Available

Available online
online at
at www.sciencedirect.com
www.sciencedirect.com

ScienceDirect
Available
Available online
online at at www.sciencedirect.com
www.sciencedirect.com
Structural
Structural Integrity
Integrity Procedia
Procedia 00
00 (2017)
(2017) 000–000
000–000
www.elsevier.com/locate/procedia
ScienceDirect
ScienceDirect
www.elsevier.com/locate/procedia

Procedia Structural
Structural Integrity 00
Integrity Procedia 7 (2017)
(2016)11–18
000–000
www.elsevier.com/locate/procedia

3rd International Symposium on Fatigue Design and Material Defects, FDMD 2017, 19-22
September 2017, Lecco, Italy

Defect Analysis
XV Portuguese and Fatigue
Conference Design
on Fracture, Basis
PCF 2016, forFebruary
10-12 Ni-based
2016, Superalloy
Paço de Arcos,718
Portugal
manufactured by Additive Manufacturing
Thermo-mechanical modeling of a high pressure turbine blade of an
airplane gas turbine engine
Yoichi Yamashitaaa, Takao Murakamiaa, Rei Miharaaa, Masami Okadabb,Yukitaka
a b c
P. BrandãoMurakami
, V. Infante
b,c *, A.M. Deus *
b,c

a
Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,
aIHI Corporation, Yokohama,
Yokohama, Japan
a Portugal
b IHI Corporation, Japan
IDMEC, Department of Mechanical Engineering,
bKMTL (Kobe Material InstitutoTesting
SuperiorLaboratory
Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,
Laboratory Co. Co. Ltd.),
Ltd.), Kobe,
Kobe, Japan
b
KMTL (Kobe Material TestingPortugal Japan
c
cc
CeFEMA, Department of Mechanical Engineering, Kyushu
Kyushu University,
University,
Instituto Fukuoka,
Fukuoka,
Superior Técnico, Japan
Japande Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa,
Universidade
Portugal

