Materials Science and Engineering C 49 (2015) 400–407

Contents lists available at ScienceDirect

Materials Science and Engineering C

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

Feasibility study of the production of biomedical Ti–6Al–4V alloy by

powder metallurgy
L. Bolzoni ⁎, E.M. Ruiz-Navas, E. Gordo
Departamento de Ciencia e Ingeniería de Materiales e Ingeniería Química, Universidad Carlos III de Madrid, Avda. de la Universidad, 30, 28911 Leganés, Madrid, Spain

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

Article history: Titanium and its alloys are characterized by an exceptional combination of properties like high strength, good
Received 12 March 2014 corrosion resistance and biocompatibility which makes them suitable materials for biomedical prosthesis and de-
Received in revised form 24 December 2014 vices. The wrought Ti–6Al–4V alloy is generally favored in comparison to other metallic biomaterials due to its
Accepted 8 January 2015
relatively low elastic modulus and it has been long used to obtain products for biomedical applications. In this
Available online 10 January 2015
work an alternative route to fabricate biomedical implants made out of the Ti–6Al–4V alloy is investigated. Spe-
Keywords:
cifically, the feasibility of the conventional powder metallurgy route of cold uniaxial pressing and sintering is ad-
Titanium alloys dressed by considering two types of powders (i.e. blended elemental and prealloyed). The characterization of
Ti–6Al–4V physical properties, chemical analysis, mechanical behavior and microstructural analysis is carried out in-
Powder metallurgy depth and the properties are correlated among them. On the base of the results found, the produced alloys are
Tensile properties promising materials for biomedical applications as well as cheaper surgical devices and tools.
Dynamic Young modulus © 2015 Elsevier B.V. All rights reserved.

1. Introduction machining [15]. Furthermore, using PM techniques equivalent mechan-

ical properties can be obtained and new composition unsuitable for
Titanium, a relatively new lightweight engineering material, is char- ingot metallurgy can be produced. Finally, the requirements for the pro-
acterized by the highest specific strength (strength to density ratio), ex- cessing tools (sintering tray) are less stringent since the material does
cellent corrosion resistance and outstanding biocompatibility [1]. Based not reach the molten state during processing, such as in casting, and
on this, titanium should have a wide range of applications; nonetheless, the formation of the α-case is limited or prevented. The purpose of
this metal has always been used in high performance and high demand- this study is to determine whether the biomedical Ti–6Al–4V alloy can
ing industry, such as aerospace or medicine, where its combination of be produced by means of the conventional PM route of cold uniaxial
properties dominates over affordability issues [2]. This is mainly due to press plus vacuum sinter. Specifically, the variation of some physical,
the production costs of this material, both extraction and processing mechanical and microstructural features along with one of the most im-
costs, compared to competitive materials. The most widely used titani- portant parameter for PM processing, the sintering temperature, is con-
um alloy, the Ti–6Al–4V, is mainly employed to fabricate aerospace sidered. Moreover, the analyses carried out allow discovering whether
components as well as biomedical implants [3]. Titanium and titanium there was some contamination during the processing of the materials
alloys for fixed and removable dental prostheses and devices are com- and differentiate the influence of the powder production process, pre-
monly obtained by casting processes such as centrifugal casting [4,5] al- cisely, the prealloying and the master alloying approaches. The whole
though some alternative techniques have been investigated to produce study was set to assess the feasibility of the production of cheaper struc-
[6–10] or modify their surface [11,12]. When casting is used, machining ture structural biomedical, and in particular dental, prostheses, devices
is generally employed to remove the brittle reaction layer (α-case) [13, and tools.
14] that forms from the interaction between molten titanium and the
processing tools (i.e. oxide-based ceramic molds and coating). Powder 2. Materials and method
metallurgy (PM) techniques could offer the possibility to lower the
final cost even if the starting material, a micrometric powder, is normal- 2.1. Ti–6Al–4V starting powders
ly more expensive than an ingot combined with other important advan-
tages in using PM method to fabricate titanium alloys. In particular, The staring materials were a blended powder, labeled as Ti–6Al–4V–
reduction of the fabrication costs by PM is primarily possible thanks to BE, and a prealloyed powder, identified as Ti–6Al–4V–PA. The descrip-
the more efficient material utilization (yield) and limited or unnecessary tion of the optimization of the purchased master alloy to produce the
Ti–6Al–4V–BE alloy can be found in a previously published work [16].
⁎ Corresponding author. As it can be seen in Fig. 1 which shows SEM pictures of the starting pow-
E-mail address: bolzoni.leandro@gmail.com (L. Bolzoni). ders, both powders are characterized by an irregular morphology which

http://dx.doi.org/10.1016/j.msec.2015.01.043
0928-4931/© 2015 Elsevier B.V. All rights reserved.
 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407 401

a) b)

c) d)
Ti-6Al-4V_(PA)
Al + Al:V (35:65)

