Available online at www.sciencedirect.

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

ScienceDirect
Procedia Manufacturing 00 (2021) 000–000
Procedia
Procedia Manufacturing
Manufacturing 00
00 (2021)
(2021) 000–000
000–000 www.elsevier.com/locate/procedia
Procedia Manufacturing 00 (2021) 000–000 www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
Procedia Manufacturing
Procedia 54 (2021)
Manufacturing 238–243
00 (2021) 000–000 www.elsevier.com/locate/procedia
Procedia Manufacturing 00 (2021) 000–000 www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
10th
10th CIRP
CIRP Sponsored
Sponsored Conference
Conference on
on Digital
Digital Enterprise
Enterprise Technologies
Technologies (DET
(DET 2021)
2021) –– Digital
Digital Technologies
Technologies as
as
Enablers on
10th CIRP Sponsored Conference of Industrial Competitiveness
Digital Enterprise and Sustainability
Technologies (DET 2021) – Digital Technologies as
10th Enablers of Industrial Competitiveness and Sustainability
10th CIRP
CIRP Sponsored
Sponsored Conference
Enablers on
Conference Digital
Digital Enterprise
of Industrial
on Technologies
Competitiveness
Enterprise (DET
(DET 2021)
2021) –
and Sustainability
Technologies – Digital
Digital Technologies
Technologies as
as
Design and additive
Design and additive manufacturing
Enablers of Industrial
manufacturing
Enablers of a fatigue-critical
Competitiveness and aerospace
aerospace part
Sustainability
of a fatigue-critical
of Industrial Competitiveness and Sustainability part using
using
Design
Design and additive manufacturing
topology optimization of aa and
fatigue-critical
L-PBF aerospace
process part using
Design and
and additive
additive manufacturing
topology optimization
manufacturing
topology optimization
of
of a fatigue-critical
and L-PBF
fatigue-critical
and L-PBF
aerospace
process
aerospace
process
part
part using
using
topology
Dagkoluoptimization
topology and
and L-PBF process
Dagkoluoptimization L-PBF process
a,b,∗ a b
Akin a,b,∗ , Istemihan Gokdag
a,b,∗ aa , Oguzhan Yilmazbb
Akin a,b,∗ , Istemihan Gokdaga , Oguzhan Yilmazb
Akin Dagkolu , Istemihan Gokdag , Oguzhan Yilmaz
a,b,∗ Inc., Fethiye Mh. Havacılık Blv.
a Turkish Aerospace Industries a No:17 06980 Ankara/Turkeyb
a Akin
Akin Dagkolu
a Turkish Aerospace Industries
a Turkish
b Advanced Manufacturing Technologies Dagkolu
Aerospace a,b,∗ Inc.,
Industries
Research Group, Faculty
,, Istemihan
Istemihan Gokdag
Inc., Fethiye
Fethiye Mh.
Gokdag ,, Oguzhan
Mh. Havacılık
of Engineering,Gazi Oguzhan Yilmaz
a No:17
Havacılık Blv.
Blv. No:17 06980
Yilmaz
University,06980 Ankara/Turkey
Ankara/Turkey
Celal Bayar
b
Blv, 06570, Maltepe, Ankara/Turkey
bb Advanced
Advanced Manufacturing Turkish
Manufacturing TechnologiesAerospace
Technologies Industries
Research Group, Inc.,
Group, Faculty Fethiye
Faculty of Mh. Havacılık
of Engineering,Gazi Blv. No:17 06980
Engineering,Gazi University,
University, Ankara/Turkey
Celal Bayar
Bayar Blv, 06570,
06570, Maltepe, Ankara/Turkey
Ankara/Turkey
a Research Celal Blv, Maltepe,
* Corresponding a Turkish
b Advanced Manufacturing Technologies
author. E-mail address:
Aerospace Industries
Research
akin.dagkolu@gazi.edu.tr
Turkish Aerospace Industries
Inc., Fethiye
Group, Faculty
Inc., Fethiye
Mh. Havacılık Blv.
of Engineering,Gazi
Mh. Havacılık Blv.
No:17 06980
University,
No:17 06980
Ankara/Turkey
Celal Bayar Blv, 06570,
Ankara/Turkey Maltepe, Ankara/Turkey
* b Advanced
Corresponding author. E-mail
* Corresponding author.
b Advanced E-mail address:
Manufacturing
Manufacturingaddress: akin.dagkolu@gazi.edu.tr
Technologies Research Group, Faculty of Engineering,Gazi University, Celal Bayar Blv, 06570,
akin.dagkolu@gazi.edu.tr
Technologies Research Group, Faculty of Engineering,Gazi University, Celal Bayar Blv, 06570,
Maltepe, Ankara/Turkey
Maltepe, Ankara/Turkey
* Corresponding author. E-mail address: akin.dagkolu@gazi.edu.tr
* Corresponding author. E-mail address: akin.dagkolu@gazi.edu.tr
* Corresponding author. E-mail address: akin.dagkolu@gazi.edu.tr
Abstract
Abstract
Abstract
Abstract
Additive Manufacturing (AM) is a new generation manufacturing method and AM is using digital CAD data directed to the machine to manu-
Abstract
Additive
Additive
Abstract Manufacturing (AM)
Manufacturing (AM) is is aa new
new generation
generation manufacturing
manufacturing method
method and
and AMAM isis using
using digital
digital CAD
CAD data
data directed
directed to
to the
the machine
machine toto manu-
manu-
facture.
AdditiveAM is therefore (AM)
Manufacturing regarded
is aas a direct
new digitalmanufacturing
generation manufacturingmethod
method.and
ThisAMresearch
is usingwork presents
digital CAD thedatamethodology
directed to theformachine
designingto critical
manu-
facture.
facture. AM
AM is
is therefore
therefore regarded
regarded as
as aa direct
direct digital
digital manufacturing
manufacturing method.
method. This
This research
research work
work presents
presents the
the methodology
methodology for
for designing
designing critical
critical
Additive
aerospace Manufacturing
parts used under(AM) is
fatigue a new generation manufacturing method and AM is using digital CAD data directed to the machine to manu-
facture.
Additive
aerospace
aerospace
AM is therefore
Manufacturing
parts
parts used under
used (AM)
under is aasconditions
regarded
fatigue
fatigue newa direct
conditions
conditions
for AM.
digital
generation
for AM.
for AM.
