International Journal of Fatigue 164 (2022) 107129

Contents lists available at ScienceDirect

International Journal of Fatigue

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

Fatigue performance of zirconia-reinforced Ti-6Al-4V nanocomposite

processed by laser powder bed fusion: An improvement by hot
isostatic pressing
Benjamin Guennec a, *, Amine Hattal b, c, Azziz Hocini b, Kamilla Mukhtarova d,
Takahiro Kinoshita a, Noriyo Horikawa a, Jenő Gubicza d, Madjid Djemaï c, Guy Dirras b, *
a
Toyama Prefectural University, College of Engineering, Department of Mechanical Systems Engineering, 939-0398, Kurokawa 5180, Imizu, Toyama, Japan
b
Université Sorbonne Paris Nord, Laboratoire des Sciences des Procédés et des Matériaux (LSPM) - UPR CNRS 3407, 99 avenue, Jean-Baptiste Clément, 93430
Villetaneuse, France
c
Z3Dlab, Parc Technologique, 26 Rue des Sablons, Montmagny 95360, France
d
Department of Materials Physics, Eötvös Lorand University, Budapest, P.O.B. 32, H-1518, Hungary

A R T I C L E I N F O A B S T R A C T

Keywords: For the first time, four-point bending fatigue behavior of a Ti-6Al-4V alloy reinforced with 1 wt% nano-yttria-
High cycle fatigue stabilized zirconia and processed by laser powder bed fusion was investigated. In a quest for clarification on
Metal Matrix Composites the effect of the hot isostatic pressing (HIP) on the fatigue outcomes, both untreated and HIPed specimens were
Titanium alloy
studied. The former material indicated relatively modest fatigue resistance mainly due to the presence of
Additive manufacturing
Hot isostatic pressing
inherent processing defects, whereas the latter one revealed a promising fatigue strength in comparison with
unreinforced Ti-6Al-4V and exhibited transgranular fatigue crack nucleation provoked by the crystallographic
slip in long-length layers of α grain.

1. Introduction
Additive manufacturing (AM) is now widely recognized as a decisive
paradigm change in the metallurgy field. Beyond its numerous advan­
tages deriving from its specific layer-by-layer process fabrication (i.e.,
no offcut, ability to print complex shape parts, low energy consumption
[1–5]), this technology provides a way more practical path to creating
metallic matrix composites (MMCs) than the conventional methods,
such as casting technique [6] or powder metallurgy [7]. Implementation
of the AM technology to develop ceramics reinforced Ti-alloys has
already been attempted, using B4C, TiC, TiB [8,9], and SiC [10] as
strengthening particles. Recently, Hattal et al. [11–13] have developed a
Ti-6Al-4V reinforced by 1 or 2.5 wt% of nano-yttria-stabilized zirconia
(nYSZ) melted by laser powder bed fusion (L-PBF) technique. The choice
of zirconia is motivated by its high-strengthening potential and its good
thermodynamical affinity toward Ti-alloys [11]. Even though the
addition of zirconia has involved a minor change of the general micro­
structure, a significant increase in the hardness and compression
strength have been reported. Furthermore, the samples subjected to hot
isostatic pressing (HIP) post-treatment have highlighted elevated
relative bulk densities, which reflect the curation of the structural de­
fects and a significant increase in its compressive strain at rupture [13].
Such characteristics are promising for industrial applications in several
fields such as aerospace, medical, and petrochemicals.
Even though the static strength of AMed Ti-6Al-4V is widely reported
to overcome its conventionally fabricated counterparts [14–18], the
practical application of such newly developed materials requires a solid
comprehension of its response against complex loadings. One of the
main issues is the poor fatigue strength of AMed materials with respect
to their conventional wrought counterparts. This feature is caused by
processing defects, such as lack-of-fusion (LOF) flaws or gas-induced
pores, implying premature fatigue crack nucleation [19–25]. HIP post-
processing is commonly carried out to cure such microstructural de­
fects, which results in a significant increase in the fatigue resistance of
AMed Ti-6Al-4V alloys [15,23,26–28]. Nevertheless, the fatigue resis­
tance obtained from such HIPed samples may be comparable to the fa­
tigue strength of conventionally fabricated materials [14,28,29]. These
HIPed materials have been prone to either surface-induced [26,29,30]
or interior-induced crack nucleation mechanisms [14–16,25]. Taking
aside flaw-induced fatigue crack initiation, only a restricted number of

* Corresponding authors.
E-mail addresses: bguennecpro@gmail.com (B. Guennec), guy.dirras@lspm.cnrs.fr (G. Dirras).

https://doi.org/10.1016/j.ijfatigue.2022.107129
Received 29 April 2022; Received in revised form 9 June 2022; Accepted 8 July 2022
Available online 13 July 2022
0142-1123/© 2022 Elsevier Ltd. All rights reserved.
B. Guennec et al.
articles have discussed the intrinsic crack nucleation mechanism of
AMed Ti-6Al-4V [14,22,28,29]. Such reports have outlined the genera­
tion of transgranular facets induced by cleavage rupture of hexagonal
close-packed (HCP) α-phase in the vicinity of the crack nucleation site,
similar to conventionally forged alloys [31,32]. Nevertheless, the liter­
ature lacks investigations on this issue.
The present work will investigate the four-point fatigue properties of
Ti-6Al-4V reinforced by 1 wt% nYSZ prepared by the L-PBF technique.
The effect of the HIP post-treatment on the microstructure, the dislo­
cation activity, and the fatigue mechanisms will also be scrutinized.
2. Experimental procedure
2.1. Materials, manufacturing process, and post-heat treatment
Ti-6Al-4V reinforced by 1 wt% nYSZ, hereafter denoted “ZTP1”, was
fabricated by mixing Ti-6Al-4V and nYSZ powders and then melted by
the L-PBF technique. The 20-μm average size Ti-6Al-4V particles [33]
possess a chemical composition indicated in Table 1. The nYSZ powder
is a 3 mol% yttria-stabilized tetragonal zirconia with an average particle
size of 40 nm and a chemical composition tabulated in Table 2. The parts
were manufactured on an SLM125 HL machine from SLM Solution
GmbH. L-PBF process parameters were used based on a previous study
[11]. Further details about the powder blending protocol and L-PBF
process parameters can be found in [13].
