Fatigue of Ti6Al4V manufactured by PBF-LB: A comparison of failure mechanisms between net-shape and electro-chemically milled surface conditions

Full text

Contents lists available at ScienceDirect
Engineering Failure Analysis
journal homepage: www.elsevier.com/locate/engfailanal
Fatigue of Ti6Al4V manufactured by PBF-LB: A comparison of
failure mechanisms between net-shape and electro-chemically
milled surface conditions
Tatiana Risposi a,1, Lorenzo Rusnati a,1, Luca Patriarca a, Alex Hardaker c,
Dawid Luczyniec d, Stefano Beretta a,b,∗
aPolitecnico di Milano, Department of Mechanical Engineering, via La Masa 1, 20156 Milano, Italy
bAuburn University, National Center for Additive Manufacturing Excellence (NCAME), Auburn, AL 36849, USA
cMTC–National Centre for AM, Coventry CV7 9JU, UK
dEuropean Space Agency ESTEC, Keplerlaan 1, 2200 AG Noordwijk-ZH, The Netherlands
A R T I C L E I N F O
Keywords:
Laser-powder bed fusion
Net-shape
Hirtisation
Fatigue
Contouring defects
A B S T R A C T
In the recent years, metal additive manufacturing (AM) has acquired large interest for many
industrial applications, principally due to the capability to produce parts with complex geom-
etry. The critical aspect of AM parts is the sensitivity to surface anomalies due to net-shape
surfaces, i.e surface microcracks and protrusions, localized stresses caused by coarse surface
roughness, or sub-surface features placed below the outer skin in the contour region. To reduce
the surface roughness and increase the fatigue properties, proper post-process treatments can
be applied. This work investigates the improvement in surface quality and fatigue properties
due to the electro-chemical milling process of Hirtisation®compared with net-shape condition,
on samples manufactured in Ti6Al4V by laser-powder bed fusion (PBF-LB). Post-processing led
to a reduction of surface roughness due to the removal of the peaks and sharp valleys that
act as crack initiation sites during fatigue tests, but it exposed the sub-skin contouring defects
to the free surface. These were the crack initiation sites resulting in a limited improvement of
the potential benefits produced by Hirtisation®. This was confirmed by fatigue life predictions
based on propagation of surface features and contouring anomalies.
1. Introduction
Additive manufacturing (AM) offers several significant advantages, including the ability to produce parts with complex
geometries with minimal post-processing, reduced material waste, and the flexibility to work with both metals and plastics [1].
This study focuses on the laser-powder bed fusion (PBF-LB) process, specifically using the Ti6Al4V alloy, which is widely utilized
in various industries such as aerospace [2,3], chemical, marine, automotive, and medical sectors [4–6]. The large applicability of
Ti6Al4V is attributed to its high corrosion resistance, excellent bio-compatibility, high strength, and a good balance of mechanical
properties relative to its density, in addition to its high weldability.
Although recent literature has demonstrated that the static properties of Ti6Al4V alloy produced by PBF-LB process are
comparable to those of wrought titanium alloys [7–11], the fatigue performances are generally reduced because the PBF-LB process
∗Corresponding author at: Politecnico di Milano, Department of Mechanical Engineering, via La Masa 1, 20156 Milano, Italy.
E-mail address: stefano.beretta@polimi.it (S. Beretta).
1These authors share the first authorship as they have contributed equally to this work.
https://doi.org/10.1016/j.engfailanal.2025.109403
Received 5 December 2024; Received in revised form 6 February 2025; Accepted 7 February 2025
Engineering Failure Analysis 172 (2025) 109403
Available online 14 February 2025
1350-6307/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

T. Risposi et al.
Nomenclature
Acronyms
AM Additive manufacturing
CPCA Compression pre-cracking constant amplitude
CPLR Compression pre-cracking load reduction
EBSD Electron back-scatter diffraction
EDM Electrical discharge machining
EVS Extreme values statistics
FCG Fatigue crack growth
IPF Inverse pole figure
LEVD Largest extreme values distribution
OPS Oxide polishing suspension
PBF-EB Electron beam powder bed fusion
PBF-LB Laser-powder bed fusion
SEB Single edge-notched bending
SEM Scanning electron microscope
Symbols
𝛿Scale parameter of the LEVD
𝛥𝐾𝑡ℎ,𝐿𝐶 Long crack threshold stress intensity factor range
𝛥𝐾𝑡ℎ Threshold stress intensity factor range
𝛥𝜎𝑤,0Fatigue limit stress range in absence of defects
𝛥𝜎𝑤Fatigue limit stress range
𝜀𝑓Elongation at fracture
𝜆Position parameter of the LEVD
𝜆𝑐Cut-off wavelength
𝜎Stress
𝜎log 𝑁Standard deviation of the logarithmic life
𝜎𝑤Fatigue limit stress amplitude
area Murakami’s defect size
area50% Average defect size
area0El-Haddad parameter
𝐴Fit parameter of the S–N curve
𝑎Crack depth
𝐵Slope of the S–N curve
𝐶Fit parameter of the NASGRO equation
𝑐Crack width
E Young’s modulus
𝑓Newman’s crack opening function
𝐹LEVD Cumulative density function of the LEVD
𝐾𝐼 𝑐Fracture toughness
𝑚Fit parameter of the NASGRO equation
𝑁Fatigue cycles
𝑁𝑙 𝑖𝑚 Fatigue cycles at the knee point
𝑝Fit parameter of the NASGRO equation
𝑞Fit parameter of the NASGRO equation
R Load ratio
𝑆𝑎Aerial arithmetic roughness
𝑆𝑝Maximum peak height
𝑆𝑝,𝑚 Mean profile peak height
𝑆𝑞Quadratic mean of profile height
𝑆𝑣Maximum valley depth
Engineering Failure Analysis 172 (2025) 109403
2