Selected fatigue
manufacturing
manufacturing
Selected fatigue
Selected fatigue
critical
method.
method This
and
critical
critical
aerospace
AMresearch
aerospace
aerospace
part
is usingwork was
digital
part
part was
was
topologically
presents
CAD the
topologically
topologically
optimized
datamethodology
directed to the
optimized
optimized
then
for re-designed
designing
machine
then
then
for
to critical
manu-
re-designed
re-designed for
for
facture.
aerospaceAM is therefore
manufacturability.
parts With
used underregarded
thisfatigue asconditions
optimization a direct digital
study,
for45
AM.%manufacturing
mass saving
Selected method.
was
fatigue This
obtained
critical research
while
aerospace workwas
presents
mechanical
part the methodology
requirements
topologically were for designing
satisfied.
optimized then critical
Manufacturing
re-designed for
facture. AM is therefore
manufacturability.
manufacturability. With
With regarded
this
thisfatigue as
optimization
optimization a direct digital
study,
study, 45
45 %
%manufacturing
mass
mass saving
saving method.
was This
obtained
waswas
obtained research
while work presents
mechanical
while mechanical the methodology
requirements
requirements were for designing
satisfied.
were satisfied. critical
Manufacturing
Manufacturing
aerospace
simulationsparts
for
manufacturability.used under
thermal
With distortions
thisfatigue conditions
are
optimization covered for
study, AM.
and
45 the
% Selected
optimized
mass savingfatigue
part critical
waswas aerospace
manufactured
obtained part
with
while mechanicalwas
lasertopologically
powder bed
requirements optimized
fusion then
(L-PBF)
were satisfied. re-designed
and secondary
Manufacturing for
aerospace
simulations
simulationsparts
for used under
thermal
for thermal distortions conditions
are covered for AM.
and the Selected
optimized fatigue
part critical aerospace
manufactured part
with was
lasertopologically
powder bed optimized
fusion then
(L-PBF) re-designed
and secondaryfor
manufacturability.
operations
simulations were
for With distortions
applied.
thermal distortions
are covered
this optimization
are study,and
covered 45 %
and
themass
the
optimized
savingpart
optimized waswas
part was
manufactured
obtained
manufactured
with laser requirements
while mechanical
with laser
powder bed fusion
powder bed fusion
(L-PBF) Manufacturing
were satisfied.
(L-PBF)
and secondary
and secondary
manufacturability.
operations were
operations With this optimization study, 45 % mass saving was obtained while mechanical requirements were satisfied. Manufacturing
were applied.
applied.
simulations for thermal
operations were applied.distortions are covered and the optimized part was manufactured with laser powder bed fusion (L-PBF) and secondary
simulations for thermal distortions are covered and the optimized part was manufactured with laser powder bed fusion (L-PBF) and secondary
operations
© 2021 Thewere applied.
Authors. Published by Elsevier B.V.
operations
© 2021
© 2021 The
Thewere applied.
Authors.
Authors. Published by
Published by Elsevier
Elsevier B.V.
B.V.
©
© 2021
This is an
2021 The
open
The Authors. Published
accessPublished
Authors. article by by
under the Elsevier
B.V. B.V. license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
CC BY-NC-ND
Elsevier
This
This is
is an
an open
open access
access article
article under
under the
the CC
CC BY-NC-ND
BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
This
© isan
2021 an
The
Peer-review
This is open
under
open access
Authors. article
Published
responsibility
access article under
by
underofElsevierthe
the CC
the CC
B.V.BY-NC-ND
scientific
BY-NC-NDcommittee oflicense
the 10th(http://creativecommons.org/licenses/by-nc-nd/4.0/)
CIRP Sponsored Conference on Digital Enterprise Technologies (DET
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review
© 2021 The
Peer-review under responsibility
Authors.
under Published
responsibility ofElsevier
by
of the scientific
the scientific committee
B.V. committee of the
of the 10th
10th CIRP
CIRP Sponsored
Sponsored Conference
Conference on on Digital
Digital Enterprise
Enterprise Technologies
Technologies (DET
(DET
Peer-review
2021
This is
) an
–
Peer-review under
open
Digital
under responsibility
access article under
Technologies
responsibility as of
thethe
Enablers
of the CC scientific
BY-NC-ND
of committee
Industrial
scientific of of
thethe
Competitiveness
committee 10th
10th CIRP
and
CIRP Sponsored
Sustainability.
Sponsored Conference
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Conference on Digital
on Digital Enterprise
Enterprise Technologies
Technologies (DET
2021
This
2021 is
)
) –
an
– open
Digital
Digitalaccess article
Technologies
Technologies under
as
as the
Enablers
EnablersCC BY-NC-ND
of
of Industrial
Industrial license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Competitiveness
Competitiveness and
and Sustainability.
Sustainability.
(DET
2021 ) 2020)
Peer-review – Digital
under
– Digital
Peer-review under Technologies
responsibility
Technologies
responsibility
of
asof theasscientific
the
Enablers Enablers
scientific of Industrial
committee
of Industrial of Competitiveness
the 10th
Competitiveness
committee
CIRP and Sustainability.
Sponsored
and Sustainability.
of optimization;
the 10th CIRP Sponsored
Conference on Digital Enterprise Technologies
Conference on Digital Enterprise Technologies (DET
(DET
2021
Keywords: Additive
) – Digital
Keywords:
manufacturing;
Technologies as Selective
Enablerslaser melting; Topology
of Industrial Competitiveness andHolistic process; Fatigue critical
Sustainability.
2021
Keywords: Additive
) – Digital
Additive manufacturing;
Technologies
manufacturing; Selective
as Selective laser
Enablerslaser melting;
of Industrial Topology optimization;
Competitiveness
melting; Topology Holistic
andHolistic
optimization; process;
Sustainability. Fatigue critical
process; Fatigue critical
Keywords: Additive manufacturing; Selective laser melting; Topology optimization; Holistic process; Fatigue critical
Keywords: Additive manufacturing; Selective laser melting; Topology optimization; Holistic process; Fatigue critical
Keywords: Additive manufacturing; Selective laser melting; Topology optimization; Holistic process; Fatigue critical
1.