Abstract
Abstract
Abstract

It
It is
is well
During known
welltheir
known that
that high
operation, strength
highmodern metallic
strengthaircraft materials
metallicengine with
with Vickers
components
materials are hardness
Vickers subjectedHV>400
hardness are
are very
to increasingly
HV>400 sensitive
sensitive to
very demanding small
small defects.
to operating This
conditions,
defects. This
paper
paper discusses
especially
discusses thefatigue
fatigue properties
high pressure
properties of
of aa Ni-based
turbine Superalloy
(HPT) blades.
Ni-based Such 718
Superalloy with
with HV=~470
conditions
718 cause these
HV=~470 which
parts
which was
wasto manufactured
undergo different
manufactured by
by additive
types ofmanufacturing
additive time-dependent
manufacturing
(AM).
(AM). The
The advantage
degradation, one of of
advantage of AM
which
AM is has been
hascreep.
been A emphasized
model using
emphasized as the potential
as the application
finite element
potential methodto
application high
high strength
to(FEM) or
or hard
hard steels
was developed,
strength in order
steels which
which are
to be difficult
areable to
to predict
difficult to
manufacture
the creep by
manufacture by traditional
behaviour
traditional of machining
HPT blades.
machining to
to complex shapes.
Flight data
complex However,
records
shapes. (FDR)
However, the disadvantage
thefor a specific or
disadvantage challenge
oraircraft,
challenge of
of AM
provided
AM by has
has abeen pointed
pointed out
commercial
been due
aviation
out due
to defects
defects which
to company, whichwereare inevitably
areused to obtain
inevitably contained
thermalin
contained the
inand manufacturing
the mechanical
manufacturing process.
data for three different flight cycles. In order to create the 3D model
process.
needed
Defects
Defects of forthe
of thematerial
the FEM analysis,
material investigated
investigateda HPTin blade
in this
this studyscrap
study were was
were mostly
mostlyscanned,
gas and its
gas porosity
porosity andchemical
and those
those made
madecomposition
by
by lack
lack of and material
of fusion.
fusion. The
The √areaproperties
√area parameter
parameterwere
obtained.
model
model was The data that
was confirmed
confirmed the was
the gathered
successful
successful was fed into
application.
application. the FEM
Although
Although themodel
the andof
statistics
statistics ofdifferent
extremes
extremes simulations
analysis
analysis isiswere run,
useful
useful forfirst
for thewith
the a simplified
quality
quality control of3D
control of
AM,
AM,rectangular
the blocksurface
the particular
particular shape,effect
surface in order
effect on to better
on the
the establish
effective
effective valuethe
value of model,size
of defect
defect andmust
size thenbe
must with
be the realconsidered.
carefully
carefully 3D mesh obtained
considered. Since thefrom
Since the the bladeof
orientations
orientations ofscrap.
defects
defectsThe
in AM materials are random, a defect in contact with specimen surface has higher influence and has the effective larger size termed a
in overall
AM expected
materials arebehaviour
random, ain terms
defect inof displacement
contact with was
specimen observed,
surface in
hasparticular
higher at the
influence trailing
and edge
has the of the
effectiveblade. Therefore
larger size such
termed
as model
as √areaeff
√area can be the
eff than
than useful
the realinsize,
real the goal
size, √area,
√area,of ofpredicting
of the
the defect
defect turbine
from blade
from the life, given
the viewpoint
viewpoint ofafracture
of set of FDR
fracture data. The
mechanics.
mechanics. The guide
guide for
for the
the fatigue
fatigue design
design and
and
development
development of of higher
higher quality
quality Ni-based
Ni-based Superalloy
Superalloy 718 718 by
by AMAM processing
processing based
based onon the
the combination
combination of of the
the statistics
statistics of
of extremes
extremes
on©defects
on 2016 The
defects and
and theAuthors.
the √area Published
√area parameterby
parameter Elsevier
model
model is B.V.
is proposed.
proposed.
Peer-review under responsibility of the Scientific Committee of PCF 2016.
Copyright © 2017 The Authors. Published by Elsevier B.V.
Peer-review
© 2017 under responsibility of the Scientific Committee of the 3rd International Symposium on Fatigue Design and Material Defects.
2017 The
The Authors.
© Keywords: High Pressure
Authors. Published
Turbineby
Published Elsevier
Blade;
by Creep;
Elsevier B.V.
Finite Element Method; 3D Model; Simulation.
B.V.

*
* Corresponding
Corresponding author.
author. Tel.:
Tel.: +81-92-606-3832;
+81-92-606-3832; fax:
fax: +81-92-606-3832.
+81-92-606-3832.
E-mail address: murakami.yukitaka.600@m.kyushu-u.ac.jp
E-mail address: murakami.yukitaka.600@m.kyushu-u.ac.jp

2452-3216©
2452-3216© 2017
2017 The
The Authors.
Authors. Published
Published by
by Elsevier
Elsevier B.V.
B.V.
Peer-review
Peer-review under
under responsibility
responsibility of
of the
the Scientific
Scientific Committee
Committee of
of the
the 3rd
3rd International
International Symposium
Symposium on
on Fatigue
Fatigue Design
Design and
and Material
Material
Defects.
* Corresponding author. Tel.: +351 218419991.
Defects.
E-mail address: amd@tecnico.ulisboa.pt

2452-3216 © 2016 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the Scientific Committee of PCF 2016.
2452-3216 Copyright  2017 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Scientific Committee of the 3rd International Symposium on Fatigue Design and Material Defects.
10.1016/j.prostr.2017.11.054
2 Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000

Peer-review
12 under responsibility of the Scientific Committee
Yoichi Yamashita et al.of the 3rd International
/ Procedia Symposium
Structural Integrity on Fatigue Design and Material
7 (2017) 11–18
Defects.