T. Risposi et al.
𝑆𝑧Maximum peak to valley height
𝑡𝑚𝑎𝑥 Maximum crack depth
UTS Ultimate tensile strength
𝑌Shape factor
YS Yield strength
Table 1
Chemical analysis of Ti6Al4V powder, expressed as Wt.%.
Al V Ti O N H Fe C Y Others
6.29 4.04 bal. 0.09 0.01 0.001 0.04 0.01 <0.001 <0.40
produces inherently rough surfaces from which cracks can nucleate [12–14]. On PBF-LB surfaces, features like ridges and troughs
with steep walls can be distinguished [15,16]. These surface features are particularly detrimental for structural components as they
create high stress concentration areas [17–19]. Additionally, sub-surface defects, such as pores and lack of fusion, can also contribute
to crack nucleation [14,20–22]. The occurrence of these defects depends on process parameters like laser power, scanning speed,
hatch spacing, and layer thickness [19,23,24]. Another important factor affecting the presence of surface and sub-surface defects is
the scanning strategy. Typically, contour scanning strategies are used to improve surface quality [25–27], although multiple defects
often appear in the sub-surface, contour region [28]. The location of these defects depends on the contour depth and parameters,
and they can have sharper shapes than defects found in the hatch [29].
Recent studies have shown that the fatigue properties of net-shape PBF-LB materials can be enhanced through appropriate
post-processing treatments that reduce surface roughness. This is particularly important because fatigue cracks often initiate at
multiple locations on the surface’s valleys [18,30–32]. One method for reducing surface roughness is chemical milling, a process
that removes layers of material using liquid chemical compounds. By utilizing chemical-based surface treatments, there are no
geometric constraints on the areas that can be treated, making them particularly effective for hard-to-access locations that are
challenging for traditional machining tools [31,33]. As a result, chemical treatments have received special attention, and their
effects on AM parts were explored in several studies [34,35]. However, certain structures, such as strut-based lattices, exhibit
limitations in chemical etching techniques due to restricted fluid flow, leaving unmelted particles trapped deep within. These
particles act as stress concentration points and are detrimental for the fatigue performances of the manufactured parts [36]. To
overcome some of the limitations of the traditional chemical etching techniques, a new chemically-based post-process treatment,
called Hirtisation®, was developed and it is currently supplied by RENA Technologies GmbH (Austria). Hirtisation®is a process that
combines electro-chemical pulse methods, flow and particle-assisted chemical removal, and chemical surface treatment, thereby
integrating the advantages from each process route. Published works have investigated the effect of Hirtisation®on AlSi10Mg
samples manufactured by the PBF-LB process [31,37,38], as well as on Ti6Al4V manufactured by the electron beam powder bed
fusion (PBF-EB) process [29,39–41]. However, to the authors’ knowledge, there is limited information available regarding the effect
of the Hirtisation®process on the fatigue properties of AMed materials, particularly Ti6Al4V, which is one of the most widely used
materials in aerospace applications.
This study examines the surface quality of net-shape Ti6Al4V produced by the PBF-LB process before and after undergoing the
Hirtisation®post-process. In particular, it focuses on the impact of residual porosities and surface roughness on fatigue performances.
A comprehensive testing campaign was designed to evaluate the fatigue and fracture properties of Ti6Al4V manufactured by the
PBF-LB process, comparing fatigue specimens differing by their surface condition.
2. Materials and experiments
2.1. Specimens manufacturing
The test campaign presented in this study involved tests conducted on specimens oriented with the load direction parallel to the
building direction. The specimens’ geometries are reported in Fig. 1.
All the Ti6Al4V specimens were manufactured by the research organization Manufacturing Technology Centre (MTC, Coventry,
UK) on a Renishaw 500Q PBF-LB machine equipped with 4×500 W Prism Ytterbium fibre pulsed lasers with dynamic focusing.
The oxygen content was controlled and measured by sensors in the Renishaw AM500Q, a maximum O2limit of 1000 ppm is set on
the machine, however the build process is performed in a lower O2atmosphere, with a measured content around 100 ppm through
the build. Powder composition is reported in Table 1. Particle size distribution was estimated through laser size diffraction and
showed a median dimension equal to 31.2 μm, with the 10% and 90% percentiles equal to 21.4 μm and 45.1 μm respectively. The
scan strategy used two contour scans, offset from each other by 50 μm and with the outer contour offset from the nominal part
geometry by 100 μm,Fig. 1(c). The contour was completed before the bulk hatching in a spot melting mode, which translates to
very rapid exposure of beam as an array of individual spots.
Engineering Failure Analysis 172 (2025) 109403
3

T. Risposi et al.
Table 2
Build parameters for Ti6Al4V specimens.
Region Parameter Value
Layer thickness [μm] 60
Scan strategy Stripe
Power [W] 180
Bulk Point distance [μm] 65
Exposure time [μs] 110
Hatch distance [μm] 95
Contour offset [μm] 50
Power [W] 320
Contour Point distance [μm] 30
Exposure time [μs] 20
Hatch distance [μm] 60
Table 3
Tensile properties of PBF-LB/Ti6Al4V alloy in net-shape condition.
UTS [MPa] YS [MPa] E [MPa] 𝜀𝑓[%]
1187 1146 114,469 12
The build was conducted in an inert build chamber evacuated and filled with Argon shielding gas. During manufacturing, the
build plate was maintained at 170 ◦C. A summary of the MTC’s standard process parameters is reported in Table 2. The specimens
were then stress-relieved through a heat treatment consisting of heating to 650 ◦C at atmospheric pressure in Argon for a soaking
time of 2 h. This was followed by a controlled cooling phase inside the furnace, completed in less than 5 h.
The cylindrical specimens for axial fatigue (Fig. 1(a)) were characterized by a diameter of the cross section of 6 mm, in
compliance with ASTM E466 [42]. The specimens were successively machined in the grip region, while the cross-section was left
in the net-shape condition.
Tensile properties were evaluated on two net-shape specimens tested under a MTS RT100 system with a load capacity of 100
kN. The results are listed in Table 3.
The single edge-notched bending (SEB) specimens (Fig. 1(b)) were manufactured with the crack plane parallel to the building
plate. The external surfaces were milled to guarantee the tolerances required. The notches were obtained by means of electrical
discharge machining (EDM) which guaranteed a notch tip with a radius of 150 μm, free of alpha case formation.
2.2. Hirtisation®
The fatigue specimens were split into two groups: net-shape condition and treated with Hirtisation®. Hirtisation®was conducted
by RENA Technologies GmbH. The treatment comprises two consecutive steps: (i) softening of support structures and powder
adhesions in Ti-Auxilex electrolyte, (ii) detachment of softened support structure, detachment of powder adhesions and polishing in
Ti-Delevatex electrolyte. Between each treatment step, the parts were rinsed in a de-ionized water basin and were dried in a vacuum
drying unit. Prior to the manufacturing of the specimens, different trials were conducted on coupons to tune the electro-chemical
milling parameters to achieve a removal up to 200 μm from the treated surface. Nonetheless, the post-process, once applied to the
fatigue specimens, removed material for an average of 66 μm in radial direction. This limited material removal is the cause of the
peculiar failure mechanism described in the following sections.
2.3. Microstructure
The near surface microstructure was evaluated on cross sections taken from the gauge length of two axial specimens, one in net-
shape condition and one treated with Hirtisation®. The samples were cut in a direction parallel to the load axis, were mounted into
an epoxy resin and polished starting with sand-paper with granularity of 220, followed by a9 μm diamond paper and concluding
with a Colloidal Silica oxide polishing suspension (OPS). An electron back-scatter diffraction (EBSD) analysis was performed by
scanning electron microscopy (SEM) in order to investigate the microstructure.
Porosity measurements were conducted on witness samples printed with the specimens. Four samples were located in the corners
of each build. The witness specimens were prepared by cutting along the vertical plate and polishing the samples to enable analysis
over the full height of the cubes. Image fractions were then taken from the images of the sample sections to give an indication
of porosity levels in the builds, in accordance with ASTM F3637 standard [43]. Porosity levels measured on the witness coupons
portrayed a stable density across the builds. The average relative density was 99.97% with a standard deviation of 0.02%; the
maximum density was recorded at 99.99%, whereas the minimum value was 99.89%.
Engineering Failure Analysis 172 (2025) 109403
4

T. Risposi et al.
Fig. 1. (a) Axial specimen for high cycle fatigue tests; (b) Single edge-notched bending specimen for fatigue crack growth tests; (c) Printing strategy represented
on the cross section of the axial specimens.
2.4. Roughness measurement
Roughness measurements on net-shape surface and on surface after Hirtisation®were performed with Alicona InfiniteFocus FV
(Focus-Variation). One specimen for each surface condition was analysed. The measurements were done along the cylinders’ axial
direction, on 4 areas of 1 mm ×15 mm each; each region was distanced by a 90 degrees rotation. These acquisitions were carried
out at a vertical resolution of 60 μm and lateral resolution of 2 μm. The primary profiles extracted from Alicona were filtered with
a Robust Gaussian filter with a cut-off wavelength 𝜆𝑐=3 mm.
Engineering Failure Analysis 172 (2025) 109403
5

