Met. Mater. Int., Vol. 17, No. 1 (2011), pp. 00~00
doi: 10.1007/s12540-010- Published 26 February 2011
Study of Compaction and Ejection of Hydrided-Dehydrided
Titanium Powder
P. G. Esteban
1
, Y. Thomas
2
, E. Baril
2
, E. M. Ruiz-Navas
1
, and E. Gordo
1,
*
1
Department of Materials Science and Engineering, IQMAAB, University Carlos III ofe Madrid,
Avda. de la Universidad 30, 28911 Leganés (Madrid), Spain
2
Industrial Materials Institute (IMI), National Research Council Canada (CNRC–NRC),
75 de Mortagne Boulevard, Boucherville (Quebec), J4B 6Y4, Canada
(received date: 2010 / accepted date: 2010)
Three similar varieties of pure Ti hydride-dehydried (HDH) powders were tested for the understanding of
the variables that have an influence on the compaction process of Ti powders. The study shows that small
differences in the characteristics of the powders lead to very different behaviours in the compaction stage.
Compressibility curves, friction with the die walls and ejection forces are discussed in this study. The results
are compared with a commercial iron powder as a reference to complete the discussion, as well as to show
the enhancements and modifications that should be performed in Ti powders to design an optimized powder
suitable for being pressed in an industrial process.
Keywords: titanium hydride-dehydried(HDH), uniaxial pressing, compressibility
1. INTRODUCTION
The excellent combination of specific strength and corro-
sion resistance of titanium and titanium alloys [1,2] encour-
ages the development of low cost processes to obtain Ti
parts. Moreover, some studies [3] predict the decrease of the
prices of Ti due to new production techniques for obtaining
Ti from its ores. Among these new techniques, the Arm-
strong process [4] is ready for production, and other incom-
ing developments based on electrolytic methods could be
even more efficient in obtaining Ti at low cost. Among the
electrolytic methods, the FFC development [5] seems to be
the most promising to obtain Ti powder by the direct reduc-
tion of TiO
2
. Other works [6,7] have studied the benefits that
Ti would provide in the reduction of emissions in vehicles.
This conjunction of factors encourages industry to develop
the powder metallurgy technology for Ti.
Powder metallurgy has provided a low-cost route for man-
ufacturing iron-based parts and the related processes are now
optimized since they have been studied for a long time. On
the other hand, Ti is emerging as a good candidate for being
processed by conventional powder metallurgy techniques,
but improvements should be done to reach the optimization
of the processes. Ti powder industry has been restricted to
high-added value applications where the cost of the materi-
als and processes are not fully optimized for large scale
productions. The traditional and potential applications
include military, sports, aerospace, medical and automo-
tive [8,9]. Biomedical applications are promising to con-
tinue the development of Ti technology, including PM
processing [10-13].
Few papers have studied the compaction of Ti powders
[14-18]. This work is focused on the different aspects related
to the uniaxial pressing of Ti powder in comparison with the
well-known behaviour of commercial iron powder. As it will
be shown in this study, the significant difference between
titanium and iron is reflected in their relative pressing perfor-
mances.
Regarding the type of Ti powder used for this study, the
Hydried-dehydried powder (Ti HDH) has been selected due
to the higher purity than sponge powder, and irregular shape
compared to atomised powder, providing better characteris-
tics for powder metallurgy processing. Ti HDH is a common
variety of Ti powder which is produced by the comminution
of Ti solids which have been previously embrittled. In a first
stage, Ti solids are heated into a furnace in a hydrogen atmo-
sphere. Then, the solids loose their natural ductility by
hydrogen embrittlement, leading to hydrided Ti. In a second
stage, hydrided Ti solids are easily comminuted to the desire
powder size. In a final stage, hydrided Ti powder must be
dehydried to recover its original ductility and properties.
This is achieved by heating Ti powder into a vacuum fur-
nace, where degassing of the material occurs.