Keywords:Fatigue, Additive Manufacturing (AM), Defects, Statistics of extremes, the √area parameter model

1. Introduction
The advantage of AM has been emphasized as the potential application to high strength or hard steels which are
difficult and costly to manufacture by traditional machining to complex shapes. However, the disadvantage or
challenge of AM has been pointed out due to defects which are inevitably contained in the manufacturing process and
detrimental to fatigue strength.
Many literature on fatigue properties of AM materials have been published in recent years. The review paper by
Berreta, S and Roman, S (2017) gives a detailed and thorough analysis on the fundamental problems to be studied on
AM materials. Günther, J. et als. (2017) carried out precise experimental observations on Ti-6Al-4V in high cycle
fatigue and very high cycle fatigue and based on the observation they discussed the problem from the viewpoint of
statistical scatter of defect size. They pointed out the problem raised by the interaction between defects and specimen
surface.
This paper discusses fatigue properties of a Ni-based Superalloy 718 manufactured by AM in terms of the effect of
defects. The guide for the fatigue design and development of high quality Ni-based Superalloy 718 by AM processing
will be presented based on the combination of the statistics of extremes on defects and the √area parameter model.

2. Material, Specimen and Experimental method

The material was made by AM processes, Selective Laser Melting (SLM) method. Solution treatment is based on
AMS5663. Specimens were machined from the two kind of raw plate materials, denoted by A and B, produced by
AM. Specimens were cut in two directions, i.e. as-built direction (L direction) and perpendicular (T direction) to as-
built direction as shown in Fig. 1 (Case for Material A and similarly for Material B). Figure 2 shows the shape and
dimension of specimen. The mechanical properties were measured by using the fatigue specimens. The 0.2% proof
stress ranged from 1227MPa to 1329MPa and the ultimate tensile strength ranged from 1306MPa to 1499MPa. The
elongation with 8mm gauge length ranged from 13.6% to 31.8% and the reduction area from 8.6% to 30.7%. It was
revealed by observation of fracture surfaces that the scatter of the mechanical properties were caused by various
defects contained in specimens. Specimen surface was polished by emery paper with #600. The remaining materials
of Fig.1 after cutting specimens were used to investigate the microstructure and statistical distribution of defects.
The analysis of statistics of extremes was applied to the largest defects observed on 9 sections with the observation
area S0 = 80.97mm2 for Material A and S0 = 116.49mm2 for Material B. The largest defects were separately analysed
on pores, linear defects and equivalent elliptical defects for interactive adjacent defects．The Vickers hardness HV
(P=5kgf) was measured at 5 points. HV＝465±1.7% for A-T and HV＝474±1.2% for B-T. Tension compression
fatigue tests were carried out with hydraulic tension-compression testing machine at 30Hz with strict specimen
alignment within ±5% for the values of 4 strain gauges attached to each specimen at ±500µε and ±1000µε. The fatigue
fracture origins were mostly at defects.

A-T1

A-T2
（φ6.8)

A-T3
R40
φ5
65

A-T4

A-T5

A-T6

A-T7
65
（18） 12 （18）
A-L1 A-L2 A-L3 A-L4 A-L5 A-L6 A-L7
（64）
15

Fig. 1 Raw plate material (Material A) and cutting layout for specimens Fig. 2 Shape and dimension of tension-compression specimen
 Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000 3
Yoichi Yamashita et al. / Procedia Structural Integrity 7 (2017) 11–18 13

3. Results and Discussion

When we discuss the fatigue properties of materials made by AM, we should first pay attention to the ideal or upper
bound of fatigue strength which we can expect for the case without influence of defects. It is well known as shown in
Fig. 3 that there is a very good correlation between fatigue limit and Vickers hardness HV up to HV=~400. For HV>400,
fatigue limit drops drastically due to presence of small defects (Murakami, Y. 2002). There is the robust empirical
formula (Eq. (1)) between fatigue limit σw and HV for HV<400 (Nishijima, S. 1980, and Murakami, Y. 2002).

σw,ideal = 1.6HV ± 0.1HV (1)

Where, the units are σw,ideal in MPa and HV in kgf/mm2. Since the Vickers hardness of the material investigated in
this study ranges from HV=465 to 474, the ideal fatigue strength can be estimated to be around σw,ideal = 744-758
MPa.

(a) N: normalized, Q.T: quenched and tempered, (b) Zero mean stress. [Garwood, M.F. et al.1951]

R=-1 zero mean stress specimens. [Nishijima, Y. et al. 1980]

Fig. 3 Relationship between hardness and the fatigue limit (Murakami, Y. (2002)).