T. Risposi et al.
2.5. Fatigue tests
The fatigue tests were performed at a stress ratio R= −1, in air, at room temperature, on the MTS 810 servo-hydraulic machine
with a load capacity of 100 kN. The run-out condition was set at 107cycles, although a single test was marked as run-out at
5 × 106cycles. The testing frequency was chosen within the range of 20 ÷40 Hz to balance the accuracy of the load control by the
servo-hydraulic equipment and the testing time. The load level was calculated for each specimen, based on the average diameter
measured on 3 sections at different heights along the gauge length. All fracture surfaces were analysed by SEM, using the Zeiss
EVO50 system, to identify the killer defect at the origin of the fatigue failure.
2.6. Fatigue crack growth tests
SEB specimens were initially pre-cracked under the MTS 810 servo-hydraulic system with load capacity of 100 kN. The selected
pre-cracking procedure consisted in the application of compressive cyclic loadings with the aim of nucleating an initial closure-free
crack at the EDM notch [44]. In particular, the first compressive load cycle generates yielding in the region of the notch, thus leading
to residual tensile stresses that, subjected to the subsequent load cycles, nucleate a crack. Successively, the residual stresses decrease
with the crack advancement until the crack driving force is below the threshold and the crack arrests [45]. Specimens were loaded
in fatigue with a maximum compressive load of 40 kN for 2× 106cycles to develop an average pre-crack size of 47 μm.
After pre-cracking, crack growth rate tests were conducted for four load ratios: R=0.7, R=0, R= −1 and R= −2. Two types
of tests were performed: (i) determination of the Paris region via Compression Pre-cracking-Constant Amplitude (CPCA) procedure,
(ii) measure of the crack growth threshold, 𝛥𝐾𝑡ℎ,𝐿𝐶 , via Compression Pre-cracking-Load Reduction method (CPLR). During the tests,
a clip-on-gauge was utilized to monitor the crack length through the compliance method. At the end of the fatigue crack growth
tests, the specimens were statically broken and the fracture surfaces were investigated by means of a stereo microscope; the real
final crack lengths were measured and the acquired experimental data were corrected accordingly.
3. Results
3.1. Micrographs of surfaces and microstructural characterization
Micrographs of polished longitudinal sections near the sample surfaces were examined using SEM, both in net-shape condition
and after Hirtisation®, as shown in Fig. 2. The difference in surface texture between the two conditions is clearly visible, with the
post-Hirtisation®surface appearing smoother and more weavy. In contrast, the net-shape surface exhibits deep and sharp valleys
that act as crack initiation sites during fatigue tests. Sub-surface features in the contour region can also be observed for both surface
conditions, as indicated by the yellow dashed circles in Fig. 2.
The use of Hirtisation®led to an average reduction of the diameter of the specimens’ gauge region of 132 μm. The removal
exceeded the ten-point height parameter 𝑆10𝑧of the surface roughness evaluated on the net-shape specimens (additional details are
reported in Section 3.2). As a result, all surface features from the net-shape surfaces were removed. However, this material removal
brought sub-surface contouring anomalies closer to the surface. For example, Fig. 2shows a sub-surface defect located at a depth
of 52 μm in a surface-treated specimen, compared to 139 μm in the net-shape condition. These defects, being closer to the surface,
may become critical and act as killer defects.
Fig. 2also presents inverse pole figure (IPF)-z maps of the regions surrounding the sub-surface pores. Consistent with previous
studies [8,46,47], the microstructure of Ti6Al4V shows very fine 𝛼′martensite with columnar 𝛽grains, aligned opposite to the
cooling direction due to the high cooling rate of the PBF-LB process. This microstructure is responsible for the high strength and
low ductility of laser-based AM Ti6Al4V alloy [48]. In both conditions, the microstructures near the surface are very similar, with
no observable microstructural alterations after the Hirtisation®post-process.
3.2. Roughness measurement
The topography of the surfaces of the fatigue specimens’ measured surfaces is presented in Fig. 3, alongside detailed SEM
images of the surfaces in the two conditions. As expected, the net-shape specimens exhibit higher surface roughness and a
significant presence of partially molten particles, see Fig. 3(a). Roughness data from net-shape fatigue specimens are in line with
the expectations as the values are comparable with legacy data by The MTC, acquired on previous prints on Renishaw AM500Q,
showing an arithmetic roughness 𝑅𝑎= 11.94 μm. After Hirtisation®, the high peaks are generally removed. However, as shown in
Fig. 3(b), slight waviness and ridges with scattered craters in between can still be observed in accordance with [31].
Areal surface roughness values for the two surface conditions, by averaging four measurements performed on each sample, are
listed in Table 4, along with the percentage reduction in roughness parameters after Hirtisation®. The roughness profiles obtained by
the four areal measurements performed on one specimen for each surface condition are reported in Fig. 4(a) for net-shape condition,
and Fig. 4(b) after Hirtisation®. The graphs evidence that the post-process changed the surface texture, since all the roughness
parameters are decreased in value. The maximum peak height, 𝑆𝑝, is the parameter most affected by Hirtisation®post-process with
a reduction of 57%. This is also visible from the height maps of the surfaces before and after the post-process, reported in Fig. 3.
Engineering Failure Analysis 172 (2025) 109403
6

T. Risposi et al.
Fig. 2. Micrographs of the sample surfaces and related IPF-z maps: (a) net-shape condition shows sharp valleys indicated with white arrows and sub-surface
defects highlighted in yellow, (b) surface after Hirtisation®post-processor shows a waviness surface with sub-surface defects closer to the surface.
Table 4
Areal roughness parameters from Alicona acquisitions, for Ti6Al4V specimens in two surface conditions.
Surface condition 𝑆𝑎[μm] 𝑆𝑞[μm] 𝑆𝑝[μm] 𝑆𝑣[μm] 𝑆𝑧[μm]
Net-shape 11.6 15.0 88.9 73.0 162.0
Hirtisation®6.8 9.0 37.9 49.1 87.0
Reduction 41% 40% 57% 32% 46%
3.3. Fatigue tests
The fatigue results are presented as S–N diagrams for both net-shape and after Hirtisation®conditions in Fig. 5. The regression
line, representing 50% failure probability, was obtained by fitting the experimental data in the finite region with Eq. (1), in
accordance with ASTM E739 [49]:
𝑁=𝐴𝜎𝐵(1)
The fitting parameters, determined using the maximum likelihood method, are listed in Table 5. Scatter in the data was accounted
for by fitting the experimental data with a log-normal distribution, assuming a constant standard deviation 𝜎𝑙 𝑜𝑔 𝑁.Table 5also reports
the experimental fatigue limit range, 𝛥𝜎𝑤, and the number of cycles to failure 𝑁𝑙 𝑖𝑚 at the knee-point of the S–N curves. The fatigue
limits were calculated applying the Dixon up and down method [50].
A comparison between the S–N curves for net-shape condition and after Hirtisation®is presented in Fig. 5, in which the
experimental points, the log-normal fits and the 95% scatter bands are reported. It is possible to observe how the specimen subjected
to Hirtisation®have enhanced fatigue performance compared to the net-shape condition. In the finite life region, fatigue life is
Engineering Failure Analysis 172 (2025) 109403
7

T. Risposi et al.
Fig. 3. Elaboration of the surfaces acquired by focus variation microscopy and SEM images of cylindrical specimens’ surfaces: (a) net-shape surface, (b) surface
after Hirtisation®.
Fig. 4. Roughness profiles extracted from the areal measurements: (a) surface in net-shape condition, (b) surface after Hirtisation®.
Engineering Failure Analysis 172 (2025) 109403
8