Ti
Al:V (35:65)

Fig. 1. SEM pictures showing the morphology (a and b) and the composition (c and d) of the starting powders, respectively: a) Ti–6Al–4V–BE and b) Ti–6Al–4V–PA alloys.

is dictated by their production method (i.e. HDH), a comminution pro- window was set to be laid between 1250 °C and 1350 °C whilst the
cess. The powders have a similar particle size distribution (b90 μm) dwell time was fixed at 2 h. It is worth mentioning that a minimum of
but the Ti–6Al–4V–PA powder is somewhat finer than the Ti–6Al–4V– three samples were considered for each processing condition.
BE powder. From the BSE micrograph of the Ti–6Al–4V–BE powder
(Fig. 1c), the starting powders used to fabricate this alloy can be identi- 2.3. Characterization of the sintered Ti–6Al–4V materials
fied: elemental titanium, an Al:V master alloy and elemental aluminum.
The particle size of the Al:V master alloy was reduced by high energy The variation of the dimensions induced by the sintering step was
ball milling meanwhile mixed with the elemental aluminum powder considered and for that the dimensions of green and sintered specimens
[16]. Due to the intrinsic nature of the materials (i.e. ductile aluminum were measured. The relative density values (ρr) were calculated as:
and brittle Al:V), the highly deformed aluminum powder particles (ρs / ρnom ∗ 100) where ρs is the density of the sintered samples and
tend to embed some of the Al:V particles. From Fig. 1d, the Ti–6Al– the (ρnom) is the nominal density of the Ti–6Al–4V alloy (4.43 g/cm3)
4V–BE powder is characterized by a uniform distribution of the alloying [17]. As for the powders, the calibrated LECO TC500 was used to mea-
elements. The contents of oxygen and nitrogen (ASTM: E1409) deter- sure the contents of interstitial elements of the sintered materials. The
mined in the Ti–6Al–4V–BE and Ti–6Al–4V–PA powders are O = preparation of the samples for their microstructural analysis included:
0.43 wt.% and N = 0.012 wt.% and O = 0.42 wt.% and N = 0.010 wt.%, cut, grinding with SiC papers of different granulometries, polishing
respectively. The total amount of oxygen dissolved in the starting pow- with silica gel and etching with Kroll' reactant. Vickers hardness
ders is already higher than the value of the wrought Ti–6Al–4V (i.e. (HV30) measurements were performed in a Wilson Wolpert DIGI-
0.20 wt.%) whilst the content of nitrogen is lower than the limit speci- TESTOR 930 Universal Hardness Tester. Tensile tests, done on the base
fied for the ELI (extra-low interstitials) grade of the Ti–6Al–4V alloy of the ASTM: E8 standard, were performed on a MicroTest universal ma-
whose maximum is 0.03 wt.% [17]. chine. The crosshead speed was set to 1 mm/min. Yield strength values
were calculated by means of the off-set method. The fractographic study
2.2. Consolidation of the Ti–6Al–4V materials of the tensile samples was done using a Philips XL-30 SEM. For tensile
specimens, the dynamic elastic modulus (E) determined by the speed
The loose powders were poured inside the cavity of a floating die of sound (ν) and the relative density was considered to avoid possible
whose die-walls were lubricated with zinc stearate and consolidated artifacts intrinsic of the stress–strain curves and it calculated using
with a uniaxial press. It is important to remark that no lubricant was Eq. (1):
added on purpose to the Ti–6Al–4V powders to limit as much as possi- pffiffiffiffiffiffiffiffiffiffi
ble the contamination of the starting powders. The geometry of the v¼ E=ρr : ð1Þ
specimens chosen for this study is the one known as “dogbone”
(ASTM: B925) to be tested to measure tensile properties. For the An ultrasonic transducer (Grindosonic) having a frequency range in
sintering of the green samples an electrical resistance tubular furnace between 20 Hz and 100 kHz and an accuracy better than 0.005% was
was used and sintering was carried out under high vacuum (10− 5 employed for determining the speed of sound. Throughout the whole
mbar). Based on previous studies [18,19], the sintering temperature work, the properties (i.e. microstructure, hardness, yield strength,
402 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407

ultimate tensile strength, strain and Young modulus) of the sintered Ti– Table 1
6Al–4V materials are compared to those of the wrought Ti–6Al–4V in Oxygen and nitrogen contents of the sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys.