After L-PBF process, the manufactured parts were subjected to stress
relief heat treatment, operated at 600 ◦ C for 2 h, in an argon-protected
environment (AMS 2801 standards). At this stage, the obtained parts
will be denoted as “HT” in the rest of the present article. To analyze the
HIP post-processing effect, some parts were subjected to F3001 standard
HIP for Ti-6Al-4V materials (920 ◦ C, 100 MPa, 2 h), afterward. Conse­
quently, the resulting specimens will be referred to as “HT + HIP”.
2.2. Four-point bending fatigue testing conditions
Fatigue tests were carried out by Instron Universal 5848 machine, in
air, at room temperature. Four-point bending loading frame has
considered an outer- and inner-span distances of So = 30 mm and Si =
10 mm, respectively. Parallelepiped shape fatigue specimen with nom­
inal dimensions of length L = 40 mm, width b = 3 mm, and thickness h =
2 mm were prepared. These fatigue testing conditions have already been
reported to be adapted to the investigation of the fatigue properties of
Ti-6Al-4V materials [34,35], using limited sized specimens in compar­
ison with fatigue axial loading technique for the sake of material con­
sumption concern. The building direction is parallel with the specimen
thickness, implying that the principal stress direction is orthogonal to
the building direction. Every specimen was polished by SiC grinding
paper, using #1200 grade for the final polishing operation carried out on
side and tensile surfaces, except further indications. The obtained
arithmetical surface roughness was estimated at Ra = 60 ± 10 nm, using
a LEXT OLS44500 laser microscope device (Olympus Corporation).
Stress concentration at specimen edges was relieved by introducing
chamfers. Under this configuration, the bending stress σb exerted on the
tensile surface of the specimen can be assessed by Eq. (1): [36].
3P(So − Si )
σb = (1)
2bh2
where P is the total applied load. Load control fatigue testing procedure
was operated at a stress ratio R (=σmin/σmax) of 0.1. Loading frequency
Table 1
Chemical composition of the as-received Ti-6Al-4V powder.
Element Ti Al V O C N H Fe
wt% Bal. 6.2 4.0 0.07 0.01 0.02 0.002 0.14
2
International Journal of Fatigue 164 (2022) 107129
Table 2
Chemical composition of the as-received nYSZ nanopowder (Data from
USNano™).
Particle ZrO2 Y2O3 Al Mg Si Ca S Nb
Ppm Bal. 51,000 20 65 102 75 165 119
was set at 5 Hz, and tests were terminated after 107 cycles unless earlier
specimen failure occurrence.
2.3. Stress intensity factor calculation
The stress intensity factor (SIF) K related to the onset of crack
propagation at a flaw has been estimated via the method proposed by
Murakami and Endo [37], as expressed in the following formula:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
K = αK σf π × area1/2 (2)
where σf is the stress level at the flaw location, area is the projected area
of the flaw in the principal stress direction, and αK is either 0.65 or 0.5,
corresponding to flaw positioned on or beneath the tensile surface. In
accordance with the fully elastic bending model (i.e., Eq. (1)), if the
defect is located inside the specimen, the value of σ f is determined by the
distance between the specimen neutral fiber and the flaw position. The
SIF range will be denoted as ΔK (=Kmax − Kmin) in the rest of the present
article.
2.4. Analysis conditions for studying the density and the microstructure
Density measurements were performed on dried parts before and
after HIP post-treatment (parts were cleaned in an ultrasonic bath and
then dried for 24 h) using an Accupyc II 1340 He pycnometer apparatus.
Further details are given in [13] and [38]. The obtained experimental
density is then divided by the reference density assessed via the mixture
law, thereby obtaining the relative density. The initial microstructure
characterization was performed by scanning electron microscopy (SEM)
using FEG SUPRA 40VP-ZEISS microscope on samples that were
extracted from the parts. Electron backscatter diffraction (EBSD) in­
vestigations were further carried out on electropolished samples
considering scan step size and surface area of 80 nm and 110 µm × 110
µm, respectively. OIM Analysis v.7 software was used to perform EBSD
data analysis. The β-phase fraction was investigated on backscattered
electron microscopy. To this end, the sample surfaces were etched using
Kroll’s reagent. After etching, the β-phase appeared in bright contrast, as
shown in Fig. S1 of the Supplementary material. In addition, various
failed specimen observations, including the study of the fracture surface,
were carried out by JEOL JSM-7800F SEM. This microscope was also
equipped with an EDAX TSL camera for EBSD acquisition. Such data
were post-processed by OIM Analysis v.8 software.
Moreover, X-ray line profile analysis (XLPA) has been operated to
assess the crystallite size and the dislocation density in the fatigued
specimens [39]. The X-ray diffraction (XRD) patterns were then
measured by a RA-MultiMax9 type high-resolution rotating anode
diffractometer (Rigaku Corporation), using CuKα1 radiation at a wave­
length of λ = 0.15406 nm. The height and the width of the rectangular X-
ray spot on the sample surface were 1.5 and 0.2 mm, respectively. Due to
the mechanical polishing process performed during the preparation of
the fatigue samples, the uppermost surface layer with a thickness of
5–10 µm was removed by HF etching for 20 s before XLPA execution.
The measured peak profiles of the matrix were evaluated by the con­
volutional multiple whole profile (CMWP) fitting procedure [40]. In this
method, the diffraction pattern is fitted by the sum of a background
spline and the diffractions peaks obtained as the convolution of the
measured instrumental peak and the theoretical profiles caused by the
finite crystallite size and the dislocations. The instrumental peaks were
measured on a standard NIST SRM660a LaB6 sample. Under these
B. Guennec et al.
circumstances, the area-weighted mean crystallite size and the disloca­
tion density were determined by the CMWP procedure. In addition, the
population of the different dislocation slip systems can also be obtained
by an analysis of the dislocation contrast factors determined by the
CMWP method. As an example, Fig. S2 in the supplementary file shows
the CMWP fitting for a fatigued HT specimen, revealing the excellent
correlation between the fitting curve and the diffraction signal. More
details of this evaluation procedure are given in [39]. Furthermore, the
fractions of the remnant Ti β-phase were estimated by the intensity
fraction of this phase in the whole diffractogram.
