− 1 strain rates during compression testing at RT and 700 ◦ C, respectively. Note the white boride-rich needles dispersed within the α2
(bright) and γ (dark) phases.

13
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

volume was inversely proportional to the flow stress at 700 ◦ C. This [16] P.R. Smith, A. Rosenberger, M.J. Shepard, R. Wheeler, Review a P/M approach for
the fabrication of an orthorhombic titanium aluminide for MMC applications,
dependence was shown to be qualitatively consistent with the assump­
J. Mater. Sci. 35 (2000) 3169–3179.
tion that dislocation glide is controlled by the thermally activated [17] P.R. Smith, A. Rosenberger, M.J. Shepard, Tape cast second generation
overcoming of forest junctions. Overall, the RT deformation was orthorhombic-based titanium Aluminide alloys for MMC applications, Scr. Mater.
considered to be controlled by dislocation glide, while at elevated 41 (1999) 221–228.
[18] C.J. Cowen, C.J. Boehlert, Comparison of the microstructure, tensile, and creep
temperature, the likely thermally activated process controlling disloca­ behavior for Ti-22Al-26Nb(at%) and Ti-22Al-26Nb-5B(at%), Met. Trans A. 38 (1)
tion glide was associated with the forest junctions acting as pinning (2007) 26–34.
points. [19] R. Muñoz-Moreno, In Situ Analysis of the High Temperature Deformation and
Fracture Mechanisms of a Gamma-TiAl Alloy (PhD dissertation), Universidad
Carlos III, Madrid, Spain, 2014.
Data availability [20] R. Muñoz-Moreno, E.M. Ruiz-Navas, B. Srinivasarao, J.M. Torralba,
Microstructural development and mechanical properties of PM Ti-45Al-2Nb-2Mn-
0.8 vol.%TiB2 processed by field assisted hot pressing, J. Mater. Sci. Technol. 30
The raw data required to reproduce these findings cannot be shared (11) (2014) 1145–1154, https://doi.org/10.1016/j.jmst.2014.08.008.
at this time as the data also forms part of an ongoing study. [21] K. Matsugi, N. Ishibashi, T. Hatayama, O. Yanagisawa, Microstructure of spark
sintered titanium-aluminide compacts, Intermetallics. 4 (1996) 457–467.
[22] E. Schwaighofer, H. Clemens, S. Mayer, J. Lindemann, J. Klose, W. Smarsly,
Declaration of Competing Interest V. Guther, Intermetallics. 44 (2014) 128–140.
[23] M. Thomas, J.L. Raviart, F. Popoff, Intermetallics. 13 (2005) 944–951.
[24] T. Voisin, J.P. Monchoux, L. Durand, N. Karnatak, M. Thomas, A. Couret, An
The authors declare that they have no known competing financial innovative way to produce γ-TiAl blades: spark plasma sintering, Adv. Eng. Mater.
interests or personal relationships that could have appeared to influence 17 (10) (2015) 1408–1413.
the work reported in this paper. [25] R. Gerling, H. Clemens, F.P. Schimansky, Powder metallurgical processing of
intermetallic gamma titanium aluminides, Adv. Eng. Mater. 6 (1–2) (2004) 23–37.
[26] Standard Test Methods for Determining Average Grain Size, ASTM Designation
Acknowledgments E112–96-3, American Society for Testing and Materials, West Conshohocken, PA,
2004.
[27] C. Yang, D. Hu, A. Huang, M. Dixon, Microstructures and tensile properties of hot
This work was partially supported by a Fulbright Senior Scholar isostatic pressed Ti4522XD powders, Intermetallics. 32 (2013) 64–71.
Award. The authors are grateful to Dr. Rocio Muñoz-Moreno (Hewlett [28] J.J. Valencia, C. McCullough, C.G. Levi, R. Mehrabian, Solidification
Packard, Barcelona, Spain) for sample preparation and taking some of microstructure of supercooled Ti-Al alloys containing intermetallic phases, Acta
Metall. 37 (9) (1989) 2517–2530.
the EBSD and SEM images. The authors are grateful to Ms. Amalia San [29] T. Voisin, L. Durand, N. Karnatak, S. Le Gallet, M. Thomas, Y. Le Berre, J.