*Corresponding author: egordo@ing.uc3m.es
©KIM and Springer
2 P. G. Esteban et al.
2. EXPERIMENTAL PROCEDURE
All experiments were carried out using hydride-dehydride
(HDH) titanium powders and water-atomized iron powder,
compacted with admixed lubricant or with die-wall lubri-
cant. Table 1 describes the type of powders tested, their supplier,
and the lubrication mode used in each case. In particular, two
batches of the similar powder (Ti, HDH process, < 75 µm)
sold under the same tradename were evaluated (Batch 1 and
Batch 2).
Apparent density was measured using the standards MPIF
4 and MPIF 28 [19,20]. Particle size distributions of the dif-
ferent powders were measured with a laser diffraction parti-
cle size analyzer (Beckman Coulter LS 13 320, USA).
Chemical analyses were carried out on LECO analyzers,
LECO TCH-600 for oxygen, nitrogen and hydrogen, and
LECO CS-200 for carbon and sulphur. Specific surface of
the powders was measured by the BET technique, using a
Monosorb Surface Area equipment, from Quantachrome
Corporation (USA), model MS-13.
Micro-hardness of the different powders was also evalu-
ated in a Vickers micro-hardness tester, model HVS-1000
(TIME Technology Europe). Ten HV
0.01
measurements were
carried out for each type of powder, at a load of 0.098N. The
same range of particle size was selected to better compare
the powders, and avoid size-effects in the measurements.
The behaviour of the different powders during compaction
and ejection was evaluated using an instrumented laboratory
press, the Powder Testing Centre (PTC) [21]. This apparatus
consists of an instrumented cylindrical die operating in a sin-
gle action mode. This press allows continuous recording of
the applied pressure and the pressure transmitted to the sta-
tionary punch during the compaction and ejection processes.
Assuming a rigid behaviour of the die, this press allows the
quantification of the three key properties or factors affecting
the green density, namely the friction at die walls, the pow-
der intrinsic compressibility and the expansion at ejection, as
described in the next section [22].
For all experiments, cylindrical specimens of 7 mm in
height were compacted at 500 MPa at room temperature in a
WC-Co die of 9.525 mm in diameter, and at a pressing rate
of 1 mm/s. At least seven samples were tested for each con-
dition, and the two first tests were used to condition the die
walls and were not considered for the calculations. Results
presented in this study show good reproducibility and corre-
spond to average values obtained from at least five specimens.
As described in Table 1, die wall lubrication was used to
minimize contamination of titanium. A thin layer of zinc
stearate was applied on the die walls using a semi-automated
device adapted to the PTC, based on the same principles of
the patented electrostatic system for industrial presses
[23,24]. In this system, lubricant particles are tribostatically
charged when they are carried by a flow of air through a
small Teflon tube, and are injected in the die cavity in such a
way to minimize turbulence. Excess of lubricant is evacu-
ated through exhaust vents located on the die cover plate. On
the other hand, iron powders were compacted either with
admixed lubricant or using die wall lubrication.
2.1. Analysis of the compaction process
The PTC enables to analyze the compaction process in a
rigid die as a function of two fundamental parameters: the
slide coefficient η, which measures the friction between
powder particles and die walls, and the intrinsic compress-
ibility, which measures the reaction of a powder to an out-
side pressure.
The intrinsic compressibility can be expressed by the rela-
tion between the average in-die density and the average pres-
sure seen by the compact. Considering that the density varies
linearly along the compaction axis as shown by several
researchers [25,26], it can be stated that the density at mid-
height is equal to the average density. Therefore, the average
pressure or net pressure, P
NET
can be evaluated at mid-height
of the compact with equation 1 for a cylindrical compact:
(1)
where Pa is the pressure applied to the compacting punch, Pt
the pressure transmitted to the stationary punch, η the sliding
coefficient, H the height of the compact, and D the diameter
of the compact.