Figure 4 shows S-N data for Material A. There is no apparent difference in fatigue strength between T- and L-
directions. All specimens fractured from defect and fatigue failure results show a large scatter due to scatter of the
defects at fracture origin. The locations of defects of fracture origins are mostly in contact with specimen surface. The
specimens which ran out for N=107 cycles were tested again at higher stress to identify the fatal inclusion which lead
specimen to failure. The second or third test was carried out at stress 40MPa higher than the previous test to avoid the
coaxing effect. Figures 5(a) and (c) are typical defects at fracture origins in Material A. Figure 5 (b) is identified as a
pore.
800

□ ： Material A （dierction - L）
700 ■ ： Material A （dierction - T）

600
Stress Ampｌitude σ a [MPa]

500
B3
A2
C2
400 B2
C1
300 B1
A1

200

100

Runout
0
1.E+04
10
4 105
1.E+05 106
1.E+06 107
1.E+07 108
1.E+08

Cycles to Failure Nf [cycles]

Fig. 4 S-N data of Material A

414 Yoichi Yamashita
Yoichi Yamashita et al. / Procedia
et Al./ Structural Integrity Structural
Procedia 00Integrity
(2017)7000–000
(2017) 11–18

(a) (b) (c)

Fig. 5 Defects at fatigue fracture origin of Material A

800

○ ： Material B （dierction - L）
700 ● ： Material B （dierction - T）

600
Stress Ampｌitude σa [MPa]

500
b2
a2
400

300
b1

200
a1

100

Runout
0
1.E+04
10 4 105
1.E+05 106
1.E+06 10 7
1.E+07 10 8
1.E+08

Cycles to Failure Nf [cycles]

Fig. 6 S-N data of Material B

(a) (b)

Fig. 7 Defects at fatigue fracture origin of Material B

Figure 6 shows S-N data for Material B. Figure 7 shows the defects observed at fracture origin for Material B.
These defects have various kinds and irregular shapes.
Figure 8 shows a defect which was observed at fracture origin in contact with specimen surface in Material A. It
was very common at fracture origin to observe defects inclined to specimen surface. Compared to rolled steels, the
direction of defects in AM materials is random and not aligned in the identical direction. The varieties of defects are
originated from lack of fusion or various qualities of powder. Since it is very difficult to identify the presence, shape
and size of defects in advance of fatigue test, it is extremely difficult to predict the fatigue limit of individual specimens
prior to fatigue test. Nevertheless, in order to quantify the fatigue limit, we need to pay attention to the mechanics
aspect of defects. The representative dimension of a defect can be expressed with √area in terms of fracture mechanics
concept (Murakami, Y. 2002). From the viewpoint of the statistics of extremes, defects larger than the defect observed
at the fracture origin should possibly exist in the specimen. Then, why did the defects at fracture origin appear mostly
 Yoichi Yamashita et al. / Procedia Structural Integrity 7 (2017) 11–18 15
Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000 5

in contact with specimen surface? We can explain the reason as follows.

The initial fatigue crack growth for irregularly shaped cracks and defects as shown in Fig. 9 (Murakami, Y. 2002)
starts from the deepest concave corner point due to the extremely high stress intensity factor at that point. However,
as the crack grows and the shape of the crack becomes round, the stress intensity factor once decreases and continues
growing to failure if the value of ∆K is higher than ∆Kth. If the value of ∆K is lower than ∆Kth, the crack stops growth
and becomes nonpropagating crack. Therefore, if a defect exists at or near specimen surface, we must consider the
effective size of defect which is larger than the real size of defect. Thus, in case of Fig. 8 the effective equivalent crack
size √areaeff must be estimated by the dotted line.
√areaeff

(a) (b) d2 e d1
(a) (b)
(d) e<d2 →√areaeff

√areaeff √areaeff
d

(c) e < d→√areaeff (e) √areaeff = Dotted area

Fig. 9 Effective size (dotted line) of irregularly shaped defect
Fig. 8 Inclined defect in contact with specimen surface
and the effective defect size (dotted line).