3. Experimental results and discussion
3.1. Microstructure of the investigated materials
The microstructures of both HT and HT + HIP metallurgical states
are shown in the inverse pole figure (IPF) maps in Fig. 1. It is clear that
these microstructures are mainly composed of HCP-Ti α grains with
different thicknesses and lengths in addition to a heterogeneous crys­
tallographic orientation, due to the thermal history of the L-PBF process
[41]. Due to the near β-transus temperature of the HIP post-treatment,
HT + HIP microstructure presents α-grain thicker laths. No significant
alteration of the crystallographic orientation by the HIP post-treatment
is observed, in line with its primary porosity curation goal. The relative
densities of the HT and HT + HIP samples were measured as 99.33 ±
Fig. 1. GB and IPF maps of the initial microstructure of samples (a) HT, Ti α-phase; (b) HT, β-phase; (c) HT + HIP, α-phase; and (d) HT + HIP, β-phase.
3
International Journal of Fatigue 164 (2022) 107129
0.03 and 99.99 ± 0.01 %, respectively. These results underline the
effectiveness of the HIP post-treatment in the elimination of structural
defects, such as encapsulated porosities inherited from atomization
process. The BCC β phase content was also investigated in this study
using EBSD and XRD. The former technique has detected a proportion of
11.3 % and 5.1 % for HT and HT + HIP samples, respectively. Using the
diffraction data of phase Ti0.7V0.3 (card No. 01–081-9817), the XRD
peak intensity fractions were 10 ± 2 % and 4 ± 1 % for HT and HT + HIP
samples.
Furthermore, the dislocation density imposed by the four-point
loading fatigue has been measured by XLPA on initial and three
distinct ruptured specimens for each material, paying attention to
conduct the measure sufficiently away from the fracture surface.
Regardless of the investigated material and the applied stress level, the
dislocation density was practically unchanged at a value of approxi­
mately 1.0 × 1014 m− 2. The negligible effect of HIP after the heat
treatment on the dislocation density can be explained by the high
temperatures of both processing steps. Namely, the dislocation density
in the L-PBF samples decreased to a low level even after the HT treat­
ment, therefore the high temperature of HIP did not cause additional
significant reduction in the dislocation density. Furthermore, both ma­
terials were subjected to fatigue experiment in the high-cycle domain. In
this domain, the majority of grains are deformed cyclically at a very
limited plastic strain, which can yield only a marginal change in the
dislocation density. It should be noted that small changes in the
B. Guennec et al.
dislocation density cannot be detected due to uncertainty of the exper­
imental values by XLPA method (typically ~ 0.2 × 1014 m− 2). In addi­
tion, no considerable evolution of the crystallite size was observed
during HIP or fatigue testing. It is also worth noting that XLPA enables to
analyze the type of dislocations. For every sample, 70 to 90% of dislo­
cations have 〈a〉-type Burgers vector, and more than half of these dis­
locations are lying on prismatic or pyramidal glide planes. It was also
found by XRD that the fatigue loading had no significant effect on the β
phase fraction.
3.2. Fatigue resistance
The fundamental fatigue resistance of the investigated materials has
been depicted in Fig. 2, in the form of an S-N diagram. The fatigue
strength regressions, drawn by solid lines, obey the JSMS-SD-6-08
Standard [42]. It should be noted that the regression corresponding to
HT + HIP material (red line) excludes the unique “HT + HIP, OPS”
datum (open dot) due to a distinct tensile surface finishing condition
(further details below). Furthermore, five sets of fatigue data obtained
on L-PBF melted unreinforced Ti-6Al-4V in the literature have been
included [20,26,27]. All these works involved ground surface specimens
to avoid fatigue crack nucleation phenomenon generated by surface
roughness. It should be mentioned that these data were obtained from
axial loading fatigue tests at R = 0.1, whereas the present study has
assessed the fatigue strength of the ZTP1 in plane bending loading
condition. Due to the stress gradient imposed by the bending loading
condition, a moderate increase in the fatigue strength of approximately
10% in comparison with the axial one in the long-life region has been
reported for steels and aluminum alloys [43]. Recently, Aguado-
Montero et al. [44] have carried out four-point bending fatigue experi­
ments on AMed Ti-6Al-4V alloys subjected to several surface post-
treatments involving compressive residual stress generation (i.e., sand
blasting, shot peening or laser peening). Even though such experimental
conditions are too distant for valuable straight comparison with the ones
analyzed in the present work, this report has underlined no specific
discrepancies between the fatigue lives obtained from axial and four-
point bending loadings, considering similar outcomes available in the
literature [44]. Based on such experimental results, the fatigue resis­
tance levels of AMed Ti-6Al-4V deriving from axial and four-point
bending loading conditions should be close one to the other, provided
that such results are related to same stress ratio R. Consequently, the
direct comparison of the fatigue data depicted in Fig. 2 is relevant to
highlight the general trend of the fatigue resistance of the investigated
materials, as a first approximation.
Fig. 2. S-N diagram of the investigated materials along with external fatigue
data from unreinforced Ti-6Al-4V processed by L-PBF [20,26,27]. Asterisk-
marked specimen notify the unique HT + HIP specimen failed by microstruc­
tural defects.
4
International Journal of Fatigue 164 (2022) 107129
In accordance with the usual trend already mentioned in the intro­
duction, HIP post-treatment has significantly improved the fatigue
strength of the nYSZ reinforced Ti-6Al-4V sample. Indeed, the nYSZ
reinforced HT material has revealed a fatigue strength equivalent to its
annealed unreinforced Ti-6Al-4V counterparts, resulting in a fatigue
endurance level at 107 cycles of 431 MPa. On the other hand, the HT +
HIP material has underlined a considerable increase in fatigue strength,
reaching the superior fatigue endurance level of 844 MPa. Therefore, the
fatigue endurance level was approximately doubled by HIP post-
treatment. Fig. 2 outlines that this endurance level is significantly
higher than other unreinforced HIPed Ti-6Al-4V. Better still, this fatigue
endurance level exceeds the one from wrought Ti-6Al-4V, which high­
lights a fatigue endurance of approximately 800 MPa under similar
conditions [45]. Despite the disparity of the loading conditions with the
data presently available in the literature, which may slightly alter the
fatigue strength, ZTP1 prone to adapted post-treatments exhibits a sig­
nificant fatigue strength gap of approximately 150 MPa with its unre­
inforced AMed Ti-6Al-4V counterparts. Such a stress level discrepancy
strongly suggests an improvement in the fatigue endurance at 107 cycles
by the addition of nYSZ strengthening particles. Nevertheless, from a
rigorous experimental methodology viewpoint, such a development still
needs to be confirmed by further experimental results obtained from
strictly equivalent fatigue condition.