Roman of IMDEA Materials, Madrid, Spain for her technical assistance F. Castagné, A. Couret, Temperature control during spark plasma sintering and
with the testing and metallographic aspects of the work. The authors application to up-scaling and complex shaping, J. Mater. Process. Technol. 213
(2013) 269–278.
also acknowledge Mrs. Cristina Moral Gil of Universidad Carlos III for [30] J. Guyon, A. Hazotte, J.P. Monchoux, E. Bouzy, Effect of powder state on spark
her technical assistance with SEM imaging. The authors also acknowl­ plasma sintering of TiAl alloys, Intermetallics. 34 (2013) 94–100.
edge Mr. Shubhashis Dixit and Mr. Barun Dash of the Indian Institute of [31] H. Clemens, S. Mayer, Design, processing, microstructure, Properties, and
applications of advanced intermetallic TiAl alloys, Adv. Eng. Mater. 15 (2013)
Technology Madras for their technical assistance with vacuum encap­ 191–215.
sulation and heat treatment. The authors would like to acknowledge Dr. [32] N. Afrin, D.L. Chen, X. Cao, M. Jahazi, Strain hardening behavior of a friction stir
Manuel Avella (IMDEA Materials Institute) for assistance with TEM welded magnesium alloy, Scr. Mater. 57 (11) (2007) 1004–1007.
[33] P. Ludwik, Influence of the Deformation Rate with Special Consideration of the
characterization.
after-Effects. Elements of Technological Mechanics, Springer, Berlin, Heidelberg,
1909, pp. 44–53.
References [34] B.P. Kashyap, P.D. Hodgson, Y. Estrin, I. Timokhina, M.R. Barnett, I. Sabirov,
Plastic flow properties and microstructural evolution in an ultrafine-grained Al-
Mg-Si alloy at elevated temperatures, Metall. Mater. Trans. A 40 (2009)
[1] S.C. Huang, E.L. Hall, The effects of Cr additions to binary TiAl-base alloys, Metall.
3294–3303, https://doi.org/10.1007/s11661-009-0036-6.
Mater Trans A. 26 (1991) 2619–2627.
[35] I. Sabirov, M.R. Barnett, Y. Estrin, P.D. Hodgson, The effect of strain rate on the
[2] M.R. Kabir, L. Chernova, M. Bartsch, Numerical investigation of room temperature
deformation mechanisms and the strain rate sensitivity of an ultra-fine-grained Al
deformation behavior of a duplex type γ-TiAl alloy using a multi-scale modeling
alloy, Scr. Mater. 61 (2009) 181–184, https://doi.org/10.1016/j.
approach, Acta Mater. 58 (2010) 5834–5847.
scriptamat.2009.03.032.
[3] F. Appel, R. Wagner, Microstructure and deformation of two-phase γ-titanium
[36] W. Kingkam, N. Li, H.X. Zhang, C.Z. Zhao, Hot Deformation Behavior of High
aluminides, Mater. Sci. Eng. R 22 (1998) 187–268.
Strength Low Alloy Steel by Thermo Mechanical Simulator and Finite Element
[4] F. Appel, J.D.H. Paul, M. Oehring, γ Titanium Aluminides, Wiley-VCH Weinheim,
Method, The 2nd International Conference on Materials Engineering and
Germany, 2011.
Nanotechnology, in: IOP Conf. Series: Materials Science and Engineering, 205, IOP
[5] X.H. Chen, L. Lu, Work hardening of ultrafine-grained copper with nanoscale twins,
Publishing, 2017, 012001, https://doi.org/10.1088/1757-899X/205/1/012001.
Scr. Mater. 57 (2007) 133–136.
[37] D. Caillard, J.L. Martin, Thermally Activated Mechanisms in Crystal Plasticity,
[6] Z.J. Pu, K.H. Wu, J. Shi, D. Zou, Development of constitutive relationships for the
Pergamon Materials Series 8, Elsevier, Oxford, United Kingdom, 2003.
hot deformation of boron microalloying TiAl-Cr-V alloys, Mater. Sci. Eng. A 192
[38] J.S. Lian, C.D. Gu, Q. Jiang, Z.H. Jiang, Strain rate sensitivity of face-centered-
(193) (1995) 780–787.
cubic nanocrystalline materials based on dislocation deformation, J. Appl. Phys. 99
[7] B. Liu, Y. Liu, W. Zhang, J.S. Huang, Hot deformation behavior of TiAl alloys
(2006), 076103.
prepared by blended elemental powders, Intermetallics. 19 (2011) 154–159.
[39] F. Diologent, T. Kruml, Measurement of the effective activation volume in 45XD
[8] J. Luo, M.Q. Li, W.X. Yu, H. Li, The variation of strain rate sensitivity exponent and
titanium aluminides by repeated transient tests, Mater. Sci. Eng. A 487 (2008)
strain hardening exponent in isothermal compression of Ti-6Al-4V alloy, Mater.
377–382.
Des. 31 (2010) 741–748.
[40] C.J. Boehlert, C.J. Cowen, S. Tamirisakandala, D.L. McEldowney, D.B. Miracle, D.