The effective maximum value of the stress intensity factor K for this kind of inclined defect is attained at the angle
between defect and specimen surface with 45deg. Therefore, when we apply the statistics of extremes to fatigue design
of AM materials, we need to consider the modification of defect size based on this fact.
As shown in Fig.4, the specimen A1 which ran out N=2x107 and B1, B2 and C1 which ran out N=107 were tested
again at higher stress to identify the fatal defect. The data tested at higher stress are denoted by A2, B2, B3 and C2 in
Fig.4. Similar tests were carried out for specimens a1, b1, a2 and b2 of Material B in Fig. 6. Based on the values of the
effective defect size √areaeff and the Vickers hardness HV, the fatigue limit σw can be estimated by the following
equation of the √area parameter model.

σw = 1.43(HV+120)/(√areaeff)1/6 (2)

Where, the units are σw: MPa, HV:kgf/mm2, √areaeff: µm.

The normalized S-N data were made as in Fig. 10 (a) and (b) where the applied stress σ is normalized by the
estimated fatigue limit σw. It can be seen that the value of σ/σw for failed specimens are mostly larger than ~0.9 and
the estimation based on the √area parameter model works well. However, some specimens of Material B failed at
2.0 2.0
○ ： Material B （direction - L）
1.8 □ ： Material A （direction - L） 1.8 ● ： Material B （direction - T）
■ ： Material A （direction - T）
1.6 1.6

1.4 1.4 b2

1.2 1.2
σa/σw

A2
σa/σw

a2
1.0 C2 1.0
C1
0.8 0.8
B3 A1
B2 b1
0.6 0.6
B1
0.4 0.4 a1

0.2 0.2
Runout Runout
0.0 0.0
1.E+04
10 4 1.E+05
10 5 1.E+06
106
1.E+07
107 1.E+08
108 104
1.E+04 1.E+05
10 5 1.E+06
10 6 1.E+07
10 7 1.E+08
10 8

Cycles to Failure Nf [cycles] Cycles to Failure Nf [cycles]

(a) Material A (b) Material B

Fig. 10 Normalized S-N curve: σ/σw - Nf
16 Yoichi Yamashita et al. / Procedia Structural Integrity 7 (2017) 11–18
6 Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000

the values σ/σw close to 0.8 around Nf = 106. At present, the reason is not clear. Although the Vickers hardness of
Material B is almost the same as Material A, the fatigue strength is lower than Material A. The same trend was
observed for additively manufactured Ti-6Al-4V with larger particle (Masuo, H. et al. 2017). These results suggest
that more precise estimation of the effective size √areaeff is necessary for the complex configuration of multi-defects
as shown in Fig. 9. The values of σ/σw for A1, B1, B2, and C1 in Fig. 4 and a1 and b1 in Fig.6 which ran out are
confirmed lower than 1.0 by calculation with the defect sizes obtained by the subsequent tests at higher stress. For
example, the values of σ/σw for the specimens of Material A which ran out longer than N=107 were summarized as
follows.
① Applied stress σ = 300MPa, N = 2×107, √area=179µm. σw=352MPa. σ/σw＝0.85.
② Applied stress σ = 365MPa, N = 107, √area=19µm. σw =512MPa. σ/σw＝0.71.
③ Applied stress σ = 350MPa, N = 107, √area=74µm. σw =408MPa. σ/σw＝0.86.
Thus, it is confirmed that these previous tests were carried out below the fatigue limit of each specimen.
Since the fatigue limit of AM specimens is influenced by the size of defects contained in individual specimens, we
must understand that we cannot define a constant fatigue limit for a material in question from the usual S-N data.
Figure 11 shows the statistics of extreme analysis of the largest defects which appeared on the sections cut from the
raw plate materials within the area S0 = 80.97mm2 for Material A and S0 = 116.49mm2 for Material B. Regarding linear
defects, the length of linear defect was also plotted in addition to the √area, because the effective value of √area must
be estimated by using the length of defect based on the concept explained in Figs. 8 and 9. Although these analyses
are based on 2D measurement, the statistics of extremes analysis such as Fig. 11 will be useful to improve the quality
of AM processes. From the viewpoint of the quality control of AM materials based on defect size, A is graded higher
than B.
8.0 8.0

7.0 7.0

6.0 6.0

Pore
ｙj （%）
ｙj （%）

5.0
Pore 5.0

4.0 Elliptical defect

4.0
Elliptical defect
Reduced variate
Reduced variate

3.0 3.0

2.0 2.0

1.0
1.0

0.0
0.0
Linear defect ℓ
-1.0
-1.0
Linear Defect ℓ
S0 = 80.97mm2 S0 = 116.49mm2
-2.0
-2.0
0 20 40 60 80 100
0 50 100 150 200
√area max（μm） or ℓ √area max（μm） or ℓ

(a) Material A (b) Material B

Fig.11 Statistics of extremes of the defects of the raw plate materials.