1. Introduction
Introduction proposed
proposed as as
as aaa seminal
seminal
seminal work work
work [1].[1]. Many
[1]. Many contributions
Many contributions
contributions such such
such asas
as
1. Introduction proposed
stress constraint and filters to deal with the numerical prob-
1. Introduction proposed
stress as
constrainta seminal
and work
filters to[1].
dealMany
with contributions
the numerical such as
prob-
1. The
Introduction stress
proposed constraint
as design and
a seminal filters
work to deal with the numerical prob-
1. The aviation
Introduction
aviation and
and space
space industries
industries are
are the
the most
most critical
critical sec-
sec- lems
stress
proposed
lems during
constraint
during as design andhave
a seminal filters
have been
work
beento[1].
made
deal
[1].
made
Many
Many to
with
to
contributions
improve
contributions
improve TO
the numerical
TO
such
formula-as
prob-
such
formula-as
The aviation and space industries(AM) are theis most critical sec- lems
stress
tions during design
constraint
using the andhave
initial filtersbeento made
methodology deal withto improve
[4, the In
15]. TOaerospace
numerical
the formula-
prob-
tors
torsThewhere
where additive
aviation and
additive manufacturing
space industries
manufacturing are
(AM) theis most
most
most potentially
critical sec-
potentially lems
stress
tions during design
constraint
using the and
initialhave
filtersbeento
methodology made
deal to
with
[4, improve
the
15]. TO
numerical
In the formula-
prob-
aerospace
tors where additive manufacturing (AM)
arearetheis mostcritical
potentially tions
lems using the initial methodology [4, 15]. In the TOaerospace
used
tors
used
The
The andaviation
where
and the andrevenues
total
theadditive
aviation
space industries
from
from AM
manufacturing
andrevenues
total space industries AM (AM)
arearetheis
most
expected to
mostcritical
most
expected
sec-
to gradu-
potentially
sec-
gradu- tions during
industry,
lems using
during
industry, where
where
design
the designs
initial
design
designs
have
have arebeen
arebeen
madeto
expected
methodology madeto
expected
tobeimprove
[4, lightweight
15].
tobeimprove In the
lightweight
formula-
and high
TOaerospace
formula-
and high
high
used
tors and
where the total
additive revenues from
manufacturing AM (AM)are expected
is most to gradu-
potentially industry,
tions
in using
strength, where
the
many designs
initial are expected
methodology
components have to
[4,
been be15].lightweight
In
designed the and
aerospace
with the fre-
ally
used
tors
ally increase
where
increase over
and theadditive
over the
totalthe next
revenues 20
manufacturing
next 20 years
from
years AM[8].
(AM)
[8]. Such
are
Suchis an
expected
most
an extraordinary
to gradu-
potentially
extraordinary industry,
tions
in where
using many
strength, designs
the initial are expected
methodology
components have to be
[4, 15].
been lightweight
In the
designed and high
aerospace
with the fre-
ally
used increase
and the over
total the next
revenues 20 years
from AM[8]. Such
are an
expectedextraordinary
to gradu- in strength,
industry,
quently usedmany
where components
designs
topology are have
expected
optimization been
to be designed
method, lightweight
such with
as the
and fre-
high
pylons,
trend
ally
used
trend of
and
of AM
increase is of
over
the istotal
AM of course
the next
revenues
course the
20
the AM
years
from
AM Technologies
AM[8]. Such an
are expected
Technologies is regarded
extraordinary
is to gradu-
regarded as
as in strength,
industry,used
quently many components
wheretopology
designs are have been
expected tomethod,
optimization designed
be lightweight
such with
as the
and fre-
high
pylons,
trend
ally of AM
increase is ofdigital
over course
the next the
20 AM yearsTechnologiesSuch an is
[8].(DDM). regarded
extraordinary as quently
in strength,
ribs, and usedmanytopology
brackets components
[20, optimization
7, 13]. have been
Besides, method,
designed
the suchwith
reason as pylons,
why the fre-
topol-
the
trend
ally
the novel
of AM
increase
novel direct
is of
over
direct course
the
digital manufacturing
next the
20 AM
years
manufacturing Technologies
[8]. Such
(DDM). isan DDM
is regardeddefines
extraordinary
DDM as
defines quently
in strength,
ribs, used
andused manytopology
brackets components
[20, 7, optimization
7, have
13]. Besides,been
Besides, method,
designed
the reason such
reason as
with
why pylons,
the fre-
topol-
the
trendnovel
of AMdirect
is ofdigital
course manufacturing
the AM (DDM).
Technologies DDMregardeddefines
as ribs,
quently
ogy and brackets
optimization topology
is[20,
used 13].
optimization
so frequently the
method,
is it can such why
offerasa topol-
pylons,
design
as
the the
trendnovel
as the
the de-centralized
of AMdirect
is of
de-centralized manufacturing
digital
course manufacturing
the
manufacturing AM of the
Technologies
of the
the parts
(DDM).
parts inin
DDM
is
in accordance
regardeddefines
accordance as ribs,
quently
ogy and brackets
used
optimization topology
is[20,
used 7, 13].
so Besides,
optimization
frequently the
method,
is it reason
can such why
offerasa topol-
pylons,
design
as
the novelde-centralized
direct manufacturing
digital manufacturing of parts
(DDM). DDM accordance
defines ogy
ribs, optimization
solutionand forbrackets is[20,
aerospace used 7, so
13].
parts frequently
Besides,
that are istheit reason
subjected can offer
to whya design
topol-
static, dy-
with
as
the
withtheproper
novel
proper qualification
de-centralized
direct digital
qualification and
and certification
manufacturing
manufacturing of the
certification [5].
partsAnother
(DDM).
[5]. in
DDM
Another reason
accordance
defines
reason ogy optimization
ribs,
solutionand forbrackets is[20,
aerospace used 7, so
13].
parts frequently
Besides,
that istheit reason
are subjected
subjected can offer
to whya design
topol-
static, dy-
with
as theproper qualification
de-centralized and
manufacturing certification
of the [5].
parts Another
in reason
accordance solution
ogy
namic, for
optimization
and aerospace
is used
thermo-elastic parts
so that
loads are
frequently
[17, is
14, it can
9]. In to
offerstatic,
a dy-
design
addition, the
behind
with
as the
behind the
proper extensive
de-centralized
the extensive usage
qualification and of
manufacturing
usage of AM
AM can
certification
of
can be
the
be correlated
[5].
parts Another
in
correlated with
accordance
with the
reason
the solution
ogy
namic, for
optimization
and aerospace
is used
thermo-elastic parts
so that
loads are
frequently
[17, subjected
is
14, it can
9]. In to
offerstatic,
a dy-
design
addition, the
behind
with the
proper extensive usage
qualification of AM can
andoftocertification be correlated
[5]. Another with the
reason namic,
solution
fact that and
for
the thermo-elastic
aerospace
difficulties partsloads
brought [17,
thatby 14,
arethe14, 9].