T. Risposi et al.
Table 5
Parameters of the S–N curves for the two series of Ti6Al4V specimens.
Surface condition 𝜎𝑙 𝑜𝑔 𝑁𝑙 𝑜𝑔10(𝐴)B𝑁𝑙 𝑖𝑚 [cycles] 𝛥𝜎𝑤[MPa]
Net-shape 0.1394 5.1384 −4.53 629,265 420
Post-Hirtisation®0.1043 5.1672 −4.71 580,369 480
Fig. 5. S–N curves of Ti6Al4V alloy from fatigue tests conducted at R= −1.
significantly increased by a factor ≈1.6, with the S–N diagrams having a similar slope. However, the increase of the fatigue limit
is less pronounced: the Hirtisation®process resulted in a fatigue limit of 𝛥𝜎𝑤=480 MPa, compared to 𝛥𝜎𝑤=420 MPa for the
net-shape condition.
It is important to note that removing the surface layer has two primary effects, particularly effective in the fatigue limit region.
First, it reduces surface roughness, which positively impacts fatigue life. Simultaneously, the Hirtisation®process reduces the average
distance between sub-surface defects and the surface, increasing the criticality of these defects. This second effect partially offsets
the positive impact of surface roughness reduction. A detailed discussion of the different types of killer defects identified through
fracture surface analysis is provided in Section 4.
3.4. Fatigue crack growth tests
The results of the experimental crack growth tests are presented in Fig. 6(a) for the four load ratios investigated in the present
study (i.e. R=0.7, 0, −1 and −2), while Fig. 6(b) presents the experimental 𝛥𝐾𝑡ℎ,𝐿𝐶 as a function of the load ratio R. These quantities
are also reported in Table 6. The procedure adopted to fit the 𝛥𝐾𝑡ℎ,𝐿𝐶 values, the solid red line in Fig. 6(a), is the same followed by
Beretta et al. [51] and by Barricelli et al. [52]. The crack growth properties of the present Ti6Al4V alloy are coherent with values
from Polimi database on the same material, the same printing direction and with heat treatments at temperature that guarantee a
similar microstructure [52,53], as well as with literature data [54,55].
The solid lines in Fig. 6(a) represent the fitting of experimental crack growth data with the NASGRO equation [56]:
𝑑 𝑎
𝑑 𝑁=𝐶1 −𝑓
1 −𝑅𝛥𝐾 𝑚(1 −𝛥𝐾𝑡ℎ
𝛥𝐾 )𝑝
(1 −𝐾𝑚𝑎𝑥
𝐾𝑐
)𝑞
(2)
in which the fitting parameters 𝐶,𝑚and 𝑝were determined according to the available experimental data, while 𝑞was set equal
to zero since it describes the zone of the Paris region in which there is unstable crack propagation (𝐾𝑚𝑎𝑥 ⟶𝐾𝑐) which was not
covered by the experimental tests. Similar results in the Paris region were obtained by Cain et al. [57].
4. Analysis of critical defects
4.1. Measurement of surface features
The fracture mechanics approach proposed by Murakami [58] was adopted for the analysis of the fatigue limit’s dependence
on defect size. The fracture surfaces of fatigue samples were analysed by means of a SEM, revealing two crack initiation features:
Engineering Failure Analysis 172 (2025) 109403
9

T. Risposi et al.
Fig. 6. Fatigue crack growth tests on Ti6Al4V alloy: (a) NASGRO fit obtained from experimental crack growth data, (b) 𝛥𝐾𝑡ℎ,𝐿𝐶 versus load ratio R variation
as experimental data and description by the NASGRO model.
Table 6
Experimentally determined and results of the NASGRO fit values of 𝛥𝐾𝑡ℎ,𝐿𝐶 , expressed in
[MPa𝑚], for the tested stress ratios.
R Experimental value NASGRO fit
−2 7.17 7.24
-1 4.62 4.82
5.01
0 2.58 2.58
0.7 1.79 1.82
1.84
valleys in the net-shape surface, and sub-surface defects in the contour region for the samples subjected to Hirtisation®. In net-shape
specimens, Figs. 7(a) and 7(b), crack initiation was predominantly linked to deep, sharp features. These features consisted of partially
molten powder particles attached to fully molten material, leading to high stress concentration zones. Furthermore, fractographic
analyses revealed defects in the contour regions, marked in purple.
In contrast, fractographies of samples subjected to Hirtisation®revealed three distinct types of crack initiation zones: (i) crack
initiating at the bottom of deep valleys (Fig. 8(a)), similar to those observed in net-shape samples but with wavy character; (ii) defects
connected to the surface affected by corrosion at the top, suggesting exposure during the Hirtisation®process, with the bottom
Engineering Failure Analysis 172 (2025) 109403
10

T. Risposi et al.
Fig. 7. Fractographies of net-shape specimens: (a) and (b) crack initiation from deep and sharp features typical of net-shape topography with detail on internal
defect in contour region.
remaining unaffected, likely due to limited electrolyte penetration (Fig. 8(c)); and (iii) cracks originating from surface-adjacent
defects without signs of internal corrosion (Fig. 8(b)).
For cracks initiated from troughs (type i), both in the net-shape condition and after Hirtisation®, considering the crack depth as
the whole height of what is detectable on the fracture plane leads to an overestimation of the defect size. Based on the approach
proposed by Barricelli et al. [52], a reference line was positioned using a surface roughness-based method to accurately calculate
the defect size. The method includes the following steps:
•Identify the external features (i.e. un-molten particles attached to the surface which affect the roughness measurements but
do not contribute the defect depth) in the fractography images;
•Starting from the highest feature, move downwards by a distance equal to the mean peak height value 𝑆𝑝,𝑚 obtained from
roughness measurements for the specific surface condition;
•Define the defect size at the fracture origin as the area of the region beneath the reference line.
The Murakami’s parameter (area)was then measured with a digital image analysis software [59]. The following rules were
applied for determining the Murakami’s parameter:
•Defect characterized with an aspect ratio (width over depth) 2c/a <10, area was calculated as the square root of the
measured area at the fracture origin;
•Defect characterized by an aspect ratio 2c/a ≥10, area was calculated as 10 ⋅𝑡𝑚𝑎𝑥, where 𝑡𝑚𝑎𝑥 is the maximum depth
measured from the reference line to the defect contour.
4.2. Failure mechanisms
Both micrographs and fracture surfaces of net-shape samples reveal defects located in the contour region of the material that
were highlighted in purple on the fractographies, see for example Figs. 7(a) and 7(b). These defects can be attributed to voids
generated at the neighbouring spots as well as the contouring-hatching interface [60]. Such defects did not lead to crack initiation
and then failure when the specimens were in net-shape condition. The case of a fatigue specimen provides a notable exemplification
of the phenomenon. The SEM imaging of its fracture surface, reported in Fig. 9(a), depicts multiple crack initiation points originating
from a shallow net-shape feature; with micrography, on the other hand, it was possible to observe a typical contour anomaly located
beneath the surface (see Fig. 9(b)) and from which a crack was nucleated, although it did not lead to the failure of the specimen.
Sharper edges and, effectively, a larger size caused the net-shape feature to develop a stronger crack driving force with respect to
the sub-surface feature, in spite of the larger effective dimension of this latter class of anomalies (as shown in Fig. 9(e)). For these
reasons surface roughness, for non-treated condition, is more detrimental than sub-surface defects, in agreement with previous
studies [12,61–66].
Engineering Failure Analysis 172 (2025) 109403
11

T. Risposi et al.
Fig. 8. Fractographies of specimens after Hirtisation®: (a) crack initiation from deep wavy valleys, (b) crack initiation from closed defects without sign of
corrosion, (c) crack initiation from defect etched at the top (red region) with enclosed portions (light blue zone).
The reduction of the cross section in samples subjected to Hirtisation®is represented by the grey area between the net-shape
surface and the contour region in Fig. 10. This process led to an improvement in surface roughness due to the removal of the deep
and sharp valleys typical of net-shape surface condition, as it was highlighted in Section 3.2. However, the Hirtisation®treatment
caused the largest contouring defects to become the anomalies at the origin of fracture, as seen from the fracture surfaces; these
flaws, in fact, were either characterized by a shorter distance to the surface of the samples (as in Fig. 8(b)) or completely exposed
to the open environment (as in Fig. 8(c)).
4.3. Statistical analysis of critical defects
Following the concepts of extreme value statistics (EVS) [67], the size of the most critical flaws observed from fracture surface
imaging was fitted with a largest extreme value distribution (LEVD) [68]:
𝐹LEVD(𝑥) = exp − exp −(𝑥−𝜆)
𝛿 (3)
in which 𝑥is the defect size, while 𝜆and 𝛿are,respectively, the location and shape parameters of the distribution.
A summary of the LEVD parameters for both surface conditions is reported in Table 7, along with the 50% percentile defect size
area50% and the average aspect ratio (𝑎
𝑐)𝑎𝑣𝑔, calculated averaging the (𝑎
𝑐)values observed on the fracture surfaces. Fig. 11 displays
the probability plots of the largest defects found on the fracture surfaces of net-shape and Hirtisation®specimens. The solid line
represents the 50% regression line, while the dashed lines represent the 95% confidence band.
The Gumbel probability plots of the failure initiation defects detected on the samples subjected to Hirtisation®(Fig. 11(b))
distinguished the two types of critical defects: those originated from elongated valleys and those related to sub-surface features
(i.e., contouring). The latter type of defects was found to be larger in size, confirming that the benefits introduced by the Hirtisation®
process, such as the improvements in surface roughness, are partially offset by the presence of more critical contour defects. If crack
Engineering Failure Analysis 172 (2025) 109403
12