Figure 12 shows the statistics of extremes analysis of the defects which were observed at fatigue fracture origins.
In this analysis, the modification of √area was applied based on the rule of Figs. 8 and 9 and the √areaeffmax was
estimated.
From the above discussion, it is necessary for the fatigue design of AM components to consider the method of the
statistics of extremes based on not only Fig. 11 (even with the 3D measurement) but also the concept of Fig.12 (Figs.
8& 9). Considering the volume and number of productions of the components in question, the effective largest defect
√areaeffmax contained in large or many components can be predicted. The lower bound of the fatigue limit σwl based
on √areaeffmax can be determined by the following equation.

σwl = 1.43(HV+120)/(√areaeffmax）1/6 (3)

 Yoichi Yamashita et al. / Procedia Structural Integrity 7 (2017) 11–18 17
Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000 7

Where, the units are σwl: MPa, HV:kgf/mm2, √areaeffmax: µm.

8.0

7.0

6.0
Material A

Reduced variate ｙj （%）

5.0

4.0

Material B
3.0

2.0

1.0

0.0

-1.0
V 0 = 235.6mm3
-2.0
0 100 200 300 400 500

√area max（μm）

Figure 12 Statistics of extremes of the defects at fracture origin of Material A and B in terms of √areaeffmax.
(V0 : Specimen volume)

Regarding the fatigue limit of Ni-based superalloy 718, a supplementary description may be necessary in terms of
VHCF. There are several literature which report fatigue failure at cycles longer than 107 cycles from subsurface grain
showing a facet at fracture origin. Thus, the possibility of “no fatigue limit” has arose for this alloy. Similar results
were reported on Ti-6Al-4V. However, this phenomenon can be understood as the failure from a large grain as the
extreme value of grain size distribution which is preferentially oriented to applied stress and is eventually regarded
equivalent to a defect with delayed crack initiation in interior of specimen. Actually, the application of the √area
parameter model to these large grains with facet works quite well.

4. Conclusions
The fatigue properties of a Ni-based Superalloy 718 manufactured by AM were studied and the perspective for
fatigue design was discussed from the viewpoint of the effect of defects.
1. The Ni-based Superalloy 718 manufactured by AM contained various irregularly shaped defects which were
observed by microstructural investigation on the as-built material and also at fatigue fracture origins.
2. Defects were mostly gas porosity and those made by lack of fusion.
3. There is no apparent difference in fatigue strength between T- and L- directions both for Material A and B.
4. Particular effect of surface on defects present near surface must be carefully considered. Since the orientations of
defects in AM materials are random, a defect in contact with specimen surface has higher influence in terms of
the effective defect size √areaeff than the real size of the defect from the viewpoint of fracture mechanics.
5. The statistics of extremes analysis based on the fracture mechanics evaluation of defects is useful for the quality
control of AM.
6. Considering the volume and number of productions of components, the lower bound of the fatigue limit σwl based
on √areaeffmax can be predicted by the √area parameter model.

References
Beretta, S., Roman, S., 2017.A comparison of fatigue strength sensitivity to defects for materials manufactured by AM or traditional processes,
Int. J. Fatigue, 94, 178–191.
Günther,J., Krewerth, D., Lippmann, T., Leuders, S., Tröster, T., Weidner, A., Biermann, H.,Niendorf, T.,2017. Fatigue life of additively
manufactured Ti–6Al–4V in the very high cycle fatigue regime, International Journal of Fatigue 94, 236–245., and References included in
this paper.
18 Yoichi Yamashita et al. / Procedia Structural Integrity 7 (2017) 11–18
8 Yoichi Yamashita et Al./ Structural Integrity Procedia 00 (2017) 000–000