the typical mill annealed state (750 °C for 4 h, furnace cooled) whose Material Processing conditions Interstitials [wt.%]
microstructure is composed of globular particles of β in a matrix of α Oxygen Nitrogen
[17].
Ti–6Al–4V–BE 1250 °C–2 h 0.44 ± 0.01 0.025 ± 0.006
1300 °C–2 h 0.46 ± 0.05 0.024 ± 0.001
3. Results 1350 °C–2 h 0.46 ± 0.02 0.023 ± 0.002
Ti–6Al–4V–PA 1250 °C–2 h 0.47 ± 0.01 0.023 ± 0.003
1300 °C–2 h 0.51 ± 0.01 0.019 ± 0.006
3.1. Physical properties of the sintered Ti–6Al–4V materials 1350 °C–2 h 0.56 ± 0.03 0.022 ± 0.002

The green density of the cold uniaxially pressed tensile test samples
with “dogbone” geometry is 86.56% ± 0.38% for Ti–6Al–4V–BE and
From the data of the chemical analysis carried out on the sintered Ti–
81.98% ± 0.49% for Ti–6Al–4V–PA, respectively. The variations of the di-
6Al–4V–BE and Ti–6Al–4V–PA alloys to determine the amount of inter-
mensions of the samples and of the relative density are presented in
stitial elements dissolved (Table 1), it can be seen that the oxygen con-
Fig. 2.
tent increases with the increment of the processing temperature with
Analyzing the data of the shrinkage underwent from the green sam-
the only exception of the Ti–6Al–4V–BE samples sintered at 1350 °C
ples during the sintering step (Fig. 2a), it can be seen that the shrinkage
which are characterized by the same content of oxygen of that of the
increases linearly with the increment of the sintering temperature from
specimens processed at 1300 °C. Conversely, in the case of nitrogen con-
1250 °C to 1350 °C where this increment is much more noticeable for
tent there is not a clear trend with the increment of the processing tem-
the Ti–6Al–4V–BE alloy than for the Ti–6Al–4V–PA alloy. More in detail,
perature for either of the alloys considered. From the data of Table 1, it
in the case of the Ti–6Al–4V–BE alloy the increment is somewhat less
can also be noticed that, although of the employment of high vacuum,
pronounced when raising the temperature from 1250 °C to 1300 °C
there is both oxygen and nitrogen pick-up with respect to the amount
(i.e. 0.22%) than from 1300 °C to 1350 °C (i.e. 0.27%) whilst the incre-
of these interstitials present in the starting powder regardless of the
ment of the shrinkage with the processing temperature is constant
processing conditions used to sinter the green samples. On the one
(i.e. 0.04%) for the Ti–6Al–4V–PA alloy. For the Ti–6Al–4V–BE alloy
hand, the oxygen pick-up for the Ti–6Al–4V–BE alloy is quite limited
part of thermal energy supplied to the system is spent for the diffusion
(i.e. maximum 0.03 wt.%) whilst it is greater, between 0.05 wt.% and
of the alloying elements towards the titanium matrix, which is not the
0.14 wt.%, for the Ti–6Al–4V–PA alloy. On the other hand, the amount
case for the Ti–6Al–4V–PA alloy where the total composition is already
of nitrogen pick-up seems to be rather constant, because the final values
fully homogeneous. This aspect is responsible for the lower absolute
lay between 0.019 wt.% and 0.025 wt.%, and, thus, better limited by the
shrinkage values that characterized the Ti–6Al–4V–BE alloy (about
sintering under high vacuum. The total amount of interstitial elements,
4%) with respect to the Ti–6Al–4V–PA alloy (greater than 6%). As it
but in this particular case especially oxygen because it is higher than
can be seen in Fig. 2b, generally, the relative density of the Ti–6Al–4V–
that of the wrought alloy, will influence the mechanical behavior of
BE and Ti–6Al–4V–PA alloys slightly increases with the sintering tem-
the sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys. This is because
perature. Specifically, the increment that the Ti–6Al–4V–BE samples ex-
the interstitials harden, strengthen and diminish the ductility of titani-
perience during sintering is greater when increasing the processing
um and its alloys [20–22].
temperature from 1250 °C to 1300 °C rather than from 1300 °C to
1350 °C where, actually, this last increment is less than half of the pre-
3.3. Microstructural analysis of the sintered Ti–6Al–4V materials
vious one. Conversely, in the case of the Ti–6Al–4V–PA alloy the incre-
ment in terms of relative density is quite homogeneous with the
A microstructural analysis was carried out to study the development
processing temperature. Because of these two relative trends the differ-
of the microstructure and the evolution of the porosity and the repre-
ence between the final relative density values of the Ti–6Al–4V–BE and
sentative micrographs of both materials are displayed in Fig. 3.
Ti–6Al–4V–PA alloys becomes smaller but still the final relative density
The microstructure of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys
of the Ti–6Al–4V–PA alloy is, at least, 1% higher than that of the Ti–6Al–
sintered at 1250 °C (Fig. 3a and b) is composed of α grains and α + β
4V–BE alloy, which starts from a higher green density but undergoes
lamellae of different orientation that remained after the growth of the
lower shrinkage.
α phase during cooling. Another feature of the microstructure of the
sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys is the residual porosity
3.2. Chemical analysis of the sintered Ti–6Al–4V materials that can be clearly identified in their microstructures. On the one side, a
great percentage of the residual porosity of the Ti–6Al–4V–BE alloy is
The results of the chemical analysis done on sintered samples are still interconnected and irregular in shape indicating that the pore coa-
presented in Table 1. lescence and coarsening did not take place completely yet even if