It should be emphasized that the intrinsic compressibility
is only dependant on the intrinsic mechanical behavior of the
powder during compaction. On the other hand, the com-
pressibility, which is defined as the pressure required to
reach a given density or the density obtained for a given
P
NET
Pa*η
H
2D
-------
⎝⎠
⎛⎞
Pa*Pt()
12⁄
==
Table 1. Powders and lubrication used
Powder Supplier Mode of Lubrication Lubricant
Titanium Powders
Ti – Batch 1 GfE * Die-Wall Zinc Stearate
Ti – Batch 2 GfE* Die-Wall Zinc Stearate
Ti – Batch 3 GfE* Die-Wall Zinc Stearate
Iron Powders
ASC100.29 Höganäs** Die-Wall Zinc Stearate
ASC100.29 Höganäs** Admixed 0.7% EBS***
*GfE Metalle und Materialien Gmbh, Germany
**Höganäs, Sweden
***EBS: ethylene bisstearamide (ACRAWAX C from Lonza)
Study of Compaction and Ejection of Hydrided-Dehydrided Titanium Powder 3
pressure, is influenced not only by the powder intrinsic com-
pressibility but also by the friction at die walls and by the
expansion at ejection. In particular, the compact size or
aspect ratio strongly affects the amount of friction at the die
walls and therefore the compressibility, while the intrinsic
compressibility is, on the contrary, independent of the com-
pact aspect ratio.
The compaction process can also be described by the
determination of a slide coefficient η, which gives an evalu-
ation of the level of friction between powder particles and
die walls. The slide coefficient η characterizes the efficiency
of transferring the compaction force throughout the part and
the densification uniformity. The slide coefficient is given by
equation 2,
(2)
where Pa is the pressure applied to the compacting punch, Pt
the pressure transmitted to the stationary punch, F the cross-
section area, S the cross-section perimeter and H the height.
The factor 4F/SH represents the compact aspect ratio or
compact geometry factor. For a cylindrical compact, the fac-
tor 4F/SH is equal to D/H where D is the diameter of the
compact. η can vary between 0 and 1, 0 representing an infi-
nite friction and 1 no friction. Thus, the higher the η, the
lower the friction loss and the better the lubrication and den-
sification uniformity. For a given in-die density, the value of
the slide coefficient proved to be a good parameter to com-
pare the lubrication behavior of similar steel powder mixes
containing different types of lubricants[22,27, 28]. However,
the value of slide coefficient is far from being constant
through the pressing process. The variation of the slide coef-
ficient results, in fact, from the complex evolution of the fric-
tion coefficient and the angle of pressure transmission or
radial to axial stress ratio. However, at high pressures, the
relative movement of particles becomes negligible and the
slide coefficient varies mainly as a function of the friction
coefficient at die walls. The evolution of the coefficient of
friction and the stress ratio during compaction is discussed
elsewhere [29].
A complete ejection curve, as recorded by the PTC, is
shown in Figure 1. The stripping pressure corresponds to the
maximum ejection force, developed at the start the ejection
process divided by the friction surface area. The ejection unit
energy is defined, by the PTC developer, as the energy
required to move the compact from the 0.01 mm to the 2.55
mm punch position (area under the curve within this inter-
val) divided by the friction area of the test compact and by
the travel distance (2.54 mm). The unit is N*m/m
2
/m or J/m
3
.
3. RESULTS AND DISCUSSION
3.1. Particle size and morphology
Images of the iron and titanium powders used in this study
are shown in Fig. 2. The first difference that can be noticed
between Fe and Ti powders is their particle morphology.
Iron ASC100.29 has an irregular morphology, typical of
powder particles produced by water atomization. Among the
three Ti powders, Batch 2 and Batch 3 particles seem to be
more irregular in shape, while Batch 1 seems to be more
angular. The angular morphology of Ti powders derives
from the HDH process, in which titanium is hydrogenated in
order to make it brittle, which provides these fragile fracture
surfaces to the powder particles after milling. Then this pow-
der is dehydrogenated to be converted back to metal Ti and
the particles retain their angular morphology from the former
hydrogenated particles. The HDH process is a relatively
low-cost way to produce Ti powders with low oxygen and
low chlorine contents [30], which is essential to obtain the
highest mechanical properties of Ti.
Figure 3 shows the volume weighed particle size distribu-
η
Pt
Pa
------
⎝⎠
⎛⎞
4F
SH
-------
=
Fig. 1. Typical ejection curve and ejection characteristics measured.
Fig. 2. SEM images of the different powders. (a) Ti-Batch 1, (b) Ti–
Batch 2, (c)Ti-Batch 3, (d) Fe ASC100.29.