T. Risposi et al.
Fig. 9. A relevant example of failure mechanism in net-shape case: (a) fracture origin with positioning of the reference line; (b) defect in the contour region
from which initiate crack without causing failure; (c) microstructure in the region around defect; (d) evaluation of the killer defects size from the fracture
surface; (e) evaluation of the effective sub-surface defect size.
Fig. 10. Scheme of the cross section of axial specimens before and after Hirtisation®post-process: the surface features of net-shape condition were removed
(dashed black line) and the defects located inside contour region (red features in dashed grey area) approaching the surface.
Table 7
Summary of LEVD parameters: comparison between net-shape surface and post-Hirtisation®.
Surface condition 𝜆[μm] 𝛿[μm] area50% [μm] (𝑎∕𝑐)𝑎𝑣𝑔
Net-shape 88 26 98 0.2
Post-Hirtisation®66 16 72 1
initiation had been solely due to surface roughness features, the improvements in fatigue strength would have been more significant;
this is further discussed in Section 5.2.
In the probability plot related to net-shape specimens, Fig. 11(a), no distinctions are highlighted since all killer defects are
represented by sharp features typical of the surface topography.
5. Modelling of fatigue properties
5.1. Fatigue strength model
The Kitagawa–Takahashi diagram [69] describes the variation of the fatigue limit as a function of defect size. One of the most
widely accepted models for this diagram is the equation proposed by El-Haddad [70], which is expressed with the Murakami’s
Engineering Failure Analysis 172 (2025) 109403
13

T. Risposi et al.
Fig. 11. LEVD probability plot of the largest defect found on fracture surfaces: (a) net-shape specimens, (b) post-Hirtisation specimens.
parameter as follows:
𝛥𝜎𝑤=𝛥𝜎𝑤,0⋅


area0
area +area0
(4)
area0=1
𝜋
⋅𝛥𝐾𝑡ℎ,𝐿𝐶
𝑌⋅𝛥𝜎𝑤,02(5)
𝛥𝜎𝑤is the fatigue strength of defective component, while 𝛥𝜎𝑤,0is the theoretical fatigue strength for the defect-free component and
thus it is computed as 𝛥𝜎𝑤,0= 2⋅0.4⋅UTS; 𝛥𝐾𝑡ℎ,𝐿𝐶 was obtained from the fit of the NASGRO equation, see Fig. 6and Table 6. The
shape factor was set as Y=0.65, in accordance with the surface killer defects detected from fractographies.
Table 8summarizes the parameters for the El-Haddad model of the Kitagawa–Takahashi diagram (4).Fig. 12 presents the
Kitagawa–Takahashi diagrams for both net-shape specimens and those treated with Hirtisation®, alongside the experimental results.
The comparison between the fatigue strength model and test data shows a good correlation between the predicted fatigue strength
and the experimental results: in details, it can be seen in Figs. 12(a) and 12(b) that the fatigue strength model accurately describes the
non-propagation of defects for the run-out specimens . This supports the use of fracture mechanics-based assessments in predicting
the fatigue properties of Ti6Al4V produced by AM.
Engineering Failure Analysis 172 (2025) 109403
14

T. Risposi et al.
Table 8
El-Haddad model parameters for the present PBF-LB/Ti6Al4V alloy.
R𝛥𝐾𝑡ℎ,𝐿𝐶 [MPa𝑚]𝛥𝜎𝑤,0[MPa] area0[μm]
−1 4.827 949.6 19.4
Fig. 12. Kitagawa diagrams for Ti6Al4V compared with experimental results on specimens: (a) Net-shape, (b) post-Hirtisation®.
Indeed, Fig. 12(b) also shows that failures due to defects generated by contouring are characterized by larger sizes compared to
crack initiation from mild valleys in specimens treated with Hirtisation®.
5.2. Life predictions of specimens
Life predictions were conducted by considering the initial defect size and aspect ratio, based on the failure initiating defects
observed on the fracture surfaces of the tested samples. Table 7lists the initial crack sizes and aspect ratios used in the crack
propagation algorithm for each series of samples, i.e., net-shape and after Hirtisation®.
The fatigue life estimates were carried out under the following assumptions:
•the crack growth model adopted was the NASGRO equation fitted as described in Fig. 6;
Engineering Failure Analysis 172 (2025) 109403
15

T. Risposi et al.
Fig. 13. Comparison between life predicted with propagation algorithm based on NASGRO equation and the experimental life for net-shape condition and after
Hirtisation®.
•the stress intensity factor solution is the one proposed by Newman and Raju [71];
•the failure condition depends on the following:
1. for surface cracks: crack depth, a, equal to one third of specimen’s thickness or crack width, c, equal to 0.4 the specimen’s
width;
2. for embedded cracks: crack depth, a, equal to a quarter of specimen’s thickness or crack width, c, equal to a quarter of
specimen’s width;
3. maximum net-section stress higher than 90% of the flow stress, 𝜎𝑓 𝑙 𝑜𝑤 ;
4. maximum stress intensity factor, 𝐾𝑚𝑎𝑥, higher than 0.7⋅𝐾𝐼 𝐶, where 𝐾𝐼 𝐶is the fracture toughness.
A comparison between the experimental data and predictions using the fatigue crack growth algorithm is presented in Fig. 13 for
both surface conditions. The continuous black line represents the locus of agreement between the predictions and the experimental
results, while the dashed lines indicate a factor 2 and 3 deviation between predictions and experiments. It can be observed that the
majority of predictions are conservative, because in the analysis the nucleation of cracks from the defects is not considered. This
effect is potentially more significant at the lowest stress levels.
This result can be seen also from the S–N diagrams reported in Figs. 14(a) and 14(b). The experimental S–N diagrams were
generated by fitting the data points in the finite life region with an equation of the type 𝑁𝑓=𝐴⋅𝛥𝜎𝐵, together with the 95% failure
probability regions marked with the dashed grey lines. These results confirm that the failure of PBF-LB/Ti6Al4V components can be
well described with a fatigue crack growth model based on the NASGRO equation. However, further discussion is required for the
case of life predictions for the post-Hirtisation®specimens. The failure of these specimens can be attributed to two types of defects:
surface troughs or subsurface contouring defects. Fig. 14(b) presents life estimates for three different initial crack assumptions: (i)
the continuous blue line represents life predictions for a crack size corresponding to the 50% percentile of the probability plot,
considering all defect types observed on the fracture surfaces; (ii) the dotted blue line reflects predictions for a defect size equal to
the mean value of wavy deep trough defects (see Fig. 11(b)); and (iii) the dashed blue line indicates life predictions for a defect
size corresponding to the mean value of contour defects.
These life predictions suggest that, for specimens treated with Hirtisation®, fatigue performance in the fatigue limit region could
be improved by approximately 35% in the case of failure driven by surface defects than the case of sub-surface contouring defects.
However, the presence of contour defects, which are brought to the surface by the Hirtisation®process, reduces this potential benefit.
The dimensions of these contour defects are comparable to those of surface defects found on the fracture surfaces of net-shape
samples. As a final result, experiments show that the Hirtisation®process still yields a 14% increase in the fatigue limit (𝛥𝜎𝑤=480
MPa vs. 𝛥𝜎𝑤=420 MPa), though this improvement is less pronounced than expected given the enhancement in surface quality.
In perspective, this finding suggests that the adoption of the Hirtisation®process should be accompanied by careful selection of
process parameters to avoid subsurface defects, even at the expense of surface quality, which can subsequently be improved by the
process itself.
Engineering Failure Analysis 172 (2025) 109403
16