Fig. 2. Variation of the thickness (a) and of the relative density (b) as a function of the sintering temperature for the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys.
 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407 403

Fig. 3. Optical micrographs of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys sintered at 1250 °C: a) Ti–6Al–4V–BE and b) Ti–6Al–4V–PA, at 1300 °C: c) Ti–6Al–4V–BE and d) Ti–6Al–4V–PA
and at 1350 °C: e) Ti–6Al–4V–BE and f) Ti–6Al–4V–PA.

spherical pores are present. On the other hand, the residual porosity of employed the complete diffusion and homogenization of the alloying
the Ti–6Al–4V–PA alloy is practically spherical with very few coarse elements are obtained.
pores of irregular shape. An increment of the sintering temperature to The total amount of residual porosity, which can be clearly seen in
1300 °C or 1350 °C does not change the nature of the microconstituents the BSE images reported in Fig. 4, decreases with the increment of the
(i.e. α grains and α + β lamellae) but it does induce their coarsening, sintering temperature. The great majority of the pores are spherical in
which is especially notable in the components sintered at 1350 °C. shape and located at the boundaries between adjacent α grains divided
Therefore, the microstructure of the Ti–6Al–4V–BE alloy is still coarser by the lamellae. The size distribution of the pore ranges between 20 μm
than that of the Ti–6Al–4V–PA alloy though the difference is not that and 50 μm with some few pores of bigger size (b 100 μm).
marked as in the case of the samples processed at 1250 °C. Concerning The results of the EDS analysis carried out to further confirm the ho-
the residual porosity, the increment of the sintering temperature leads mogeneous distribution of the alloying elements, which is paramount to
to its volumetric percentage reduction but the mean size seems to be in- guarantee reliable and consistent properties, are reported in Table 2.
creased due to the fact that the pores tend to coalesce. In the case of the
Ti–6Al–4V–BE alloy the residual porosity of the specimens processed at
1350 °C is mainly spherical, even if some irregular pores are still present, 3.4. Mechanical properties of the sintered Ti–6Al–4V materials
which it is not the case for the material sintered at 1300 °C.
Fig. 4 shows the BSE micrographs of the Ti–6Al–4V–BE and Ti–6Al– The results of hardness measurements for Ti–6Al–4V–BE and Ti–
4V–PA alloys where it can be seen that with the sintering parameters 6Al–4V–PA alloys are shown in Fig. 5.
404 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407

Fig. 4. SEM (BSE) micrographs of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys sintered at 1250 °C: a) Ti–6Al–4V–BE and b) Ti–6Al–4V–PA, at 1300 °C: c) Ti–6Al–4V–BE and d) Ti–6Al–4V–
PA and at 1350 °C: e) Ti–6Al–4V–BE and f) Ti–6Al–4V–PA.

From the data of the Vickers Hardness measurements performed The typical tensile stress–strain behavior found for Ti–6Al–4V–BE
into the cross-section of the sintered specimens (Fig. 5), the hardness and Ti–6Al–4V–PA materials sintered in the range 1250–1350 °C is
follows the same trend of the relative density and, therefore, it increases displayed in Fig. 6.
continuously with the increment of the sintering temperature. By a clos-
er look to the absolute values, it can be noticed that the increment in
terms of hardness that takes place between 1250 °C and 1300 °C is
slightly greater than that between 1300 °C and 1350 °C. This fact is in
agreement with the densification and the relative density (Fig. 2b)
data and it is mainly a consequence of the phenomena that govern the
sintering step.

Table 2
EDS chemical composition of the sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys.

Material Processing conditions Ti [wt.%] Al [wt.%] V [wt.%]

Ti–6Al–4V–BE 1250 °C–2 h 90.14 5.97 3.89

1300 °C–2 h 90.30 5.89 3.81
1350 °C–2 h 89.64 6.21 4.15
Ti–6Al–4V–PA 1250 °C–2 h 90.05 5.81 4.14
1300 °C–2 h 89.91 6.25 3.84
Fig. 5. Variation of the hardness as a function of the sintering temperature for the Ti–6Al–
1350 °C–2 h 90.13 6.09 3.78
4V–BE and Ti–6Al–4V–PA alloys.
 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407 405

Fig. 6. Representative stress–strain behavior of sintered Ti–6Al–4V alloys: a) Ti–6Al–4V–BE and b) Ti–6Al–4V–PA.