3.3. Fatigue crack initiation observation and mechanism analysis
A summary of the fracture surface observation has been depicted in
Fig. 3. Fracture surfaces of the HT material have revealed the presence of
obvious defects, as shown in Fig. 3(a) and (b). It should be noted that
these defects may be located either on (Fig. 3(a)) or beneath the tensile
surface (Fig. 3(b)), even though the latter configuration was the most
observed. In accordance with these defects’ sizes (diameter ~ 100 μm)
and the relative dispersion of their aspect ratios (ranging from 1.12 to
3.20), such flaws were clearly related to local LOF phenomenon, which
is extremely common for AMed materials [19,20,23,25,28,30,46].
Analysis of the SIF range induced by such defects ΔKLOF has been carried
out, resulting in the diagram presented in Fig. 4(a). Taking aside one
diverging point located in the long-life region, the distribution of ΔKLOF
has presented a decreasing logarithmic regression with the fatigue life
Nf. Such a trend is very common in conventional metallic materials [37].
As highlighted by the inset of Fig. 4(a), the crack initiation site of the
diverging point has underlined the presence of two distinct LOF defects
close one the other. Under these circumstances, it is suggested to
consider the effective area determined by both flaws [47], as highlighted
by the red dotted line in the inset of Fig. 4(a). Such a very extended flaw
area results in a relatively large ΔKLOF, despite the low applied stress
level of 480 MPa. Therefore, the dissonant character of this peculiar
datum may indicate that only one of these defects has effectively
contributed to the fatigue crack nucleation mechanism. Furthermore,
since LOF defects intrinsically govern the fatigue failure mechanism of
HT material, one can estimate the theoretical fatigue endurance σw,HT by
analyzing the distribution of the defect size. Following the procedure
undertaken in [47], the defect size has been fitted by generalized
extreme value (GEV) distribution, as shown in Fig. 4(b). Considering the
5% largest flaw size to assure a high confidence level of theoretical fa­
tigue endurance obtained, the corresponding LOF defect size was
(areaLOF)1/2
max = 242 μm. Consequently, the theoretical fatigue endurance
σ w,HT (expressed in maximum stress level) has been estimated as [48]:
( )− α
HV + C 1− R
σ w,HT = C1 ( √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅)2 1/6 , (3)
areaLOF max 2
where HV is the average material hardness (440 HV [13]), C1 is set at
1.53 (corresponding to interior-induced fatigue crack initiation), C2 is
120 HV, and α = 0.28 is the Walter coefficient. Numerical application
has provided a value σw,HT = 438 MPa, which is very close to the
B. Guennec et al. International Journal of Fatigue 164 (2022) 107129

Fig. 3. SEM observation of the crack initiation site. (a) HT material, a LOF flaw located on the tensile surface is indicated by a white arrow, σmax = 690 MPa, Nf =
2.30 × 104; (b) HT material, a LOF flaw located inside the specimen is indicated by a white arrow, σmax = 510 MPa, Nf = 6.91 × 104; (c) HT + HIP material, a
microstructural defect located on the tensile surface is indicated by a white arrow, σmax = 850 MPa, Nf = 8.17 × 105; and (d) Tilted micrograph of the HT + HIP
material, σmax = 1100 MPa, Nf = 1.80 × 105.

Fig. 4. (a) Stress intensity factor range induced by defects at the crack initiation site over the specimen fatigue life, and (b) LOF defect size fitted by GEV distribution.

experimental one.
Considering now the case of the HT + HIP material, LOF defects were
not detected at the crack initiation site. This observation is consistent
with the beneficial effect of the HIP on the material density, as discussed
earlier. Therefore, this result demonstrates that standard HIP post-
processing operation effectively nullifies the impact of LOF flaws on
the fatigue properties of ZTP1. In addition, every specimen has high­
lighted a crack initiation site on the tensile surface. A unique specimen
has underlined the presence of a microstructural defect at the crack
initiation site, as depicted in Fig. 3(c). The corresponding fatigue datum,
highlighted by an asterisk in Fig. 2, has revealed poorer fatigue
resistance than other specimens. Therefore, this defect has undoubtedly
provoked a premature cracking phenomenon. Based on this defect
configuration, the SIF range induced by the fatigue crack initiation ΔKd
has been included in Fig. 4(a). Compared to the general trend observed
from the HT material, this datum has been shifted into the long-life re­
gion. This behavior suggests that the HT + HIP material is more resistant
to fatigue crack propagation in the near-threshold region than the HT
one. Even though a specific study of the fatigue crack propagation
characteristics is mandatory to confirm this development, this result is in
line with the significant beneficial effect of the HIP treatment on the SIF
range threshold value and the near-threshold crack propagation rate of

5
B. Guennec et al.
unreinforced Ti-6Al-4V [22,27]. Taking aside the previously elaborated
specimen, fatigue crack initiation sites in HT + HIP material have
depicted a typical configuration presented in Fig. 3(d), where the
specimen was slightly tilted to observe the fracture and the tensile sur­
faces. Despite the presence of debris on the tensile surface, one can
detect the presence of a series of plastic features on the tensile surface
below the crack initiation site (see inset). Given the microstructure of
such AMed materials, this configuration suggests that an entire α-phase
grain colony has been plastically deformed. Since the orientation of the
crack initiation is strictly parallel with these plastic marks, the crack
initiation is strongly thought to be nucleated by an intragranular slip
mechanism. It should be noted that no facets were observed in the vi­
cinity of the crack initiation site of the HT + HIP specimens.