subjected
productionIn addition,
to the
static, dy-
method of
flexibility
behind
with the
proper
flexibility in design.
extensive Owing
usage
qualification
inextensive
design. Owing
Owing and the
AM manufacturing
can
certification
toAM be
the manufacturing correlated
[5].
manufacturing freedom
Another with
freedomreasonof
the
of namic,
solution
fact and
thatandfor thermo-elastic
aerospace
thethermo-elastic parts
difficulties brought loads
brought that [17,
byare 9].
subjected
the14, productionIn addition,
to methodthe
static, dy-
of
flexibility
behind the in design. usage ofto the can be correlated freedom
with of
the fact
namic,
the that
designedthe difficulties
part (i.e. as loads
draw by
[17,the
direction, production
9].
symmetryIn method
addition,
plane, of
the
and
the AM
flexibility
behind
the AM the processes,
in design.
extensive
processes, more complex
Owing
usage
more to
of AM
complex thegeometries
manufacturinghaving
can be correlated
geometries lighter
freedom
havingfreedom and
with and
lighter of
the fact
namic,
the that the difficulties
and thermo-elastic
designed part (i.e.
(i.e. as brought
loads
asbrought
draw by the production
[17, 14,symmetry
direction, method
9]. In addition, of
the
plane, and
and
the AM
flexibility processes,
in design. more complex
Owing to thegeometries
manufacturinghaving lighter and
of the
fact designed
that
overhang the part
difficulties
angle) can be draw
integrateddirection,
by the
into symmetry
production
the plane,
method
optimization of
pro-
stiffer
the AM properties
flexibility
stiffer processes,
in can
design.
properties can be
more designed
complex
Owing
be complex to the and
designed geometries manufactured.
geometries
manufacturinghaving
and manufactured.
manufactured. In
freedomdesign
lighter and
In design
designof the
fact designed
that
overhang the part
angle) (i.e.
difficulties
can as
be draw
brought
integrateddirection,
by the
into the symmetry
production
the plane,
method
optimization and
of
pro-
stiffer
the AM properties
processes, can be
more designed and having In
lighter and overhang
the
cess designed
as angle)
part can
(i.e. be
as integrated
draw into
direction, optimization
symmetry plane, pro-
and
for
for
for
AM,
stiffer
the AM,
AM,
stiffer
topology
AMproperties
topologycan
processes,
topology
properties
optimization
be complex
more designed
optimization
optimization
can be designed
(TO)
(TO)
(TO)
andmethod
andmethod
method
has
manufactured.
geometries has
has
manufactured.
been
having preferred
In
beenlighter
been
design
and
preferred
preferred
the
cess
cess as aaa constraint
overhang
designed
as
overhang
constraint
angle)
partcan
constraint
angle) can
improves
(i.e. be
improves
improves
drawthe
as integrated
the
be integrated
design
into processes
thedirection,
design
design
into
the
processes
processes
[18,
optimization
symmetry [18,
[18,
the optimization
19].
19].
Fa-
19].pro-
plane, and
Fa-
Fa-
pro-
due
for to
AM,
stiffer
due to its significant
topology
properties
its can
significant role in
optimization weight
be designed
role in weight(TO) reduction.
andmethod has been In
manufactured.
reduction.
design
preferred
In design tigue
cess asis
overhang
tigue is aconsidered
constraint
angle) can
considered as
asbeaa integrated
improves constraint
constraint during
the design
into
during thethe
processes design
optimization
the design to
[18, 19].
to meet
Fa-
pro-
meet
due
for to
AM, its significant
topology role in
optimization weight(TO) reduction.
method has been preferred tigue
cess
all as
the is aconsidered
constraint
requirements as a
improves
since constraint
the the during
design
structural the
processes
parts useddesign
[18,
in to meet
19]. Fa-
aerospace
dueTopology
for to
AM, its
Topology optimization
significant
topology role in
optimization
optimization is a mathematical
weight
isweight(TO) reduction.
a mathematicalmethod
mathematical method
has been
method that that deter-
preferred
that deter-
deter- tigue
cess
all asisa considered
constraint
theisrequirements
requirements as
sincea
improves constraint
the during
design
the structural
structural the
processes
partsthe useddesign
[18, to meet
19].
in aerospaceFa-
aerospace
due Topology
to its optimization
significant role inis a reduction. method all
tiguethe
applications considered
are since
as
commonly a the
constraint parts
during used in
design to meet
mines
due
mines to the
Topology optimum
its significant
the optimum material
optimization
material adistribution
role inisweightmathematical under
reduction.
distribution under defined
method
definedthatloading
deter-
loading all theisrequirements
tigue
applications considered
are commonlysince
as
commonly theunder
a constraint
under cyclic
structural
during
cyclic loads
parts the
loads in
used service.
design
in Dur-
in aerospace
to meet
service. Dur-
mines the
Topology optimum material
optimization is adistribution
mathematical under defined
method loading
that deter- applications
all the
ing the are
requirements
design since theunder cyclic
structural loads
parts in
used service. Dur-
in aerospace
and
mines
and boundary
the optimum
Topology
boundary conditions
material
optimization
conditions within
within aa specified
is adistribution
mathematical under
specified design
defined
method
design space
that
space [2].
loading
deter-
[2]. applications
all
ing the
the areof
requirements
design of aa fatigue
commonly
fatigue critical
since theunder part,
cyclic
structural
critical part, crack
loads
parts
crack usedstates
in of
service. initia-
Dur-
in aerospace
states of initia-
and
mines boundary
the conditions
optimum materialwithin a
distributionspecified
under design
definedspace [2].
loading ing
tion the
applicationsdesign
caused are
by of a
shear fatigue
commonly
stress critical
under
and part,
cyclic
propagation crack
loads states
in
caused of
service.
by initia-
Dur-
normal
Initial
and
mines
Initial TO
boundary
the formulation
TO optimum conditions
formulation that
material enables
within a
distribution
thatwithin
enables to find
specified
to find
find optimum
under design
optimumdefinedshape
space and
[2].
loading
shape and ing
tion the
applicationsdesign
caused are
by ofof a fatigue
commonly
shear stress and critical
under
and part,
cyclic
propagation crack
loads states
in
caused of
service. initia-
Dur-
byofnormal
normal
Initial
and TO
boundary formulation
conditions that enables a to
specified optimum
design shape
space and
[2]. tion
ing
stress caused
the design
should by sheara stress
fatigue propagation
critical part, crack caused
states by initia-
material
Initial
and TOdistribution
boundary
material formulation during
conditions
distribution during the
the design
thatwithin
enables of
of aaaoptimum
to find
a specified
design structural part
part was
shape
design space
structural and
[2].