As it can be seen in Fig. 6, the typical tensile stress–strain curves 4. Discussion

for the Ti–6Al–4V processed by pressing and sintering are composed
of an elastic part, up to approximately 800 MPa, and then some plas- From the characterization of the starting powders and of the behav-
tic deformation, which is around 3%. This plastic deformation gets ior of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys, it can be stated that
lower with the sintering temperature. From the curves it can also the irregular morphology of the Ti–6Al–4V powders makes them ideal
be noticed that, independently of the sintering temperature, all the and suitable materials to be processed by the conventional route of
materials show similar shape indicating similar behavior and cold uniaxial pressing and sintering. This is because during cold pressing
Young modulus. the asperities of the irregular powder particles deform and form the in-
The variation of the yield stress, ultimate tensile strength (UTS) and terparticle locking which constitutes the green strength of the green
strain at fracture (ε) with the sintering temperature in comparison to components. This guarantees the handling of the parts without break-
the value of the wrought alloy is shown in Fig. 7. age or delamination which is paramount for maintaining the original
Analyzing the data shown in Fig. 7, it can be seen that the yield shape of biomedical devices. Moreover, the difference in maximum par-
stress slightly increases with the processing temperature and, in ticle size plays a role during the consolidation and the densification of
most of the cases, the values obtained are higher than the one spec- the Ti–6Al–4V components because the smaller the particle size the
ified for the wrought Ti–6Al–4V alloy. With respect to the UTS of the greater the specific surface area where the reduction of the total surface
Ti–6Al–4V–BE alloy decreases from 900 MPa down to 810 MPa with area is one of the driving force of the sintering of particulate materials.
the increasing of the sintering temperature whereas that of the More in detail, Ti–6Al–4V–PA alloy has slightly lower green density
Ti–6Al–4V–PA alloy stays practically constant. Concerning the but this is due to the fact that it already has the alloying elements dis-
strain at fracture data (Fig. 7), it can be seen that the ductility of solved inside the titanium matrix in each single powder particle. This
both Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys decreases with the aspect makes them harder and less deformable and, therefore, more dif-
temperature. ficult to shape by cold uniaxial pressing. Nevertheless, the Ti–6Al–4V–
The fractographic analysis of the tensile test specimens by SEM was PA alloy is characterized by a higher contraction during sintering (i.e.
carried out and the results are presented in Fig. 8. shrinkage) and undergoes a much higher densification (somewhat
From the analysis of the fracture surfaces (Fig. 8), it can be seen that lower than 80%) with respect to the Ti–6Al–4V–BE alloy (50–60%).
the materials present ductile fracture due to microvoid coalescence This is due to the fact that none of the thermodynamic energy supplied
where the size of the dimples found in the Ti–6Al–4V–BE alloy seems to the system is invested in the diffusion of the alloying elements to-
to be larger than Ti–6Al–4V–PA. This pore-assisted fracture is the typical wards the titanium matrix and on the homogenization of the alloying
behavior of the Ti–6Al–4V titanium alloy obtained by the conventional elements throughout the whole microstructure. Because of these as-
PM route and it is similar to those formed on ingot metallurgy processed pects, the Ti–6Al–4V–BE alloy reaches final lower relative density
materials [23]. values, of at least 1% (Fig. 2b), where the difference in terms of relative
The results of the dynamic Young modulus carried out on Ti–6Al– density is more marked at lower sintering temperatures. Furthermore,
4V–BE and Ti–6Al–4V–PA alloys sintered samples are reported in the densification of the Ti–6Al–4V–PA alloy is also favored by the slight-
Table 3 where it can be seen that the measured values are very similar ly lower powder particle size or, in turns, its bigger specific surface area.
between them regardless of the sintering temperature which is in The sintering of the green Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys com-
agreement with the results of the stress–strain curves (Fig. 6). ponents has two other main effects: interstitial pick-up and grain

Fig. 7. Variation of the tensile properties as a function of the sintering temperature for the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys: left) ultimate tensile strength — UTS and right) strain at
fracture (ε).
406 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407

Fig. 8. Fracture surface of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys sintered at 1250 °C: a) Ti–6Al–4V–BE and b) Ti–6Al–4V–PA.