Since the present crack nucleation mechanism has occurred on the
tensile surface of the specimen, an additional investigation has been
conducted through the EBSD technique in the vicinity of the crack
initiation site. To this end, the tensile surface of one specimen was
ground up to #4000 grade and finished by silica suspension for effective
data acquisition. This sample was then prone to a fatigue test at σ max =
950 MPa, resulting in rupture at Nf = 4.67 × 105 cycles, as shown in the
S-N diagram (Fig. 2) by the HT + HIP, OPS nomenclature. Since this
specimen has highlighted a fatigue strength similar to regular ones, the
effect of the specimen surface condition on the fatigue strength of HIPed
ZTP1 should be minor for a surface roughness equivalent to the typical
specimens, at least.
Fig. 5. Fatigue crack nucleation mechanism analysis in HT + HIP, OPS specimen. (a) Broken piece fatigue crack initiation from tensile surface viewpoint; (b) IPF map
of the region (step size of 0.15 μm); (c) Reciprocal broken piece fatigue crack initiation from tensile surface viewpoint; and (d) IPF map of the region (step size of 0.15
μm). IPF maps depict results related to a CI larger than 0.100.
6
International Journal of Fatigue 164 (2022) 107129
The crack initiation site of the related specimen for each broken
piece has been observed from the tensile surface viewpoint, as depicted
in Fig. 5(a) and (c). In line with Fig. 3(d), the formation of aligned plastic
features can be detected on the tensile surface. High magnification mi­
crographs, inserted as insets, have underlined the concentration of
several marks, almost parallel to each other. Since the orientation of the
crack initiation on the tensile surface is the same as the direction of these
features, the crack nucleation mechanism is definitively governed by the
local plastic mechanism involved. The elucidation of this mechanism
requires additional information dealing with the microstructure in the
vicinity of the initiation site. To this end, EBSD acquisitions on the
tensile surface of each broken piece have been operated. The obtained
results have been presented via superposition of the probed region SEM
photograph and the IPF map based on the processed data software, as
shown in Fig. 5(b) and (d). As confirmed by the insertion of schematic
lattices in blue color to assess the crystallographic orientation, these
pictures have underlined the presence of α-phase oriented in the same
manner at the edge of both broken pieces. Such a configuration confirms
the occurrence of a transgranular fatigue crack nucleation induced by
plastic deformation of the scattered lath.
Furthermore, other long-length α-phase layers similarly oriented
have also been prone to severe plastic deformation (see Fig. 5(d)). This
analysis demonstrates that the long-length α layer presenting an unfa­
vorable orientation governs the crack initiation mechanism. In accor­
dance with Joseph et al. [49], fatigue loading generates dislocation pile-
B. Guennec et al.
ups lying on the principal slip planes of Ti α-phase grains, showing
limited interaction with other slip systems. These pile-ups have opposite
signs on the neighboring slip planes. Depending on the microstructure of
the α + β Ti-alloys, this dislocation configuration results in a competition
between crystallographic slip and cleavage fatigue cracks nucleation
mechanisms [31,50]. On the one hand, forged Ti-6Al-4V with equiaxed
or bimodal microstructure tends to generate α facets in the vicinity of the
crack initiation site by cleavage fracture on {1017} plane [32]. Indeed,
such microstructures tend to hinder crystallographic slip, generating
intense accumulation of the pile-ups at GBs throughout the primary
activated slip planes. This configuration results in local stress rises be­
tween close activated slip planes having pile-ups with opposite signs,
which is able to initiate the cleavage rupture [49,51].
On the other hand, lamellar structured Ti-6Al-4V alloy (also
commonly referred to as “basketweave microstructure”), which presents
elongated α grains aligned in the same direction inside each prior
β-grain, tends to highlight slip-induced fatigue crack nucleation mech­
anism [50,52]. The lamellar structure favors the plastic slip phenome­
non by the presence of extended dislocation slip distances, such as the
plastic deformation can be conveyed in a limited number of activated
planes of the primary slip system. In addition, the back stress induced by
the dislocation pile-ups necessarily generates a slip irreversibility be­
tween the loading and the unloading stages of the fatigue stressing cycle.
As discussed by Lin et al. [53], this irreversibility tends to create dislo­
cation loops due to a local accumulation of vacancies, and these vacancy
clusters are able to nucleate a fatigue crack along slip bands. Due to the
relative similarities of the microstructure in forged lamellar and AMed
Ti-6Al-4V alloys, observing a slip-induced crack initiation mechanism in
the presently investigated material is reasonable. Nevertheless, as far as
the authors know, the present work is the first clear report of plastic slip-
induced fatigue crack nucleation phenomenon in AMed Ti-6Al-4V,
despite the presence of elongated α-grain layers in these microstruc­
tures. Therefore, the high fatigue strength obtained by HT + HIP ma­
terial stems reasonably from the impeded cleavage-induced fatigue
crack nucleation. Moreover, in line with [50], a moderate limitation in
the dimension of the long-length α-grain colony to restrict the maximum
slip distance should be a promising method to reach even higher fatigue
endurance levels of the ZTP1 material.
Finally, the EBSD acquisition allows a quick analysis of the plastic
slip activity based on the crystal orientation of the scattered α-grain lath.
The maximum Schmid factor (SF) of the different slip system groups
have been tabulated in Table 3. Due to the low SF values, the activation
of 〈a〉-type prismatic and pyramidal dislocations is difficult, although
there is an abundance of these dislocations in the studied materials as
revealed by the XLPA method. Therefore, the plasticity inside these
grains requires high stresses, which may induce fatigue cracking in this
part of the material. It should be noted that although the pyramidal
〈a + c〉slip system has a high SF (see Table 3), its activation at room
temperature is also difficult due to the large Burgers vector and the high
friction stress for these dislocations [54].