Garwood, M.F., Zurburg, H.H., and Erickson, M.A., 1951. Correlation of Laboratory Tests and Service Performance, Interpretation of Tests
and Correlation with Service, ASM, Philadelphia, PA, pp. 1–77.
Masuo, H., Tanaka, Y., Morokoshi, S., Yagura, H., Uchida, T. Yamamoto, Y. Murakami, Y., 2017, Effects of Defects, Surface Roughness
and HIP on Fatigue Strength of Ti-6Al-4V manufactured by Additive Manufacturing, Structural Integrity/Procedia, 3rd Int. Sym. FDMD,
Lecco, Italy.
Murakami, Y., 2002. Metal Fatigue: Effects of Small Defects and Nonmetallic Inclusions, Elsevier.
Murakami, Y., 2012. Material defects as the basis of fatigue design, Int. J. Fatigue, Vol.41, pp.2-10.
Murakami, Y. and Endo, M., 1994. Effects of defects, inclusions and inhomogeneities on fatigue strength, Int. J. Fatigue, Vol.16, pp.163-182.
Murakami, Y., Beretta, S., 1999. Small defects and inhomogeneities in fatigue strength: experiment, models and statistical implications,
Extremes, 2(2), 123-147.
Murakami, Y. and Nemat-Nasser, S. 1983. Growth and Stability of Interacting Surface Flaws of Arbitrary Shape, Engng. Frac. Mech., 17-1,
193-210.
Nishijima, S., 1980. Statistical Analysis of Fatigue Test Data, J. Soc. Mater. Sci., Japan, 29(316) , 24–29.

Appendix Brief Summary of the √area Parameter Model

Many fatigue limit prediction models have been proposed. Regarding the material parameters, most of models adopt
one or more fatigue data such as ∆Kth or ∆JR or S-N data. If we adopt the reference fatigue data which were obtained
in a condition close to a prediction problem, the model may be accurate. However, such a model is a kind of so-called
data fitting model and cannot be easily applied to other materials.
In the √area parameter model, only one material parameter, Vickers hardness HV, is used. The reason is that there
is the robust empirical relationship between HV and σw for HV<400 as follows (Fig. 3).
σw = 1.6HV ± 0.1HV (A1)
Where, the units are σw in MPa and HV in kgf/mm2.
Based on FEM analysis, the background of Eq. (A1) can be correlated to tensile strength and cyclic stress-strain
constitutive property. For HV>400, fatigue limit cannot attain the value of Eq. (A1) and drops due to presence of small
defects and nonmetallic inclusions. It is known that if there exists defects larger than a critical size, the value of Eq.
(A1) cannot be guaranteed even for HV<400. Therefore, we need to seek the geometrical parameter which crucially
reflects the fatigue threshold. Based on the concept, the square root of the projection area of defects, √area, was
discovered as the representative geometrical parameter which expresses the maximum stress intensity factor Kmax
along the front of 3D crack as shown in Fig. A1.
Thus, using one material parameter, HV, and one geometrical parameter, √area, and examining many fatigue data
on artificial small defects, the prediction equation for small cracks and defects was developed as follows.
σw = C0(HV+120)/(√area)1/6 (A2)
Where, C0 = 1.43 for surface cracks and C0 = 1.56 for subsurface cracks, and the units are σw: MPa, HV: kgf/mm2,
√area: µm.
In Eq. (A2), the term (HV + 120) must be carefully interpreted, because σw is not simply proportional to HV. For
example, the quantity (HV + 120) is doubled for HV = 120. But it is only 1.2HV for HV = 600. Adding a constant 120
to HV reflects the easiness of crack closure for soft metals. Thus, the √area parameter model is not a simple empirical
formula but it was made based on the thorough consideration on the fracture mechanics and a strength parameter
which includes the intrinsic strength of microstructure and crack closure tendency.
Equation (A2) is not valid for a very small defect if the value of σw calculated by Eq. (A2) exceeds the fatigue limit
of unnotched specimen σw0. Such a very small defect must be regarded as non-damaging defect. On the other hand,
for valid application of Eq. (A2) there is the upper bound of √area depending on HV of materials.

Fig. A1