growth. Concerning the chemical analysis, oxygen and nitrogen pick-up comparing the absolute hardness values of the two alloys studied, it
are due to the handling of the powder, the air trapped into the green can be seen that Ti–6Al–4V–PA samples always reach higher hardness
samples and to the oxygen and nitrogen atoms adsorbed on the surface of approximately 20 HV30 with respect to the Ti–6Al–4V–BE specimens.
of the powder particles which diffuse into the material during the This is, once again, due to the combination of higher relative density (i.e.
sintering step. Nevertheless, the final amount of interstitials is also af- lower residual porosity) and higher oxygen content (Table 1) that char-
fected by the fact that sintering is carried out by batched and there acterizes the Ti–6Al–4V–PA sintered components. From the graph
could be a slight variation in the atmosphere present inside the shown in Fig. 5, it can also be seen that the sintered Ti–6Al–4V–BE
sintering furnace. In comparison to the typical values of the wrought and Ti–6Al–4V–PA alloys, generally, have comparable hardness to the
Ti–6Al–4V alloy, the amount of oxygen dissolved by the Ti–6Al–4V–BE wrought Ti–6Al–4V alloy with the exception of the samples made out
and Ti–6Al–4V–PA alloys is higher, as it was already the case for the of the Ti–6Al–4V–BE alloy sintered at 1250 °C. The fact that the sintered
starting powders, whereas the final nitrogen contents shown in Ti–6Al–4V–BE and Ti–6Al–4V–PA reach similar hardness even though
Table 1 are still lower than the limit of the ELI grade. Regarding the mi- of the presence of approximately 4–5% of residual porosity is due to
crostructural analysis, the microstructure of the Ti–6Al–4V–BE and Ti– the greater amount of oxygen dissolved into these materials. Actually,
6Al–4V–PA alloys is typical of the wrought Ti–6Al–4V alloy slow cooled Ti–6Al–4V–PA alloy, which undergoes a higher oxygen pick-up during
from a processing temperature above the β transus (i.e. 996 °C), al- sintering, is characterized by an even higher Vickers hardness in com-
though finer with respect to wrought alloys, and it is composed by α parison to the wrought Ti–6Al–4V alloy in the annealed state whose
grains which started to grow from the beta grain and α + β lamellae. hardness is 320 HV [17]. The yield stress of both Ti–6Al–4V–BE and
Moreover, by the comparison of the microstructure of the Ti–6Al–4V– Ti–6Al–4V–PA alloys increases because of the reduction of the residual
BE and Ti–6Al–4V–PA alloys, it can be noticed that the microstructural porosity and the strengthening mechanism of the oxygen present in
features of the Ti–6Al–4V–BE alloy are slightly coarser than those of the interstitial sites of the titanium lattice. The specific level of oxygen
the Ti–6Al–4V–PA due to the fact that the Ti–6Al–4V–BE has slightly of each sintering condition also helps in explaining the differences be-
bigger starting particle size distribution. Another microstructural fea- tween the two materials (i.e. Ti–6Al–4V–PA alloy has higher oxygen
ture of the sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys is the pres- and, thus, higher yield strength) and the difference with respect to the
ence of the residual porosity where this is smaller in size and lower in value of the Ti–6Al–4V wrought alloy (Fig. 7). Specifically, Ti–6Al–4V–
volumetric percentage for the Ti–6Al–4V–PA alloy in agreement with BE and Ti–6Al–4V–PA alloys have higher yield strength than the
the relative density data shown in Fig. 2b. The increment of the sintering wrought alloy, although of the presence of the residual porosity, be-
temperature leads to a coarsening of the microstructural features (i.e. cause of their higher oxygen content. The ultimate tensile strength of
grain growth and pores coalescence). On the base of the characteriza- the Ti–6Al–4V–BE alloy decreases with the sintering temperature and
tion and of the result just described, the Ti–6Al–4V–BE and Ti–6Al– that of the Ti–6Al–4V–PA alloy remains constant and this different be-
4V–PA alloys have similar or higher hardness and comparable mechan- havior is due to the combination of the lower relative density, lower ox-
ical behavior (i.e. comparable ultimate tensile strength, the same ygen content and coarser microstructure of the Ti–6Al–4V–BE and Ti–
microvoids coalescence ductile fracture mode and similar Young modu- 6Al–4V–PA alloys. Actually, it can be noticed that the difference be-
lus) but lower ductility with respect to the wrought Ti–6Al–4V alloy. In comes bigger with the increment of the sintering temperature whose
particular, the hardness of the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys effects are: reduce the difference in terms of relative density and size
increases with the processing temperature where this increment is of the microstructural features but significantly enhance the difference
mainly due to the reduction of the volumetric amount of residual poros- in terms of oxygen content between the Ti–6Al–4V–BE and Ti–6Al–
ity present in the microstructure but is also favored by the oxygen and 4V–PA alloys. By analyzing the UTS data shown in Fig. 7, it can also be
nitrogen pick-up that the component experiences during sintering. By noticed that, apart from the Ti–6Al–4V–BE alloy sintered at 1350 °C,
the UTS values obtained are similar to that of the wrought alloy (i.e.
Table 3 900 MPa) despite the fact that the Ti–6Al–4V–BE and Ti–6Al–4V–PA al-
Dynamic Young modulus results of the sintered Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys. loys are characterized by the presence of the residual porosity as
microconstituent (Fig. 3). This behavior is justified by the fact that the
Material Processing conditions Young modulus [GPa]
Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys have higher oxygen content
Ti–6Al–4V–BE 1250 °C–2 h 112 ± 3 Mean value: 111 ± 6
and finer microstructural features in comparison to the wrought Ti–
1300 °C–2 h 104 ± 1
1350 °C–2 h 116 ± 1 6Al–4V alloy in the annealed state. For both the Ti–6Al–4V–BE and Ti–
Ti–6Al–4V–PA 1250 °C–2 h 109 ± 1 Mean value: 110 ± 1 6Al–4V–PA alloys the strain at fracture decreases with the increment
1300 °C–2 h 111 ± 2 of the sintering temperature principally due to the interstitial elements
1350 °C–2 h 110 ± 1 pick-up. It is worth mentioning that the greatest drop in strain for the
Ti–6Al–4V Wrought 114
Ti–6Al–4V–BE alloy takes place from 1250 °C to 1300 °C which is exactly
 L. Bolzoni et al. / Materials Science and Engineering C 49 (2015) 400–407 407