4. Conclusions
The present work has investigated for the first time the four-point
bending fatigue features in the high-cycle domain of 1 wt% nYSZ-
reinforced Ti-6Al-4V alloy melted by laser powder bed fusion technol­
ogy (i.e., ZTP1). A particular attention was paid to the crack initiation
mechanism of such novel metal matrix composites fabricated via AM
technology. The stress-relieved (HT) composite material exhibited
similar or slightly lower fatigue strength than its annealed unreinforced
AMed Ti-6Al-4V counterparts due to early fatigue crack nucleation at
lack-of-fusion flaws. On the other hand, HIP post-processed composite
material (HT + HIP) showed an elevated fatigue endurance level (σ max
= 844 MPa), which strongly suggests an improvement of the fatigue
endurance in respect to its unreinforced AMed Ti-6Al-4V counterparts.
7
International Journal of Fatigue 164 (2022) 107129
Table 3
Crystal orientation of the α-grain alongside crack initiation site and associated
maximum Schmid factor values.
Crystal Orientation* Max SF Max SF Max SF Max SF
φ1 (◦ ) Φ φ2
Basal Prismatic Pyramidal Pyramidal 〈a
〈a〉 + c〉
(◦ ) (◦ )
170.8 88.3 30.0 0.151 0.013 0.078 0.469
*Euler angles expressed in Bunge convention.
Such an endurance value also suggests similar fatigue resistance of HT +
HIP material to conventional wrought Ti-6Al-4V, at least. Furthermore,
for the first time, a clear crack nucleation mechanism governed by
plastic slip has been reported for AMed Ti-6Al-4V without cleavage
facets. In a way similar to forged lamellar Ti-6Al-4V, this phenomenon is
promoted by the elongated α-grain colonies, which potentially generate
long slip distances if the activated systems are approximately parallel
with the geometrical orientation of the long-length α layers. Combined
with its high static strength compared to unreinforced AMed Ti-6Al-4V,
this singular crack nucleation mechanism is thought to be the main
reason for the high fatigue strength. Moreover, the occurrence of this
nucleation mechanism implies that a limitation of the maximum plastic
slip distance by the reduction of the α-grains length should consist in an
effective way to reach even higher fatigue strength. Both HIP processing
and fatigue loading have no considerable effects on the overall dislo­
cation density, regardless of the applied stress level. Furthermore, most
dislocation had 〈a〉-type Burger vector and prismatic or pyramidal glide
planes. Consequently, ZTP1 subjected to adapted post-treatment pre­
sents promising characteristics for actual applications as a structural
material from its fatigue characteristic viewpoint.
5. Data availability
The raw/processed data required to reproduce these findings cannot
be shared at this time as the data also forms part of ongoing research
project.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgments
The support by the Université Sorbonne Paris Nord (USPN)
“CARAMEL” platform facility funded by Région Île-de-France and the
CNRS for the HIP post-treatments is very welcomed.
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ijfatigue.2022.107129.
References
[1] Emmelmann C, Sander P, Kranz J, Wycisk E. Laser Additive Manufacturing and
Bionics: Redefining Lightweight Design. Phys Proc 2011;12:364–8. https://doi.
org/10.1016/j.phpro.2011.03.046.
[2] Herzog D, Seyda V, Wycisk E, Emmelmann C. Additive manufacturing of metals.
Acta Mater 2016;117:371–92. https://doi.org/10.1016/j.actamat.2016.07.019.
B. Guennec et al.
[3] Khanna N, Zadafiya K, Patel T, Kaynak Y, Rashid RAR, Vafadar A. Review on
machining of additively manufactured nickel and titanium alloys. J Mater Res
Technol 2021;15:3192–221. https://doi.org/10.1016/j.jmrt.2021.09.088.
[4] Tshephe TS, Akinwamide SO, Olevsky E, Olubambi PA. Additive manufacturing of
titanium-based alloys- A review of methods, properties, challenges, and prospects.
Heliyon 2022;8:e09041. https://doi.org/10.1016/j.heliyon.2022.e09041.
[5] Monteiro H, Carmona-Aparicio G, Lei I, Despeisse M. Energy and material
efficiency strategies enabled by metal additive manufacturing – A review for the
aeronautics and aerospace sectors. Energy Rep 2022;8:298–305. https://doi.org/
10.1016/j.egyr.2022.01.035.
[6] Xiao L, Lu WJ, Qin JN, Chen YF, Zhang D, Wang MM, et al. Steady-state creep of in
situ TiB plus La2O3 reinforced high temperature titanium matrix composite. Mater
Sci Eng A 2009;499:500–6. https://doi.org/10.1016/j.msea.2008.09.002.
[7] Koo MY, Park JS, Park MK, Kim KT, Hong SH. Effect of aspect ratios of in situ
formed TiB whiskers on the mechanical properties of TiBw/Ti–6Al–4V composites.
Scr Mater 2012;66:487–90. https://doi.org/10.1016/j.scriptamat.2011.12.024.
[8] Ni DR, Geng L, Zhang J, Zheng ZZ. Effect of B4C particle size on microstructure of
in situ titanium matrix composites prepared by reactive processing of Ti–B4C
system. Scr Mater 2006;55:429–32. https://doi.org/10.1016/j.
scriptamat.2006.05.024.
[9] Yang ZW, Fu LQ, Wang SL, Zhang M, Wang Y, Ma ZQ, et al. Balance of strength and
plasticity of additive manufactured Ti-6Al-4V alloy by forming TiB whiskers with
gradient distribution. Addit Manuf 2021;39:101883. https://doi.org/10.1016/j.
addma.2021.101883.
[10] Sivakumar G, Ananthi V, Ramanathan S. Production and mechanical properties of
nano SiC particle reinforced Ti–6Al–4V matrix composite. T Nonferr Metal Soc
2017;27:82–90. https://doi.org/10.1016/S1003-6326(17)60009-8.
[11] Hattal A, Chauveau T, Djemai M, Fouchet JJ, Bacroix B, Dirras G. Effect of nano-
yttria stabilized zirconia addition on the microstructure and mechanical properties
of Ti6Al4V parts manufactured by selective laser melting. Mater Des 2019;180:
107909. https://doi.org/10.1016/j.matdes.2019.107909.
[12] Hattal A, Djemai M, Fouchet JJ, Chauveau T, Bacroix B, Hocini A, et al. Titanium
Ti6Al4V alloy reinforced by the addition of nano yttria stabilized zirconia
fabricated by selective additive manufacturing: microstructure and mechanical
investigations. MATEC Web Conf 2020;321:03018.