was tion the
ing
stress caused
design
should bybe
beofinvestigated
shear stress and
a fatigue
investigated and
criticalthe critical
andpropagation
thepart, crack
critical stress
caused
states
stress levels
levels must
byofnormal
initia-
must
material
Initial TOdistribution
formulation during the
that enables design of
to find structural part was stress
tionlower should
caused bybe investigated
shear stress andand the critical
propagation stress
caused levels must
by normal
material
Initial TOdistribution
formulation during the design
that enables of aoptimum
to find structuralshape
optimum
and
part was
shape and be
stress
tion
be caused
lower than
should
thanbybeallowable
investigated
shear
allowable stressstresses.
and
stresses. Constraints
andpropagation
the
Constraints of
critical caused
stress
of high
high bycycle
levels fa-
must
normal
cycle fa-
material
2351-9789
2351-9789 distribution
© ©2021
2021TheThe during
Authors.
Authors. the design
Published by
Published of a
Elsevier
by structural
B.V.
Elsevier part
B.V. was be lower
stress than
should allowable
be investigated stresses.
and Constraints
the critical of
stresshigh cycle
levels fa-
must
2351-9789
material
2351-9789
This is an
© 2021
© 2021
open
The
distribution Authors.
during
Thearticle
access Authors.
under
Published
the design
Published
the CC
by Elsevier
by Elsevier
BY-NC-ND
B.V.
of a structural
B.V. part was be lower than allowable
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
stresses. Constraints
stress should be investigated and the critical stress levels must of high cycle fa-
This
This isisan
2351-9789 an©open
open2021 access article
Thearticle
access Authors.
under under
Published
the CC theElsevier
by CC BY-NC-ND
BY-NC-ND B.V. be lower than allowable stresses. Constraints of high cycle fa-
license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
This is
2351-9789an
Peer-review
Peer-review
This
open
is an open
access
©under
2021
under
article under
responsibility
Theresponsibility
access Authors.
article under
the
ofthe CC
theof
Published
CC
BY-NC-ND
scientific
by scientific
the Elsevier
BY-NC-ND B.V.
committee CIRP10th
of the
be lower
Sponsored than
CIRP Conference
allowable
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
committee of the 10th SponsoredonConference
license(http://creativecommons.org/licenses/by-nc-nd/4.0/) stresses.
Digital Enterprise Constraints
Technologies
on Digital Enterprise
of high cycle
(DETTechnologies fa-
2021)–Digital
Peer-review
2351-9789
Peer-review under
©under responsibility
2021 responsibility of the
the scientific
The Authors. Published scientific committee
by Elsevier B.V. ofof the
the 10th
10th CIRP
CIRP Sponsored
Sponsored Conference
Conference on
on Digital
Digital Enterprise
Enterprise Technologies (DET 2021)–Digital
2021)–Digital
This
(DET is an open
Technologies
2020)
Peer-review as–access
under article
Enablers
Digital underofthe
ofTechnologies
Industrial
responsibility CC BY-NC-NDcommittee
theCompetitiveness
as Enablers license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
andofSustainability.
Industrial
of the 10th Competitiveness and Sustainability.
Technologies (DET
2021)–Digital
Technologies
This is an as
open
Technologies Enablers
asaccess of
article
Enablers underofthe
of Industrial
Industrial
scientific
CC BY-NC-NDcommittee
Competitiveness
Competitiveness and
and Sustainability.
CIRP Sponsored Conference on Digital Enterprise
license(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Sustainability.
Technologies (DET
Peer-review under responsibility
10.1016/j.promfg.2021.07.037
Technologies as Enablers of the
of Industrial scientific committee of the 10th CIRP Sponsored Conference on Digital Enterprise Technologies (DET 2021)–Digital
Peer-review under responsibility of theCompetitiveness and Sustainability.
scientific committee of the 10th CIRP Sponsored Conference on Digital Enterprise Technologies (DET 2021)–Digital
Technologies as Enablers of Industrial Competitiveness and Sustainability.
Technologies as Enablers of Industrial Competitiveness and Sustainability.
 Akin Dagkolu et al. / Procedia Manufacturing 54 (2021) 238–243
A. Dagkolu et al. / Procedia Manufacturing 00 (2021) 000–000
tigue(HCF) stress and static stress aggregated by modified p-
norm were implemented to the TO problem formulation with
the objective function of minimizing mass used for the con-
ceptual design process of the lightweight structures [6]. Static
analysis instead of costly dynamics analysis was performed to
determine fatigue behavior using the modified-Goodman crite-
rion. Relaxed equivalent stresses calculated with the Sines ap-
proach and the compliance constraint were integrated into the
TO formulation [3]. The method which works fatigue damage
for HCF was proven the applicability to the additively manu-
factured parts by numerical examples solved using two alloy
steels [16]. In this study, a methodology for the design of a
fatigue-critical aerospace part for AM is presented. The pro-
posed methodology aiming to improve the fatigue behavior and
stiffness per mass ratio includes modeling for additive manu-
facturing using TO, re-design, and analytical calculations using
results of Finite Element (FE) analyses to determine the life
cycles for the cyclic loadings containing HCF and low cycle
fatigue (LCF).
2. Materials and Methods
2.1. Materials
The sample part was originally manufactured from annealed
Ti-6Al-4V material with computer numerical control (CNC)
machining process. The part is under variable loading in service
for an aerospace application. To be able to use similar material
properties, gas atomized Ti-6Al-4V (grade-5) powder is used in
the additive manufacturing of the optimized part. The chemical
composition of the Ti-6Al-4V powder supplied by the manu-
facturer is given in Table 1. This Ti alloy is frequently used
in aerospace applications where high mechanical strength and
fatigue resistance is needed. In this alloy, aluminum acts as α
stabilizer and vanadium as β stabilizer [11].
Table 1. Chemical composition of the Ti-6Al-4V powder
Element Al V O N C H Fe
5,5-6,75 3,5-4,5 2000 500 800 150 3000
wt.-% wt.-% ppm ppm ppm ppm ppm
2.2. Additive Manufacturing
All the samples and the optimized part were manufactured
by laser powder bed fusion process (L-PBF) and using the
EOSINT M280 machine. Manufacturing parameters were se-
lected as the default parameter set for Ti-6Al-4V alloy supplied
by the machine manufacturer and they are given in Table 2. The
selected parameter set is held constant through all the works.