when the amount of interstitial elements increases the most. In the case cold uniaxial pressing and (vacuum) sintering show similar mechanical
of the Ti–6Al–4V–PA alloy, the important drop in ductility occurs when behavior to that of the wrought alloy. The main advantages of the em-
increasing the sintering temperature from 1300 °C to 1350 °C which is, ployment of powder metallurgy for the production of biomedical de-
once again, when the material experiences the greatest interstitials vices are the reduction of the final production cost and the possibility
pick-up and pores coarsening. By comparing the strain at fracture data to consolidate components to their near-net shape. Irregular powders
of the two alloys, it can be seen that the Ti–6Al–4V–PA alloy performs are suitable for the pressing and sintering route where the increment
somewhat better than the Ti–6Al–4V–BE alloy despite the grater oxy- of the sintering temperature leads to higher shrinkage and densification
gen content which is mainly due to the lower relative density or volu- as well as relative density level regardless of the nature of the powder. A
metric percentage of residual porosity of the Ti–6Al–4V–BE alloy great range and combination of mechanical properties can be achieved
where the size and shape of the pores play a paramount role. Whit re- by changing the processing temperature to adjust them for specific
spect to the wrought alloy (Fig. 7), the ductility of both Ti–6Al–4V–BE products. Moreover, the employment of different types of powders
and Ti–6Al–4V–PA alloys is lower due to the greater amount of intersti- (namely, prealloyed or blending elemental) is another aspect which
tial elements and the residual porosity that characterize them. If need- permits to wide the spectrum of properties achievable and required
ed, the final amount of residual porosity of the Ti–6Al–4V–BE and Ti– for different applications.
6Al–4V–PA alloys could be reduced or even eliminated by means of a
post-processing of the materials by using a hot isostatic pressing cycle.
Acknowledgments
For that the samples would need to be encapsulated in an inert contain-
er and degasified prior to proceeding with their processing at a temper-
The authors want to acknowledge the financial support from the
ature below the beta transus of the alloy simultaneously applying
Regional Government of Madrid through the ESTRUMAT (S2009/MAT-
isostatic pressure by means of an inert gas (i.e. argon). The reduction
1585) project and from the Spanish Ministry of Science through the
or sealing of the residual porosity is expected to improve the mechanical
R&D Projects MAT2009-14547-C02-02 and MAT2009-14448-C02-02.
performances, especially under dynamic and cyclic loads (i.e. fatigue).
Finally, both Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys have similar elas-
tic modulus to the wrought material and it does not vary significantly References
with the sintering temperature. This is thought to be due to the fact
[1] C. Leyens, M. Peters, Titanium and Titanium Alloys: Fundamentals and Applications,
that the sintered alloys are characterized exclusively by closed porosity Wiley-VCH, Köln, Germany, 2003.
because their relative density is greater than 94%. Below this critical [2] K. Roncone, O. Yu, C. Boland, R. Boyer, M. Mohanty, N. Murray, A. Sherman, JOM 57
value the sintered materials present both open and closed porosity of (2005) 11–13.
[3] D.M. Brunette, P. Tengvall, M. Textor, P. Thomsen, Titanium in Medicine: Materials
relative density [15,24] and, consequently, physical and mechanical Science, Surface Science, Engineering, Biological Responses and Medical Applica-
properties are much more affected by the presence of the residual po- tions, Springer, New York, USA, 2001.
rosity. Specifically, the properties are expected to have steeper decreas- [4] K. Watanabe, O. Miyakawa, Y. Takada, O. Okuno, T. Okabe, Biomaterials 24 (2003)