[13] Hattal A, Mukhtarova K, Djemai M, Chauveau T, Hocini A, Fouchet JJ, et al. Effect
of hot isostatic pressing on microstructure and mechanical properties of Ti6Al4V-
zirconia nanocomposites processed by laser-powder bed fusion. Mater Des 2022;
214:110392. https://doi.org/10.1016/j.matdes.2022.110392.
[14] Rafi HK, Karthik NV, Gong H, Starr TL, Stucker BE. Microstructures and
mechanical properties of Ti6Al4V parts fabricated by selective laser melting and
electron beam melting. J Mater Eng Perform 2013;22:3872–83. https://doi.org/
10.1007/s11665-013-0658-0.
[15] Mower TM, Long MJ. Mechanical behavior of additive manufactured, power-bed
laser-fused materials. Mat Sci Eng A 2016;651:198–213. https://doi.org/10.1016/
j.msea.2015.10.068.
[16] Bagehorn S, Wehr J, Maier HJ. Application of mechanical surface finishing
processes for roughness reduction and fatigue improvement of additively
manufactured Ti-6Al-4V parts. Int J Fatigue 2017;102:135–42. https://doi.org/
10.1016/j.ijfatigue.2017.05.008.
[17] Ren XP, Li HQ, Guo H, Shen FL, Qin CX, Zhao ET, et al. A comparative study on
mechanical properties of Ti-6Al-4V alloy by additive manufacturing vs. traditional
processing. Mat Sci Eng A 817 2021:141384. https://doi.org/10.1016/j.
msea.2021.141384.
[18] Zhang LC, Wang J, Liu Y, Jia Z, Liang S-X. Additive manufacturing of titanium
alloys. Encyclopedia Mater.: Met. Alloys 2022;1:256–74. https://doi.org/10.1016/
B978-0-12-819726-4.00002-8.
[19] Edwards P, Ramulu M. Fatigue performance evaluation of selective laser meltedTi-
6Al-4V. Mat Sci Eng A 2014;598:327–37. https://doi.org/10.1016/j.
msea.2014.01.041.
[20] Gong H, Rafi K, Gu H, Janaki Ram GD, Starr T, Stucker B. Influence of defects on
mechanical properties of Ti-6Al-4V components produced by selective laser
melting and electron beam melting. Mat Des 2015;86:545–54. https://doi.org/
10.1016/j.matdes.2015.07.147.
[21] Li P, Warner DH, Fatemi A, Phan N. Critical assessment of the fatigue performance
of additively manufactured Ti-6Al-4V and perspective for future research. Int J
Fatigue 2016;85:130–43. https://doi.org/10.1016/j.ijfatigue.2015.12.003.
[22] Cao F, Zhang T, Ryder MA, Lados DA. A review of the fatigue properties of
additively manufactured Ti-6Al-4V. JOM 2018;70:349–57. https://doi.org/
10.1007/s11837-017-2728-5.
[23] Du L, Qian G, Zheng L, Hong Y. Influence of processing parameters of selective
laser melting on high-cycle and very-high-cycle fatigue behaviour of Ti-6Al-4V.
Fatigue Fract Eng Mater Struct 2021;44:240–56. https://doi.org/10.1016/j.
msea.2017.03.001.
[24] da Costa PR, Sardinha M, Reis L, Freitas M, Fonte M. Ultrasonic fatigue testing in
as-built and polished Ti6Al4V alloy manufactured by SLM. Forces in Mechanics
2021;4:100024. https://doi.org/10.1016/j.finmec.2021.100024.
[25] Xu Z, Liu A, Wang X, Liu B, Guo M. Fatigue limit prediction model and fatigue
crack growth mechanism for selective laser melting Ti6Al4V samples with inherent
defects. Int J Fatigue 2021;143:106008. https://doi.org/10.1016/j.
ijfatigue.2020.106008.
[26] Wycisk E, Siddique S, Herzog D, Walther F, Emmelmann C. Fatigue performance of
laser additive manufactured Ti-6Al-4V in very high cycle fatigue regime up to 109
cycles. Front Mater 2015;2:72. https://doi.org/10.3389/fmats.2015.00072.
8
International Journal of Fatigue 164 (2022) 107129
[27] Greitemeier D, Palm F, Syassen F, Melz T. Fatigue performance of additive
manufactured TiAl6V4 using electron and laser beam melting. Int J Fatigue 2017;
94:211–7. https://doi.org/10.1016/j.ijfatigue.2016.05.001.
[28] Gunther J, Krewerth D, Lippmann T, Leuders S, Troster T, Weidner A, et al. Fatigue
life of additively manufactured Ti-6Al-4V in the very high cycle fatigue regime. Int
J Fatigue 2017;94:236–45. https://doi.org/10.1016/j.ijfatigue.2016.05.018.
[29] Sun C, Chi W, Wang W, Duan Y. Characteristic and mechanism of crack initiation
and early growth of an additively manufactured Ti-6Al-4V in very high cycle
fatigue regime. Int J Mech Sci 2021;205:106591. https://doi.org/10.1016/j.
ijmecsci.2021.106591.
[30] Kahlin M, Ansell H, Basu D, Kerwin A, Newton L, Smith B, et al. Improved fatigue
strength of additively manufactured Ti6Al4V by surface post processing. Int J
Fatigue 2020;134:105497. https://doi.org/10.1016/j.ijfatigue.2020.105497.
[31] Liu X, Sun C, Hong Y. Effects of stress ratio on high-cycle and very-high-cycle
fatigue behavior of a Ti-6Al-4V alloy. Mat Sci Eng A 2015;622:228–35. https://doi.
org/10.1016/j.msea.2014.09.115.
[32] Neal DF, Blenkinsop PA. Internal fatigue origins in α-β titanium alloys. Acta Metall
1976;24:59–63. https://doi.org/10.1016/0001-6160(76)90147-4.
[33] Hattal A, Chauveau T, Djemai M, Fouchet JJ, Bacroix B, Dirras G. Data related to
spectrum analyzes for phases identification, microstructure and mechanical
properties of additive manufactured Ti6Al4V reinforce with nano yttria stabilized
zirconia. Data in Brief 2020;29:105249. https://doi.org/10.1016/j.
dib.2020.105249.