T. Risposi et al.
Fig. 14. Life predictions for specimens: (a) net-shape, (b) post-Hirtisation®.
6. Conclusions
This study analysed the impact of Hirtisation®post process on the surface quality and on the fatigue performances of PBF-
LB/Ti6Al4V alloy. A wide experimental campaign on fatigue specimens was performed exploring two surface conditions: net-shape
and after Hirtisation®. The main conclusions are here summarized:
•the Hirtisation®process significantly improves the roughness and surface quality for net-shape surface of Ti6Al4V;
•The analysis of fracture surfaces of the fatigue samples shows that failure is mainly driven by two types of defects: (i) deep
and sharp valleys on the surface, (ii) sub-surface features in the contour region;
•In net-shape condition, surface roughness is more detrimental than sub-surface defects, even if in some cases short cracks were
observed to originate from sub-surface defects without causing failure;
•After chemical milling by Hirtisation®, due to an insufficient material removal, sub-surface anomalies are brought closer to
the surface and they become the critical features triggering fatigue failures;
•The benefits introduced by Hirtisation®are not completely exploited because of the harmful contour sub-surface defects near
the surface; this led to a limited improvement of the fatigue limit, i.e. +14%, when compared to the fatigue limit estimation
with shallow valleys;
Engineering Failure Analysis 172 (2025) 109403
17

T. Risposi et al.
•The process parameters (either the PBF-LB or the chemical milling ones) should be carefully selected, to avoid that sub-surface
anomalies could impair the potential fatigue strength enhancement due to improved surface quality.
CRediT authorship contribution statement
Tatiana Risposi: Writing – original draft, Visualization, Data curation. Lorenzo Rusnati: Writing – review & editing, Visu-
alization, Data curation. Luca Patriarca: Writing – review & editing, Visualization, Methodology. Alex Hardaker: Methodology,
Investigation. Dawid Luczyniec: Writing – review & editing, Resources. Stefano Beretta: Writing – review & editing, Supervision,
Methodology, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing
interests: Stefano Beretta reports financial support was provided by European Space Agency. If there are other authors, they declare
that they have no known competing financial interests or personal relationships that could have appeared to influence the work
reported in this paper.
Acknowledgements
The study was conducted within the call-off order of the European Space Agency (ESA/ESTEC) ‘‘ESA Additive Manufacturing
Benchmarking’’, contract number 4000133245/20/NL/AR/idb. The project involved ESA/ESTEC, The Manufacturing Technology
Centre (MTC) and Politecnico di Milano (PoliMi).
Data availability
Data will be made available on request.
References
[1] H. Bikas, P. Stavropoulos, G. Chryssolouris, Additive manufacturing methods and modelling approaches: a critical review, Int. J. Adv. Manuf. Technol. 83
(2016) 389–405, http://dx.doi.org/10.1007/s00170-015-7576- 2.
[2] E. Uhlmann, R. Kersting, T.B. Klein, M.F. Cruz, A.V. Borille, Additive manufacturing of titanium alloy for aircraft components, Procedia Cirp 35 (2015)
55–60, http://dx.doi.org/10.1016/j.procir.2015.08.061.
[3] P. Singh, H. Pungotra, N.S. Kalsi, On the characteristics of titanium alloys for the aircraft applications, Mater. Today: Proc. 4 (8) (2017) 8971–8982,
http://dx.doi.org/10.1016/j.matpr.2017.07.249.
[4] A. Bandyopadhyay, F. Espana, V.K. Balla, S. Bose, Y. Ohgami, N.M. Davies, Influence of porosity on mechanical properties and in vivo response of Ti6Al4V
implants, Acta Biomater. 6 (4) (2010) 1640–1648, http://dx.doi.org/10.1016/j.actbio.2009.11.011.
[5] B. Vrancken, L. Thijs, J.-P. Kruth, J. Van Humbeeck, Heat treatment of Ti6Al4V produced by selective laser melting: Microstructure and mechanical
properties, J. Alloys Compd. 541 (2012) 177–185, http://dx.doi.org/10.1016/j.jallcom.2012.07.022.
[6] J. Sun, Y. Yang, D. Wang, Parametric optimization of selective laser melting for forming Ti6Al4V samples by Taguchi method, Opt. Laser Technol. 49
(2013) 118–124, http://dx.doi.org/10.1016/j.optlastec.2012.12.002.
[7] J.J. Lewandowski, M. Seifi, Metal additive manufacturing: a review of mechanical properties, Annu. Rev. Mater. Res. 46 (2016) 151–186, http:
//dx.doi.org/10.1146/annurev-matsci- 070115-032024.
[8] S. Liu, Y.C. Shin, Additive manufacturing of Ti6Al4V alloy: A review, Mater. Des. 164 (2019) 107552, http://dx.doi.org/10.1016/j.matdes.2018.107552.
[9] W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform. 23 (2014) 1917–1928, http://dx.doi.org/10.1007/s11665-014-0958- z.
[10] D. Herzog, V. Seyda, E. Wycisk, C. Emmelmann, Additive manufacturing of metals, Acta Mater. 117 (2016) 371–392, http://dx.doi.org/10.1016/j.actamat.
2016.07.019.
[11] H.D. Nguyen, A. Pramanik, A. Basak, Y. Dong, C. Prakash, S. Debnath, S. Shankar, I. Jawahir, S. Dixit, D. Buddhi, A critical review on additive manufacturing
of Ti-6Al-4V alloy: Microstructure and mechanical properties, J. Mater. Res. Technol. 18 (2022) 4641–4661, http://dx.doi.org/10.1016/j.jmrt.2022.04.055.
[12] E. Wycisk, A. Solbach, S. Siddique, D. Herzog, F. Walther, C. Emmelmann, Effects of defects in laser additive manufactured Ti-6Al-4V on fatigue properties,
Phys. Procedia 56 (2014) 371–378, http://dx.doi.org/10.1016/j.phpro.2014.08.120.
[13] K.S. Chan, M. Koike, R.L. Mason, T. Okabe, Fatigue life of titanium alloys fabricated by additive layer manufacturing techniques for dental implants, Met.
Mater. Trans. A 44 (2013) 1010–1022, http://dx.doi.org/10.1007/s11661-012-1470-4.
[14] T.M. Mower, M.J. Long, Mechanical behavior of additive manufactured, powder-bed laser-fused materials, Mater. Sci. Eng.: A 651 (2016) 198–213,
http://dx.doi.org/10.1016/j.msea.2015.10.068.
[15] A. Triantaphyllou, C.L. Giusca, G.D. Macaulay, F. Roerig, M. Hoebel, R.K. Leach, B. Tomita, K.A. Milne, Surface texture measurement for additive
manufacturing, Surf. Topogr.: Metrol. Prop. 3 (2) (2015) 024002, http://dx.doi.org/10.1088/2051-672X/3/2/024002.
[16] A. Townsend, N. Senin, L. Blunt, R. Leach, J. Taylor, Surface texture metrology for metal additive manufacturing: a review, Precis. Eng. 46 (2016) 34–47,
http://dx.doi.org/10.1016/j.precisioneng.2016.06.001.
[17] M. Neikter, F. Forsberg, R. Pederson, M.-L. Antti, P. Åkerfeldt, S. Larsson, P. Jonsén, G. Puyoo, Defect characterization of electron beam melted Ti-6Al-4V
and alloy 718 with X-ray microtomography, Aeronaut. Aerosp. Open Access J. 2 (3) (2018) 139–145, http://dx.doi.org/10.15406/aaoaj.2018.02.00044.
[18] J. Pegues, M. Roach, R.S. Williamson, N. Shamsaei, Surface roughness effects on the fatigue strength of additively manufactured Ti-6Al-4V, Int. J. Fatigue
116 (2018) 543–552, http://dx.doi.org/10.1016/j.ijfatigue.2018.07.013.
[19] J. Gockel, L. Sheridan, B. Koerper, B. Whip, The influence of additive manufacturing processing parameters on surface roughness and fatigue life, Int. J.
Fatigue 124 (2019) 380–388, http://dx.doi.org/10.1016/j.ijfatigue.2019.03.025.
[20] S. Beretta, S. Romano, A comparison of fatigue strength sensitivity to defects for materials manufactured by AM or traditional processes, Int. J. Fatigue
94 (2017) 178–191, http://dx.doi.org/10.1016/j.ijfatigue.2016.06.020.
Engineering Failure Analysis 172 (2025) 109403
18