Table 2. L-PBF Manufacturing Parameters
Laser Power Scanning Speed Hatch Distance Layer Thickness
(W) (mm/s) (mm) (mm)
170 1250 0,015 0,03
239
2
Mechanical properties of the used Ti-6Al-4V alloy powder
and the machine performance were validated after manufactur-
ing of test samples and testing them according to the standards
such as ASTM E8, E23, E466.
2.3. Post Processes
All of the test samples are manufactured within the same
batch simultaneously and heat-treated with identical conditions.
Heat treatment parameters are supplied by the machine manu-
facturer as holding at 800°C 2 hours in an argon atmosphere and
followed by furnace cooling to room temperature. This stress
relief heat treatment was applied because of the thermal resid-
ual stress-induced deformations during cutting the part from the
build platform [10]. Test specimens were then CNC machined
to comply with the corresponding test standards. The optimized
part and the witness fatigue test samples were heat-treated with
same parameters then Hot Isostatically Pressing (HIP) was con-
ducted for the qualification requirements at 900°C and 1000
Bar. As-built surface properties of L-PBF manufactured part
is considered insufficient for a fatigue critical part. To im-
prove the surface properties of the part, chemical and chemi-
cal/mechanical polishing processes were applied. After the sur-
face finishing process, the assembly contact interfaces of the
part were machined to achieve the required geometrical toler-
ances.
2.4. Design and Analysis
The flowchart of design, optimization and analysis of the
structurally optimized part is presented in Figure 1. Before TO,
design and non-design spaces are fully defined. The design vol-
ume has been chosen as inclusively as possible to reach the op-
timum material distribution. Non-design spaces include the as-
sembly interfaces of the application part. And during the model-
ing the non-design spaces should consider tooling accessibility
and the interaction with the other sub-system components. In
the optimization process, the prepared design volume was first
subjected to finite element analysis according to the loading
condition that constitutes the optimization boundary condition.
As the solution principle of the Solid Isotropic Material with
Penalization (SIMP) method used in topology optimization, an
artificial density value was assigned to each finite element and
it was decided whether the elements should remain in the de-
Fig. 1. Design - Optimization Flowchart
240 Akin Dagkolu et al. / Procedia Manufacturing 54 (2021) 238–243
A. Dagkolu et al. / Procedia Manufacturing 00 (2021) 000–000
sign volume according to the load exposure acquired from the
finite element analysis results.
Topology optimization problem used in the design of the
aerospace part is :
Find [ρ1 , ρ2 , ..., ρN ]T
Minimize C(ρ) = uT K(ρ)u
S ub ject to K(ρ)u = f
N (1)
i=1 ρi vi − Ṽ ≤ 0
0 ≤ MoS
0 ≤ ρm in ≤ ρ ≤ 1
where ρ is the artificial density vector containing all design
variables, C is the compliance of the structure, K is the global
stiffness matrix, u is the displacement vector, f is the external
force vector, vi is the volume of each finite element, N is the to-
tal number of elements, Ṽ is the prescribed volume limit value,
MoS is the margin of safety and ρmin is a number close to zero
to avoid the singularity problem.
The Margin of Safety (MoS) is calculated to ensure the static
requirement for the part to be optimized as follows:
σallowable
MoS = −1
σmax x S F
where σallowable is the yield stress of the material, σmax is the
von-Mises stress result obtained from FEA and SF is the safety
factor that covers AM uncertainties and the deviation based on
FEM. The safety factor (SF) is chosen as 2 because the selected
additive manufacturing method for the production of the part
has higher uncertainties than the conventional manufacturing
methods [12].
The resulting isosurface geometry has been remodeled in
a CAD environment as a Non-uniform rational B-splines
(NURB) surface for validation analysis and manufacturability
requirements. The remodeled optimized part was subjected to
validation analysis and it was checked if the final geometry
meets the mechanical strength criteria. After static analyses, fa-
tigue life calculations were made by analytical method against
fatigue loads of the optimized part, which are among the other
critical loading conditions. And as the last effort before L-PBF,
manufacturing simulations of the final part geometry were car-
ried out with thermo-mechanical analysis, and possible distor-
tion/residual stress conditions were investigated.
3. Results and Discussions
3.1. Topology Optimization
Topology optimization of the selected aircraft part is con-
ducted to obtain a design with maximum available stiffness
and 40 % weight reduction of the original part. To start opti-
mization work, the part geometry was split into 2 volumes as
design space where topology optimization was performed and
non-design space as the unused/kept volume of the part dur-
ing optimization. These volumes are given in Figure 2. Loading
conditions of the fatigue critical part are given as static with pre-
loading and cyclic loads and all loading conditions are shared
in Table 3. Since both static and fatigue analysis could not be
3
Fig. 2. Topology optimization setup including design space (red region) and
non-design space (gray regions)
solved at the same time while topology optimization, -8900 N
which is the static load and the maximum loading condition
was used during the optimization. In optimization setup, design
space was discretized with 238751 finite elements and as seen
in Figure 2, it was fixed from connector holes to be analyzed
as hard-mounted. Static load was applied to the corresponding
region of the structure with RBE3 elements.
As a result of the performed topology optimization in the
(2)
Altair Optistruct software, normalized objective function his-
tory decreased as 80 % of the initial value of the design space
in 17 iterations. According to von-Mises stress results of the fi-
nal iteration, the MoS value was calculated as nearly 12 using
Equation 2 to check the static requirements of the structure and
it was found in the feasible region. When the volume constraint
function to be needed to satisfy 80 % mass saving of the de-
sign space was investigated, it is determined to be found in the
feasible region of the optimization problem. To investigate the
optimization result and supply reference geometry for the re-
design phase, the elements with the design variable below 0.5
were removed from the final design and the reference model
was exported as a .stl file (Figure 3).