[9] R.U. Vaidya, Z. Jin, C. Cady, G.T. Gray, D.P. Butt, A comparative study of the strain
B. Mirac, In situ scanning electron microscopy observations of tensile deformation
rate and temperature dependent compression behavior of Ti-46.5Al-3Nb-2Cr-0.2W
in a boron-modified Ti-6Al-4V alloy, Scr. Mater. 55 (2006) 465–468.
and Ti-25Al-10Nb-3V-1Mo intermetallic alloy, Scr. Mater. 1 (41) (1999) 569–574.
[41] W. Chen, C.J. Boehlert, The 455◦ C tensile and fatigue behavior of boron-modified
[10] H. Clemens, H. Kestler, Processing and applications of intermetallic γ-TiAl-based
Ti-6Al-2Sn-4Zr-2Mo-0.1Si(wt.%), Int. J. Fatigue 32 (5) (2010) 799–807.
alloys, Adv. Eng. Mater. 2 (9)) (2000) 551–570.
[42] Y. Deng, Z. Yin, J. Huang, Hot deformation behavior and microstructural evolution
[11] U. Fröbel, F. Appel, Strain ageing in γ (TiAl)-based and α2 (Ti3Al) titanium
of homogenized 7050 aluminum alloy during compression at elevated
aluminides, Intermetallics 14 (10–11) (2006) 1187–1193.
temperature, J. Mater. Sci. Eng. A. 528 (3) (2011) 1780–1786.
[12] U. Fröbel, F. Appel, Strain ageing in γ (TiAl)-based titanium aluminides due to
[43] J. Li, F. Li, F. Xue, J. Cai, B. Chen, Micromechanical behavior study of forged 7050
antisite atoms, Acta Mater. 50 (14) (2002) 3693–3707.
aluminum alloy by microindentation, Mater. Des. 37 (2012) 491–499.
[13] S. Shu, F. Qiu, B. Xing, S. Jin, J. Wang, Q. Jiang, Effect of strain rate on the
[44] D. Xu, M. Zhu, Z. Tang, C. Sun, Determination of the dynamic recrystallization
compression behavior of TiAl and TiAl-2Mn alloys fabricated by combustion
kinetics model for SCM435 steel, J. Wuhan Univ. Technol.-Mater Sci Ed. 28 (4)
synthesis and hot press consolidation, Intermetallics. 43 (2013) 24–28.
(2013) 819–824.
[14] T. Kawabata, H. Fukai, O. Izumi, Effect of ternary additions on mechanical
[45] S. Cheng, Y.H. Zhao, Y.T. Zhu, E. Ma, Optimizing the strength and ductility of fine
properties of TiAl, Acta Mater. 46 (1998) 2185–2194.
structured 2024 Al alloy by nanoprecipitation, Acta Mater. 55 (2007) 5822–5832.
[15] Y.Q. Su, C. Liu, X.Z. Li, J.J. Guo, B.S. Li, J. Jia, H.Z. Fu, Microstructure selection
[46] Y.M. Wang, E. Ma, Strain hardening, strain rate sensitivity, and ductility of
during the directionally peritectic solidification of Ti-Al binary system,
nanostructured metals, Mater. Sci. Eng. A 375-377 (2004) 46–52.
Intermetallics. 13 (2005) 267–274.

14
C.J. Boehlert et al. Materials Characterization 172 (2021) 110856

[47] E.W. Hart, Theory of the tensile test, Acta Mater. 15 (1967) 351–355. [61] N.V. Dudamell, P. Hidalgo-Manrique, A. Chakkedath, Z. Chen, C.J. Boehlert,
[48] Y. Wei, A.F. Bower, H. Gao, Enhanced strain-rate sensitivity in fcc nanocrystals due F. Galvez, S. Yi, J. Bohlen, D. Letzig, M.T. Pérez-Prado, Influence of strain rate on
to grain-boundary diffusion and sliding, Acta Mater. 56 (2008) 1741–1752. the twin and slip activity of a magnesium alloy containing neodymium, Mater. Sci.
[49] A.V. Sergueeva, N.A. Mara, A.K. Mukherjee, Plasticity at really diminished length Eng. 583 (2013) 220–231.
scales, Mater. Sci. Eng. A 463 (1–2) (2007) 8–13. [62] R.J. Asaro, S. Suresh, Mechanistic models for the activation volume and rate
[50] G. Rajaram, S. Kumaran, S. Suwas, Effect of strain rate on tensile and compression sensitivity in metals with nanocrystalline grains and nano-scale twins, Acta Mater.
behaviour of Al-Si/graphite composite, Mater. Sci. Eng. A 528 (2011) 6271–6278. 53 (2005) 3369–3382.