T. Risposi et al.
[21] N. Sanaei, A. Fatemi, Defects in additive manufactured metals and their effect on fatigue performance: A state-of-the-art review, Prog. Mater. Sci. 117
(2021) 100724, http://dx.doi.org/10.1016/j.pmatsci.2020.100724.
[22] A. Mostafaei, C. Zhao, Y. He, S.R. Ghiaasiaan, B. Shi, S. Shao, N. Shamsaei, Z. Wu, N. Kouraytem, T. Sun, et al., Defects and anomalies in powder bed
fusion metal additive manufacturing, Curr. Opin. Solid State Mater. Sci. 26 (2) (2022) 100974, http://dx.doi.org/10.1016/j.cossms.2021.100974.
[23] H. Gong, K. Rafi, H. Gu, T. Starr, B. Stucker, Analysis of defect generation in Ti–6Al–4V parts made using powder bed fusion additive manufacturing
processes, Addit. Manuf. 1 (2014) 87–98, http://dx.doi.org/10.1016/j.addma.2014.08.002.
[24] R. Cunningham, S.P. Narra, C. Montgomery, J. Beuth, A. Rollett, Synchrotron-based X-ray microtomography characterization of the effect of processing
variables on porosity formation in laser power-bed additive manufacturing of Ti-6Al-4V, JOM 69 (2017) 479–484, http://dx.doi.org/10.1007/s11837-016-
2234-1.
[25] Z. Chen, X. Wu, D. Tomus, C.H. Davies, Surface roughness of selective laser melted Ti-6Al-4V alloy components, Addit. Manuf. 21 (2018) 91–103,
http://dx.doi.org/10.1016/j.addma.2018.02.009.
[26] Z. Chen, S. Cao, X. Wu, C.H. Davies, Surface roughness and fatigue properties of selective laser melted Ti–6Al–4V alloy, in: Additive Manufacturing for
the Aerospace Industry, Elsevier, 2019, pp. 283–299, http://dx.doi.org/10.1016/B978-0- 12-814062- 8.00015-7.
[27] P. Karimi, C. Schnur, E. Sadeghi, J. Andersson, Contour design to improve topographical and microstructural characteristics of alloy 718 manufactured by
electron beam-powder bed fusion technique, Addit. Manuf. 32 (2020) 101014, http://dx.doi.org/10.1016/j.addma.2019.101014.
[28] A. Kleen, E. Glaubitz, J. Gockel, Influence of Contour Offset in Laser Powder Bed Fusion on Melt Pool Behaviors, Surface Roughness and Sub-Surface
Porosity, in: Proceedings Solid Freeform Pubblication Symposium, 2024, http://dx.doi.org/10.26153/tsw/58089.
[29] V. Sandell, P. Åkerfeldt, T. Hansson, M.-L. Antti, Fatigue fracture characterization of chemically post-processed electron beam powder bed fusion Ti–6Al–4V,
Int. J. Fatigue 172 (2023) 107673, http://dx.doi.org/10.1016/j.ijfatigue.2023.107673.
[30] J. Berglund, J. Holmberg, K. Wärmefjord, R. Söderberg, Detailed evaluation of topographical effects of hirtisation post-processing on electron beam powder
bed fusion (PBF-EB) manufactured Ti-6Al-4V component, Precis. Eng. 85 (2023) http://dx.doi.org/10.1016/j.precisioneng.2023.10.007.
[31] V. Sandell, J. Nilsson, T. Hansson, P. Åkerfeldt, M.-L. Antti, Effect of chemical post-processing on surfaces and sub-surface defects in electron beam melted
Ti-6Al-4V, Mater. Charact. 193 (2022) 112281, http://dx.doi.org/10.1016/j.matchar.2022.112281.
[32] M. Kahlin, H. Ansell, D. Basu, A. Kerwin, L. Newton, B. Smith, J. Moverare, Improved fatigue strength of additively manufactured Ti6Al4V by surface
post processing, Int. J. Fatigue 134 (2020) 105497, http://dx.doi.org/10.1016/j.ijfatigue.2020.105497.
[33] M. Basha, S. Basha, V. Jain, M. Sankar, State of the art on chemical and electrochemical based finishing processes for additive manufactured features,
Addit. Manuf. 58 (2022) 103028, http://dx.doi.org/10.1016/j.addma.2022.103028.
[34] Y. Sun, S. Gulizia, C. Oh, D. Fraser, M. Leary, Y. Yang, M. Qian, The influence of as-built surface conditions on mechanical properties of Ti-6Al-4V
additively manufactured by selective electron beam melting, JOM 68 (2016) 791–798, http://dx.doi.org/10.1007/s11837-015-1768- y.
[35] S. Bagehorn, T. Mertens, D. Greitemeier, L. Carton, A. Schoberth, Surface finishing of additive manufactured Ti-6Al-4V-a comparison of electrochemical
and mechanical treatments, in: 6th Eur Conf Aerosp Sci, 2015.
[36] B. Wysocki, J. Idaszek, J. Buhagiar, K. Szlązak, T. Brynk, K.J. Kurzydłowski, W. Święszkowski, The influence of chemical polishing of titanium scaffolds
on their mechanical strength and in-vitro cell response, Mater. Sci. Eng.: C 95 (2019) 428–439, http://dx.doi.org/10.1016/j.msec.2018.04.019.
[37] N. Stelzer, T. Sebald, M. Hatzenbichler, B. Bonvoisin, B. Lubos, M. Scheerer, Properties of surface engineered metallic parts prepared by additive
manufacturing, 2021.
[38] E. Beevers, D. Neumayer, B. Bonvoisin, A. Brandão, S. Hansal, M. Doppler, T. Rohr, B. Van Hooreweder, Effect of hirtisation treatment on surface
quality and mechanical properties of AlSi10Mg samples produced by laser powder bed fusion, Mater. Today Commun. 38 (2024) 108042, http:
//dx.doi.org/10.1016/j.mtcomm.2024.108042.
[39] N. Stelzer, M. Scheerer, Z. Simon, L. Bača, T. Sebald, H. Gschiel, M. Hatzenbichler, B. Bonvoisin, Mechanical properties of surface engineered metallic
parts prepared by additive manufacturing, in: The European Conference on Spacecraft Structures, Materials and Environmental Testing, 2018.
[40] J. Gumpinger, A. Brandão, E. Beevers, T. Rohr, T. Ghidini, S. Beretta, S. Romano, Expression of additive manufacturing surface irregularities through a
flaw-based assessment, in: Symposium on Structural Integrity of Additive Manufactured Parts, ASTM International, 2020, pp. 234–249, http://dx.doi.org/
10.1520/STP162020180098.
[41] J. Berglund, J. Holmberg, K. Wärmefjord, R. Söderberg, Detailed evaluation of topographical effects of hirtisation post-processing on electron beam powder
bed fusion (PBF-EB) manufactured Ti-6Al-4V component, Precis. Eng. 85 (2024) 319–327, http://dx.doi.org/10.1016/j.precisioneng.2023.10.007.
[42] ASTM E466-21, Standard practice for conducting force controlled constant amplitude axial fatigue tests of metallic Materials1, 2021, http://dx.doi.org/
10.1520/E0466-21.
[43] ASTM F3637-23, Standard guide for additive manufacturing of metal - finished part properties - methods for relative density measurement, 2023.
[44] R. Pippan, The growth of short cracks under cyclic compression, Fatigue Fract. Eng. Mater. Struct. 9 (5) (1987) 319–328, http://dx.doi.org/10.1111/j.1460-
2695.1987.tb00459.x.
[45] M.A. James, S.C. Forth, J.A. Newman, Load history effects resulting from compression precracking, J. ASTM Int. 2 (9) (2005) JAI12025.
[46] S. Cao, Y. Zou, C.V.S. Lim, X. Wu, Review of laser powder bed fusion (LPBF) fabricated Ti-6Al-4V: process, post-process treatment, microstructure, and
property, Light: Adv. Manuf. 2 (3) (2021) 313–332, http://dx.doi.org/10.37188/lam.2021.020.
[47] M. Benedetti, E. Torresani, M. Leoni, V. Fontanari, M. Bandini, C. Pederzolli, C. Potrich, The effect of post-sintering treatments on the fatigue and biological
behavior of Ti-6Al-4V ELI parts made by selective laser melting, J. Mech. Behav. Biomed. Mater. 71 (2017) 295–306, http://dx.doi.org/10.1016/j.jmbbm.
2017.03.024.
[48] H. Galarraga, R.J. Warren, D.A. Lados, R.R. Dehoff, M.M. Kirka, P. Nandwana, Effects of heat treatments on microstructure and properties of Ti-6Al-4V
ELI alloy fabricated by electron beam melting (EBM), Mater. Sci. Eng.: A 685 (2017) 417–428, http://dx.doi.org/10.1016/j.msea.2017.01.019.
[49] E.-. ASTM, Standard practice for statistical analysis of linear or linearized stress-life, american society for testing and materials, 2010.
[50] W. Dixon, The up-and-down method for small samples, J. Amer. Statist. Assoc. 60 (312) (1965) 967–978, http://dx.doi.org/10.1080/01621459.1965.
10480843.
[51] S. Beretta, L. Patriarca, M. Gargourimotlagh, A. Hardaker, D. Brackett, M. Salimian, J. Gumpinger, T. Ghidini, A benchmark activity on the fatigue life
assessment of AlSi10Mg components manufactured by L-PBF, Mater. Des. 218 (2022) 110713, http://dx.doi.org/10.1016/j.matdes.2022.110713.
[52] L. Barricelli, L. Patriarca, A. Du Plessis, S. Beretta, Orientation-dependent fatigue assessment of Ti6Al4V manufactured by L-PBF: Size of surface features
and shielding effect, Int. J. Fatigue 168 (2023) 107401, http://dx.doi.org/10.1016/j.ijfatigue.2022.107401.
[53] L. Patriarca, A. D’Andrea, M. Cova, L. Rusnati, S. Beretta, Cyclic R-curve measurements for structural metallic alloys, Adv. Eng. Mater. 2400447,
http://dx.doi.org/10.1002/adem.202400447.
[54] S. Leuders, M. Thöne, A. Riemer, T. Niendorf, T. Tröster, H.a. Richard, H. Maier, On the mechanical behaviour of titanium alloy TiAl6V4 manufactured
by selective laser melting: Fatigue resistance and crack growth performance, Int. J. Fatigue 48 (2013) 300–307, http://dx.doi.org/10.1016/j.ijfatigue.2012.
11.011.
[55] M.T. Hasib, H.E. Ostergaard, X. Li, J.J. Kruzic, Fatigue crack growth behavior of laser powder bed fusion additive manufactured Ti-6Al-4V: Roles of post
heat treatment and build orientation, Int. J. Fatigue 142 (2021) 105955, http://dx.doi.org/10.1016/j.ijfatigue.2020.105955.
[56] S.W.R. Institute, NASGRO reference manual version 8.1, 2016.
Engineering Failure Analysis 172 (2025) 109403
19