3.2. Redesign of the Topology Optimization Result
For the manufacturing requirements and validation analyses
of the parts designed with topology optimization, the isosur-
face geometries resulting from the optimization were remod-
eled as NURB surfaces. This remodeling of the optimized part
was performed in a CAD environment with a generative shape
design module by taking the stl geometry exported from FEM
software. NURB modeling of the isosurface geometry provided
more accurate results from the validation analyses and at the
same time, it was ensured that the necessary adjustments such
Table 3. Loading conditions of the optimization part
Static Loading Condition
Load Case Load (N)
Static Limit -8900
Cyclic Loading Condition
Load Case Max Load (N) Min. Load (N)
HCF -2234 -335
LCF -3760 1916
 Akin Dagkolu et al. / Procedia Manufacturing 54 (2021) 238–243
A. Dagkolu et al. / Procedia Manufacturing 00 (2021) 000–000
Fig. 3. Topology optimization result
as size change in the part geometry, minor changes in the part
feature/section can be made parametrically after the validation
analyses. Remodeling of the optimized part and the final manu-
facturing model is given in Figure 4. Topology optimization and
the redesign efforts are considered nearly equal in engineering
time and resource usage because in additive manufacturing and
conventional manufacturing workflow since the only variable is
the manufacturing constraints.
3.3. Validation Analyses
The re-modeled CAD geometries were taken into the FEM
environment and analyzed with the identical loading and
boundary conditions with the topology optimization setup, and
then analytical fatigue life calculations were made for the fa-
tigue loading, which is one of the critical loading conditions of
the part. The von-Misses stress criterion was used in the static
strength calculation of the part. In other words, the von-Misses
stresses in the finite elements used in the analysis are expected
to remain below the yield strength of the material. The static fi-
nite element analysis setup of the optimized and redesigned part
Fig. 4. Final manufacturing design
241
4
Fig. 5. Static analysis results of optimized part
and the stress and displacement result from the static analysis
are given in Figure 5.
In addition to that, it is expected to have an endurance limit
greater than 107 cycles because the optimized part is a fatigue
critical part. For fatigue loads, the optimized part has been sub-
jected to static finite element analysis with loading values at
the lower and upper boundaries of the high cycle and low cy-
cle fatigue loads. Static analyses with fatigue loads were per-
formed under the same boundary conditions as topology op-
timization and static validation analysis setups. Critical stress
concentration regions for fatigue damage are determined by ex-
perience based on an engineering approach according to load
flow conditions. Stress concentration regions where the life cal-
culations were made are given in Figure 6. Maximum principal
and signed von-Misses stress data is read from the FE analyses
result of the defined stress concentration regions and finite life
calculations which are made according to the Soderberg Dam-
age Criterion. In the analytical calculation of the fatigue life of
the parts, there are some ”endurance limit modifying factors”
and these factors define the part by its material, manufactur-
ing, environment, and design specifications. These modifying
factors used in the fatigue life calculation are seen in Table 4.
Fatigue endurance limit calculation of the optimized part was
completed with aforementioned engineering assumptions and
fatigue life results are given in Table 5 for each load type and
notch region.
3.4. Manufacturing Simulations
After validating the structural properties of the optimized
part, L-PBF manufacturing simulations were carried out. These
simulations run a thermo-mechanical analysis in the back-
Table 4. Endurance limit modifying factors
Endurance Limit Modifying Factors Values
Surface Factor, ka 0.72
Size Factor, kb 0.90
Loading Factor, kc 1.00
Temperature Factor, kd 1.00
Reliability Factor, ke 0.81
Miscellaneous-Effects Factor, k f 1 (Region1) 0.68
Miscellaneous-Effects Factor, k f 2 (Region2) 0.79
Miscellaneous-Effects Factor, k f 3 (Region3) 0.80
Notch Sensitivity, q 0.38
242 Akin Dagkolu et al. / Procedia Manufacturing 54 (2021) 238–243
A. Dagkolu et al. / Procedia Manufacturing 00 (2021) 000–000 5

Table 5. Endurance limit calculations of the optimized part

Loading # of cycles # of cycles # of cycles
Type Region #1 Region #2 Region #3
HCF 2.5x1019 1.3x1017 2.1x1015
LCF 5.1x1018 7.3x1012 1.3x1011

Fig. 8. Generation of manufacturing model

From a time consumption and cost point of view; manufac-

turing simulations seem like an additional step in the process
since in conventional manufacturing of the part, no manufac-
turing simulation work was conducted. But in the conventional
method, the work part is being manufactured with computer nu-
merical control(CNC) machining. Because of that a computer-
Fig. 6. Critical notch regions for fatigue load
aided manufacturing(CAM) effort is necessary to create tool-
paths and machine algorithms. Moreover, this process takes
nearly the same amount of engineering time as the simulation.
ground, modeling the L-PBF process with the finite element
3.5. Additive Manufacturing of the Part
method. For calculation of the thermal distortion and residual
stress, some assumptions and simplifications were to be made
Manufacturing of the optimized part was accomplished with
to model the problem as accurately as possible while keeping
the EOSINT M280 L-PBF machine with default processing
the calculation time and cost relatively low. The predicted heat
parameters. Insight of the thermo-mechanical process simula-
loss method was used on modeling the heat transfer problem
tion results of the optimized part, build orientation, and sup-
and lattice support structures were homogenized as 3D block
port structure generation efforts were made to achieve success-
elements to lower the uncertainty in the model. Build platform
ful manufacturing without thermal distortion and cracks. Before
temperature kept constant at 200 °C and it was modeled as
the manufacturing phase of the optimized part, to comply with
an undeformable component. Thermo-mechanical simulations
the post-processing requirements, surface tolerances were ap-
predicted 0,8 mm thermal deformation on the part (see Figure
plied to outer part surfaces and assembly interfaces for chemical
7). Machining tolerances were expanded to compensate for de-
and mechanical polishing and also for machining. Integration
formation effects based on the simulation results. Solid metal
of support structures and the changes applied for post-process
supports were applied on the region where maximum distortion
efforts are shown in Figure 8.
was expected (See Figure 8).
After L-PBF manufacturing, the part was heated for thermal
In these simulations, Autodesk Netfabb software was used
stress relieving then cut from the manufacturing table with a
as the thermo-mechanical solver and the solver was validated
wire EDM process. The manufactured part is shown in Figure
experimentally with residual stress measurements taken from
9, already connected to the building table.
L-PBF manufactured samples, with XRD and hole-drilling
After cleaning the support structures and removing the ex-
methods. But this experimental validation process kept out of
cess powder, part and the witness samples were HIPed for a
the scope of this article.
fully dense part. The HIP process is applied at 900 °C and 1000
Bar. The aforementioned surface post-processes were applied
after the HIP process. Firstly optimized part was subjected to
chemical polishing followed by chemical-mechanical polish-
ing. Chemical and mechanical polishing parameters are not pre-
sented because they were not shared by the process vendor.