[34] Kikuchi S, Hayami Y, Ishiguri T, Guennec B, Ueno A, Ota M, et al. Effect of bimodal
grain size distribution on fatigue properties of Ti-6Al-4V alloy with harmonic
structure under four-point bending. Mat Sci Eng A 2017;687:269–75. https://doi.
org/10.1016/j.msea.2017.01.076.
[35] Guennec B, Ishiguri T, Kawabata MO, Kikuchi S, Ueno A, Ameyama K.
Investigation on the durability of Ti-6Al-4V alloy designed in a harmonic structure
via powder metallurgy: fatigue behavior and specimen size parameter issue. Metals
2020;10:636. https://doi.org/10.3390/met10050636.
[36] Horikawa N, Guennec B, Kashi T, Miyajima T, Kinoshita T, Okamura S, et al.
Industrial AIP and UBMS coating deposition: static and fatigue performances of
TiAlN coated SKH51 tool steel. J Mater Eng Perform 2022. https://doi.org/
10.1007/s11665-022-06917-y.
[37] Murakami Y, Endo M. Effects of defects, inclusions and inhomogeneities on fatigue
strength. Int J Fatigue 1994;16:163–82. https://doi.org/10.1016/0142-1123(94)
90001-9.
[38] Mergia K, Stefanopoulos KL, Ordás N, García-Rosales C. A comparative study of the
porosity of doped graphites by small angle neutron scattering, nitrogen adsorption
and helium pycnometry. Microporous Mesoporous Mater 2010;134:141–9. https://
doi.org/10.1016/j.micromeso.2010.05.019.
[39] Gubicza J. X-ray line profile analysis in Materials Science, IGI-Global, Hershey, PA,
USA.
[40] Ribarik G, Gubicza J, Ungar T. Correlation between strength and microstructure of
ball-milled Al-Mg alloys determined by X-ray diffraction. Mat Sci Eng A 2004;
387–389:343–7. https://doi.org/10.1016/j.msea.2004.01.089.
[41] Vrancken B, Thijs L, Kruth J-P, Van Humbeeck J. Heat treatment of Ti6Al4V
produced by Selective Laser Melting: Microstructure and mechanical properties.
J Alloy Compd 2012;541:177–85. https://doi.org/10.1016/j.jallcom.2012.07.022.
[42] JSMS Committee on Fatigue of Materials. JSMS Committee of Reliability
Engineering. Standard Evaluation Method of Fatigue Reliability for Metallic
Materials – Standard Regression Method of S-N Curves. JSMS-SD-6-08. Soc. Mater.
Jpn 2008.
[43] Kurek A, Lagoda T, Kurek M. Stress gradient as a size effect in fatigue life
determination for alternating bending. Int J Fatigue 2021;153:106461. https://doi.
org/10.1016/j.ijfatigue.2021.106461.
[44] Aguado-Montero S, Navarro C, Vazquez J, Lasagni F, Slawik S, Domingez J. Fatigue
behavior of PBF additive manufactured TI6AL4V alloy shot and laser peening. Int J
Fatigue 2023;154:106536. https://doi.org/10.1016/j.ijfatigue.2021.106536.
[45] Kahlin M, Ansell H, Moverare JJ. Fatigue behavior of notched additive
manufactured Ti6Al4V with as-built surface. Int J Fatigue 2017;101:51–60.
https://doi.org/10.1016/j.ijfatigue.2017.04.009.
[46] Brandl E, Heckenberger U, Holzinger V, Buchbinder D. Additive manufactured
AlSi10Mg samples using Selective Laser Melting (SLM): Microstructure, high cycle
fatigue and fracture behavior. Mat Des 2012;34:159–69. https://doi.org/10.1016/
j.matdes.2011.07.067.
[47] Schneller W, Leitner M, Leuders S, Sprauel JM, Grun F, Pfeifer T, et al. Fatigue
strength estimation methodology of additively manufactured metallic bulk
material. Addit Manuf 2021;39:101688. https://doi.org/10.1016/j.
addma.2020.101688.
[48] Hu YN, Wu SC, Wu ZK, Zhong XL, Ahmed S, Karabal S, et al. A new approach to
correlate the defect population with the fatigue life of selective laser melted Ti-6Al-
4V. Int J Fatigue 2020;136:105584. https://doi.org/10.1016/j.
ijfatigue.2020.105584.
[49] Joseph S, Lindley TC, Dye D. Dislocation interactions and crack nucleation in a
fatigued near-alpha titanium alloy. Int J Plast 2018;110:38–56. https://doi.org/
10.1016/j.ijplas.2018.06.009.
[50] Nalla RK, Boyce BL, Campbell JP, Peters JO, Ritchie RO. Influence of
microstructure on high-cycle fatigue of Ti-6Al-4V: Bimodal vs. Lamellar structure.
Metall Mater Trans A 2002;33:899–918. https://doi.org/10.1007/s11661-002-
0160-z.
[51] Ivanova SG, Biederman RR, Sisson Jr RD. Investigation of fatigue crack initiation in
Ti-6Al-4V during tensile-tensile fatigue. J Mater Eng Perform 2002;11:226–31.
https://doi.org/10.1361/105994902770344312.
B. Guennec et al.
[52] Crupi V, Epasto G, Guglielmino E, Squillace A. Influence of microstructure [alpha
+ beta and beta] on very high cycle fatigue behavior of Ti-6Al-4V. Int J Fatigue
2017;95:64–75. https://doi.org/10.1016/j.ijfatigue.2016.10.002.
[53] Lin M-R, Fine ME, Mura T. Fatigue crack initiation on slip bands: theory and
experiment. Acta Metall 1986;34:619–28. https://doi.org/10.1016/0001-6160
(86)90177-X.
International Journal of Fatigue 164 (2022) 107129
[54] Kishida K, Kim JG, Nagae T, Inui H. Experimental evaluation of critical resolved
shear stress for the first-order pyramidal c+a slip in commercially pure Ti by
micropillar compression method. Acta Mater 2020;196:168–74. https://doi.org/
10.1016/j.actamat.2020.06.043.