4 P. G. Esteban et al.
tions (left) as well as the cumulative volume particle size dis-
tributions (right) for all the powders tested. Batch 1 has a
slightly higher mean size and slightly wider distribution size
than Batch 2 and Batch 3 have approximately the same dis-
tribution shape, but displaced. Fe ASC100.29 shows a differ-
ent distribution shape, clearly wider than all the others. The
characteristic parameters extracted from the particle size dis-
tributions are summarized in Table 2.
3.2. Composition and hardness
In relation with the purity of the powders, it is known that
little content of interstitial elements (oxygen, carbon, nitro-
gen and hydrogen) contributes to dramatically change the
mechanical properties of titanium. These interstitials increase
the elastic modulus, the yield strength and reduce the ductil-
ity of titanium [31]. Nitrogen has generally the most signifi-
cant effect followed by oxygen and carbon [32,33]. While
nitrogen and carbon are usually not found at high concentra-
tions in dense titanium, oxygen is a common contaminant
due to the high affinity of titanium for oxygen and the high
solubility of oxygen in titanium. The total content of intersti-
tial, especially oxygen, is normally higher in the powder par-
ticles with the lower particle size due to their higher specific
surface. The analyses of these impurities have been deter-
mined for the four powders and are summarized in Table 3.
O
Eq
(%at) = O + 1.96*N + 0.52*C (3)[31]
Ti - Batch 1 and Batch 2 have very similar oxygen content,
but their nitrogen is significantly different, what is consistent
with the microhardness values. Indeed, the nitrogen content
of Batch 2 is about 10X higher than Batch 1 and the HV
0.01
of
Batch 2 is about 30 % higher than Batch 1. Conrad et al. [31]
proposed an equivalent oxygen equation to estimate the Vick-
ers hardness. In the equation 3, the effect of the nitrogen con-
tent is 1.96X that of oxygen and the carbon content is 0.52X
of oxygen. Table 3 gives the O
Eq
as calculated with the actual
composition of the powders. However, as showed in works
regarding mechanical properties of Ti foams [34], for high
specific surface materials, it is important to discriminate the
amount of oxygen coming from solid solution from that of
the surface oxide layer. As powder particles have a high spe-
cific area, the main part of oxygen is located as an oxide layer
at the surface of the particles, so this oxygen does not harden
the inside of the particles. Then, instead of total oxygen, only
the contribution of oxygen in solid solution (inside the vol-
ume of the particles) should be considered as a contributor to
the mechanical properties of the titanium powder.
Fig. 3. Particle size distributions (left), and cumulative distributions (right) for the four powders tested.
Table 2. Particle size characteristic parameters of the different
powders
Mean size
[µm]
D
10
[µm]
D
50
[µm]
D
90
[µm]
Ti - Batch 1 54 24 51 88
Ti - Batch 2 48 21 47 77
Ti – Batch 3 119 80 115 166
Fe ASC100.29 97 38 88 169
Table 3. Hardness and Chemical analyses of O, N, C and H for the four powders tested
Hardness
HV
0.01
Specific
surface [m
2
/g]
wt.% O wt.% N ppm H wt.% C
at.% O
Eq
(1)
(wt.%)
at.% O
SolEq
(2)
(wt.%)
Ti – Batch 1 127 ± 24 0.09 0.319 ± 0.003 0.008 ± 0.001 57 ± 2 0.009 ± 0.001 1.020(0.343) 0.543(0.182)
Ti – Batch 2 167 ± 32 0.1 0.343 ± 0.002 0.098 ± 0.004 106 ± 2 0.013 ± 0.002 1.693(0.572) 1.146(0.385)
Ti – Batch 3 159 ± 20 0.05 0.250 ± 0.006 0.069 ± 0.004 74 ± 2 0.013 ± 0.001 1.228(0.414) 0.955(0.321)
Fe ASC100.29 103 ± 16 0.02 0.083 ± 0.005 0.003 ± 0.001 4 ± 2 0.006 ± 0.001 - -
(1)
Total equivalent oxygen content (at.%) - see equation 3.
(2)
Equivalent oxygen calculated with interstitial in solid solution in the powder particles.