The original part selected in this study is currently being
used in manned aircraft platforms and should comply with the
aviation the qualification standards. For that reason, fatigue and
static loading tests at qualification level will be conducted on
the optimized part with the same configuration used in numeri-
cal analyses.

4. Conclusions

In this study, the design of a fatigue-critical aerospace part

Fig. 7. Displacement results of thermo-mechanical simulation of optimized part for additive manufacturing with topology optimization is pre-
 Akin Dagkolu et al. / Procedia Manufacturing 54 (2021) 238–243
A. Dagkolu et al. / Procedia Manufacturing 00 (2021) 000–000
Fig. 9. L-PBF Manufactured optimized part
sented. The optimized part was designed 45 % lighter in mass
and the mechanical strength requirements of the optimized part
were validated by finite element analyses and analytical meth-
ods. It was ensured that the part would be manufactured without
thermal distortion and in accordance with the subsequent post-
processes to be applied with the manufacturing process simula-
tions. The optimized part was successfully manufactured with
the L-PBF method and after being subjected to heat treatment,
HIP, chemical, and mechanical surface treatments, it was made
ready for real condition tests. In summary, the process of devel-
oping a fatigue-critical aerospace part with a ”design for addi-
tive manufacturing” point of view has been proposed compre-
hensively with a holistic approach that can be used as a guide
for AM design-manufacturing-validation workflows.
Acknowledgements
The authors thank the support of Turkish Aerospace under
grant project DKTM 2018/05 ongoing with collaboration be-
tween Gazi University and Turkish Aerospace.
References
[1] Bendsøe, M.P., Kikuchi, N., 1988. Generating optimal topologies in
structural design using a homogenization method. Computer Methods
in Applied Mechanics and Engineering 71, 197–224. doi:10.1016/
0045-7825(88)90086-2.
[2] Bendsøe, M.P., Sigmund, O., 2004. Extensions and applications. Springer
Berlin Heidelberg, Berlin, Heidelberg. pp. 71–158. doi:10.1007/
978-3-662-05086-6_2.
[3] Collet, M., Bruggi, M., Duysinx, P., 2017. Topology optimization for min-
imum weight with compliance and simplified nominal stress constraints
for fatigue resistance. Structural and Multidisciplinary Optimization 55.
doi:10.1007/s00158-016-1510-6.
[4] Duysinx, P., Van Miegroet, L., Lemaire, E., Brüls, O., Bruyneel, M., 2008.
Topology and generalized shape optimization: Why stress constraints are
so important? International Journal for Simulation and Multidisciplinary
Design Optimization 2(4) 2. doi:10.1051/ijsmdo/2008034.
[5] Frazier, W., 2014. Metal additive manufacturing: A review. Jour-
nal of Materials Engineering and Performance 23. doi:10.1007/
s11665-014-0958-z.
243
6
[6] Holmberg, E., Torstenfelt, B., Klarbring, A., 2014. Fatigue constrained
topology optimization. Structural and Multidisciplinary Optimization 50.
doi:10.1007/s00158-014-1054-6.
[7] Krog, L., Tucker, A., Rollema, G., 2011. Application of topology, sizing
and shape optimization methods to optimal design of aircraft components.
[8] Kumar, J., Nair, C., 2017. Current trends of additive manufacturing in the
aerospace industry , 39–54doi:10.1007/978-981-10-0812-2_4.
[9] Liu, H., Zhang, W., Gao, T., 2015. A comparative study of dynamic analy-
sis methods for structural topology optimization under harmonic force ex-
citations. Structural and Multidisciplinary Optimization 51. doi:10.1007/
s00158-014-1218-4.
[10] Mercelis, P., Kruth, J.P., 2006. Residual stresses in selective laser sintering
and selective laser melting. Rapid Prototyping Journal 12. doi:10.1108/
13552540610707013.
[11] Mierzejewska, , Hudak, R., Sidun, J., 2019. Materials mechanical proper-
ties and microstructure of dmls ti6al4v alloy dedicated to biomedical ap-
plications. Materials 12. doi:10.3390/ma12010176.
[12] Orme, M., Madera, I., Gschweitl, M., Ferrari, M., 2018. Topology opti-
mization for additive manufacturing as an enabler for light weight flight
hardware. Designs 2, 51. doi:10.3390/designs2040051.
[13] Remouchamps, A., Bruyneel, M., Fleury, C., Grihon, S., 2011. Application
of a bi-level scheme including topology optimization to the design of an
aircraft pylon. Structural and Multidisciplinary Optimization 44. doi:10.
1007/s00158-011-0682-3.
[14] Shi, G., Guan, C., Quan, D., Wu, D., Tang, L., Gao, T., 2019. An aerospace
bracket designed by thermo-elastic topology optimization and manufac-
tured by additive manufacturing. Chinese Journal of Aeronautics 33.
doi:10.1016/j.cja.2019.09.006.
[15] Sigmund, O., 2007. Morphology-based black and white filters for topology
optimization. Structural and Multidisciplinary Optimization 33, 401–424.
doi:10.1007/s00158-006-0087-x.
[16] Suresh, S., Lindström, S., Thore, C.J., Klarbring, A., 2021. Topology op-
timization for transversely isotropic materials with high-cycle fatigue as
a constraint. Structural and Multidisciplinary Optimization 63. doi:10.
1007/s00158-020-02677-2.
[17] Trivers, N., Carrick, C., Kim, I.Y., 2020. Design optimization of a busi-
ness aircraft seat considering static and dynamic certification loading
and manufacturability. Structural and Multidisciplinary Optimization 62.
doi:10.1007/s00158-020-02650-z.
[18] Vatanabe, S., Lippi, T., R. de Lima, C., Paulino, G., Silva, E., 2016. Topol-
ogy optimization with manufacturing constraints: A unified projection-
based approach. Advances in Engineering Software 100, 97–112. doi:10.
1016/j.advengsoft.2016.07.002.
[19] Xiong, Y., Yao, S., Zhao, Z.L., Xie, Y., 2019. A new approach to eliminat-
ing enclosed voids in topology optimization for additive manufacturing 32,
101006. doi:10.1016/j.addma.2019.101006.
[20] Zhu, J.H., Zhang, W.H., Xia, L., 2016. Topology optimization in aircraft
and aerospace structures design. Archives of Computational Methods in
Engineering 23, 595–622. doi:10.1007/s11831-015-9151-2.