[51] J. Luo, M.Q. Li, Strain rate sensitivity and strain hardening exponent during the [63] B. Viguier, J. Bonneville, J.L. Martin, The mechanical properties of single-phase γ-
isothermal compression of Ti60 alloy, Mater. Sci. Eng. A 538 (2012) 156–163. Ti47Al51Mn2 polycrystals, Acta Mater. 44 (1996) 4403–4415.
[52] S.Q. Wang, J.H. Liu, D.L. Chen, Effect of strain rate and temperature on strain [64] J.D.H. Paul, F. Appel, Work-hardening and recovery mechanisms in gamma-based
hardening behavior of a dissimilar joint between Ti–6Al–4V and Ti17 alloys, titanium aluminides, Metall. Mater. Trans A. 34 (2003) 2103–2111.
Mater. Des. 56 (2014) (1980-2015) 174–184. [65] A. Tikhonovsky, M. Bartsch, U. Messerschmidt, Plastic deformation of yttria
[53] S. Nagarjuna, M. Srinivas, High temperature tensile behaviour of a Cu–1.5 wt.% Ti stabilized cubic zirconia single crystals I. Activation parameters of deformation,
alloy, Mater. Sci. Eng. A 335 (1–2) (2002) 89–93. Phys. Status Solidi 201 (1) (2004) 26–45.
[54] S. Lou, D.O. Northwood, Effect of strain aging on the strength coefficient and [66] H.J. Frost, M.F. Ashby, Deformation Mechanism Maps, Pergamon Press, Oxford,
strain-hardening exponent of construction-grade steels, J. Mater. Eng. Perform. 3 United Kingdom, 1982.
(3) (1994) 344–349. [67] J. Zhang, H. Di, H. Wang, K. Mao, T. Ma, Y. Cao, Hot deformation behavior of Ti-
[55] Y. Wei, A.F. Bower, H. Gao, Enhanced strain-rate sensitivity in fcc nanocrystals 15-3 titanium alloy: a study using processing maps, activation energy map, and
due to grain-boundary diffusion and sliding, Acta Mater. 56 (8) (2008) Zener–Hollomon parameter map, J. Mater. Sci. 47 (9) (2012) 4000–4011.
1741–1752. [68] T. Yasmeen, B. Zhao, J.-H. Zheng, F. Tian, J. Lin, J. Jiang, The study of flow
[56] D. Canadinc, C. Efstathiou, H. Sehitoglu, On the negative strain rate sensitivity of behavior and governing mechanisms of a titanium alloy during superplastic
Hadfield steel, Scr. Mater. 59 (10) (2008) 1103–1106. forming, Mater. Sci. Eng. A 139482 (2020).
[57] X. Mao, F. Li, J. Cao, J. Li, Z. Sun, G. Zhu, S. Zhou, Strain rate effects on tensile [69] M.L. Meier, D.R. Lesuer, A.K. Mukherjee, α Grain size and β volume fraction aspects
deformation behaviors of Ti-10V-2Fe-3Al alloy undergoing stress-induced of the superplasticity of Ti-6Al-4V, Mater. Sci. Eng. A 136 (1991) 71–78.
martensitic transformation, Mater. Sci. Eng. A 710 (2018) 1–9. [70] H.J. McQueen, D.L. Bourell, in: A.K. Sachdev, J.D. Embury (Eds.), Formability and
[58] F. Appel, U. Lorenz, M. Oehring, U. Sparka, R. Wagner, Thermally activated Metallurgical Structure, TMS, Warrendale, PA, 1987, pp. 341–368.
deformation mechanisms in micro-alloyed two-phase titanium amminide alloys, [71] A. Couret, in: Low and High Temperature Deformation Mechanisms in TiAl Alloys,
Mater. Sci. Eng. A 233 (1997) 1–14. 15th International Conference on the Strength of Materials (ICSMA-15) IOP
[59] J. Friedel, Dislocations, Pergamon Press, Oxford, 1967. Publishing Journal of Physics: Conference Series 240, 2010, 012001, https://doi.
[60] M.A. Morris, T. Lipe, D.G. Morris, Strain-ageing, strain-rate sensitivity, and flow org/10.1088/1742-6596/240/1/012001.
stress variations at intermediate temperatures in a two-phase Ti-Al alloy, Scr. [72] F. Appel, in: Krzysztof Sztwiertnia (Ed.), Phase Transformations and
Mater. 34 (1996) 1337–1343. Recrystallization Processes During Synthesis, Processing and Service of TiAl Alloys,
Recrystallization, InTech, 2012, pp. 225–266. ISBN: 978-953-51-0122-2.

15