T. Risposi et al.
[57] V. Cain, L. Thijs, J. Van Humbeeck, B. Van Hooreweder, R. Knutsen, Crack propagation and fracture toughness of Ti6Al4V alloy produced by selective
laser melting, Addit. Manuf. 5 (2015) 68–76, http://dx.doi.org/10.1016/j.addma.2014.12.006.
[58] Y. Murakami, Metal Fatigue: Effects of Small Defects and Nonmetallic Inclusions, 2019.
[59] T. Ferreira, W. Rasband, Imagej user guide, ImageJ/ Fiji 1 (2012) 155–161.
[60] X. Zhao, S. Dadbakhsh, A. Rashid, Contouring strategies to improve the tensile properties and quality of EBM printed inconel 625 parts, J. Manuf. Process.
62 (2021) 418–429, http://dx.doi.org/10.1016/j.jmapro.2020.12.007.
[61] H. Masuo, Y. Tanaka, S. Morokoshi, H. Yagura, T. Uchida, Y. Yamamoto, Y. Murakami, Effects of defects, surface roughness and HIP on fatigue strength
of Ti-6Al-4V manufactured by additive manufacturing, Procedia Struct. Integr. 7 (2017) 19–26, http://dx.doi.org/10.1016/j.prostr.2017.11.055.
[62] S. Moon, R. Ma, R. Attardo, C. Tomonto, M. Nordin, P. Wheelock, M. Glavicic, M. Layman, R. Billo, T. Luo, Impact of surface and pore characteristics on
fatigue life of laser powder bed fusion Ti–6Al–4V alloy described by neural network models, Sci. Rep. 11 (1) (2021) 20424, http://dx.doi.org/10.1038/
s41598-021- 99959-6.
[63] R. Molaei, A. Fatemi, N. Phan, Significance of hot isostatic pressing (HIP) on multiaxial deformation and fatigue behaviors of additive manufactured
Ti-6Al-4V including build orientation and surface roughness effects, Int. J. Fatigue 117 (2018) 352–370, http://dx.doi.org/10.1016/j.ijfatigue.2018.07.035.
[64] Y. Sun, S. Lu, S. Gulizia, C. Oh, D. Fraser, M. Leary, M. Qian, Fatigue performance of additively manufactured Ti-6Al-4V: surface condition vs. internal
defects, JOM 72 (2020) 1022–1030, http://dx.doi.org/10.1007/s11837-020-04025- 7.
[65] G. Kasperovich, J. Hausmann, Improvement of fatigue resistance and ductility of TiAl6V4 processed by selective laser melting, J. Mater. Process. Technol.
220 (2015) 202–214, http://dx.doi.org/10.1016/j.jmatprotec.2015.01.025.
[66] D. Greitemeier, F. Palm, F. Syassen, T. Melz, Fatigue performance of additive manufactured TiAl6V4 using electron and laser beam melting, Int. J. Fatigue
94 (2017) 211–217, http://dx.doi.org/10.1016/j.ijfatigue.2016.05.001.
[67] S. Beretta, Y. Murakami, Statistical analysis of defects for fatigue strength prediction and quality control of materials, Fatigue Fract. Eng. Mater. Struct.
21 (9) (1998) 1049–1065, http://dx.doi.org/10.1046/j.1460-2695.1998.00104.x.
[68] E. Gumbel, Statistics of Extremes, Courier Corporation, 2012.
[69] H. Kitagawa, Applicability of fracture mechanics to very small cracks or the cracks in the early stage, in: Proc. 2nd Int. Conf. on Mechanical Behaviour
of Materials, 1976, pp. 627–631.
[70] Y. Murakami, Inclusion rating by statistics of extreme values and its application to fatigue strength prediction and quality control of materials, J. Res.-
Natl. Inst. Stand. Technol. 99 (1994) http://dx.doi.org/10.6028/jres.099.032, 345–345.
[71] J. Newman Jr., I. Raju, Analysis of Surface Cracks in Finite Plates Under Tension or Bending Loads, Tech. Rep., 1979.
Engineering Failure Analysis 172 (2025) 109403
20