1737–1743.
ing linear correlation with the decrement of the relative density for [5] P.C. Garcia Oliveira, G. Luis Adabo, R. Faria Ribeiro, S. Soares Rocha, Dent. Mater. 22
values lower than 94% [25]. The variation of the mechanical properties (2006) 1098–1102.
reported in this work with the sintering temperature are consistent [6] T. Traini, C. Mangano, R.L. Sammons, F. Mangano, A. Macchi, A. Piattelli, Dent. Mater.
24 (2008) 1525–1533.
with the results found when assessing the sinterability of the Ti–6Al–
[7] V. Srimaneepong, T. Yoneyama, E. Kobayashi, H. Doi, T. Hanawa, Dent. Mater. 24
4V–BE and Ti–6Al–4V–PA alloys over a wider range of processing tem- (2008) 839–845.
peratures (i.e. 900–1400 °C) [18,19] or the influence of the sintering [8] E. Marin, M. Pressacco, S. Fusi, A. Lanzutti, S. Turchet, L. Fedrizzi, Mater. Sci. Eng. C 33
(2013) 2648–2656.
time [26]. It is worth mentioning in previous studies that, in some
[9] Z. Lin, L. Wang, X. Xue, W. Lu, J. Qin, D. Zhang, Mater. Sci. Eng. C 33 (2013)
cases, a lower compacting pressure was used. The lower shaping pres- 4551–4561.
sure leads to lower green densities and, eventually a stronger effect of [10] G. Serra, L. Morais, C.N. Elias, I.P. Semenova, R. Valiev, G. Salimgareeva, M. Pithon, R.
the residual porosity in the final behavior of the sintered alloys. In par- Lacerda, Mater. Sci. Eng. C 33 (2013) 4197–4202.
[11] F.J. Gil, J.A. Planell, J. Padros, C. Aparicio, Dent. Mater. 23 (2007) 486–491.
ticular, in that part of this systematic study to assess different properties [12] A. Poulon-Quintin, I. Watanabe, E. Watanabe, C. Bertrand, Dent. Mater. 28 (2012)
of titanium powder metallurgy materials processed under different 945–951.
combinations of compacting and sintering parameters it was found [13] S.-Y. Sung, Y.-J. Kim, Mater. Sci. Eng. A 405 (2005) 173–177.
[14] C. Ohkubo, T. Hosoi, J.P. Ford, I. Watanabe, Dent. Mater. 22 (2006) 268–274.
that the mechanical properties increase linearly with the increment of [15] R.M. German, Powder Metallurgy Science, 2nd Edition MPIF — Metal Powder Indus-
the sintering temperature. Nonetheless, the mechanical performances tries Federation, Princeton, USA, 1994.
(transverse rupture strength and strain) decrease starting approximate- [16] L. Bolzoni, P.G. Esteban, E.M. Ruiz-Navas, E. Gordo, Powder Metall. 54 (2011)
543–550.
ly from the sintering temperature of 1300 °C. This is due to the fact that [17] R. Boyer, G. Welsch, E.W. Collings, in: A. International (Ed.), Ohio, USA, 1998.
for the Ti–6Al–4V–BE and Ti–6Al–4V–PA alloys a critical amount of in- [18] L. Bolzoni, P.G. Esteban, E.M. Ruiz-Navas, E. Gordo, J. Mech. Behav. Biomed. Mater. 15
terstitials (especially oxygen) is reached with such processing condi- (2012) 33–45.
[19] L. Bolzoni, P.G. Esteban, E.M. Ruiz-Navas, E. Gordo, J. Mech. Behav. Biomed. Mater. 14
tion. The tensile properties of the Ti–6Al–4V–BE and Ti–6Al–4V–PA (2012) 29–38.
alloys show similar behavior as discussed on the base of the results re- [20] R.I. Jaffee, I.E. Campbell, Trans. Am. Inst. Min. Metall. Eng. 185 (1949) 646–654.
ported in Fig. 7. [21] R.I. Jaffee, H.R. Ogden, D.J. Maykuth, Trans. Am. Inst. Min. Metall. Eng. 188 (1950)
1261–1266.
[22] W.L. Finlay, J.A. Snyder, J. Met. 188 (1950) 277–286.
5. Conclusions [23] J.E. Smugeresky, D.B. Dawson, Powder Technol. 30 (1981) 87–94.
[24] G. Friedman, R. Regn, Met. Powders Rep. 39 (1984) 273–281.
This feasibility study of the production of the biomedical Ti–6Al–4V [25] C. Vincent, J.F. Silvain, J.M. Heintz, N. Chandra, J. Phys. Chem. Solids 73 (2012)
499–504.
alloy by powder metallurgy demonstrated that, actually, titanium alloys [26] L. Bolzoni, E.M. Ruiz-Navas, E. Gordo, Mater. Des. 52 (2013) 888–895.
processed by means of the conventional powder metallurgy route of