materials

Article
Inverse Method to Determine Fatigue Properties of
Materials by Combining Cyclic Indentation and
Numerical Simulation
Hafiz Muhammad Sajjad 1, * , Hamad ul Hassan 1 , Matthias Kuntz 2 ,
Benjamin J. Schäfer 1,2 , Petra Sonnweber-Ribic 2 and Alexander Hartmaier 1
1 Interdisciplinary Centre for Advanced Material Simulation (ICAMS), Ruhr-Universität Bochum,
Universitätsstr 150, 44801 Bochum, Germany; [email protected] (H.u.H.);
[email protected] (B.J.S.); [email protected] (A.H.)
2 Robert Bosch GmbH—Corporate Sector Research and Advance Engineering, 71272 Renningen, Germany;
[email protected] (M.K.); [email protected] (P.S.-R.)
* Correspondence: [email protected]; Tel.: +49-234-32-22-443




Received: 3 June 2020; Accepted: 8 July 2020; Published: 13 July 2020


Abstract: The application of instrumented indentation to assess material properties like Young’s
modulus and microhardness has become a standard method. In recent developments, indentation
experiments and simulations have been combined to inverse methods, from which further material
parameters such as yield strength, work hardening rate, and tensile strength can be determined.
In this work, an inverse method is introduced by which material parameters for cyclic plasticity,
i.e., kinematic hardening parameters, can be determined. To accomplish this, cyclic Vickers indentation
experiments are combined with finite element simulations of the indentation with unknown material
properties, which are then determined by inverse analysis. To validate the proposed method,
these parameters are subsequently applied to predict the uniaxial stress–strain response of a material
with success. The method has been validated successfully for a quenched and tempered martensitic
steel and for technically pure copper, where an excellent agreement between measured and predicted
cyclic stress–strain curves has been achieved. Hence, the proposed inverse method based on cyclic
nanoindentation, as a quasi-nondestructive method, could complement or even substitute the
resource-intensive conventional fatigue testing in the future for some applications.

Keywords: cyclic indentation; Vickers hardness; inverse analysis; numerical simulations; cyclic
material properties; fatigue life

1. Introduction
Depth-sensing indentations or instrumented indentations are very useful means to characterize
and determine mechanical properties (i.e., Young’s modulus and hardness) of thin films as well as of
bulk materials [1–5]. Hyung [6] and Suresch et al. [7,8] have proposed two novel methods to identify
the elastic modulus, yield strength, and the hardening exponent through nano-indentation. In addition,
instrumented indentation experiments make it possible to determine further material properties,
such as the strain hardening coefficient and yield strength [9]. Schmaling and Hartmaier [10] have
introduced a method to identify plastic material properties (i.e., yield strength and work hardening rate)
by using an inverse analysis for the remaining imprint after indentation. A comprehensive comparison
of the hardness measurement approaches at diverse scales (i.e., nano, micro, and macro) of Brinell,
Vickers, Meyer, Rockwell, Shore, IHRD, Knoop, and Buchholz was performed by Broitman [11]. He has
not only described each indentation method but has also presented its inadequacies in evaluating
results. Furthermore, he has discussed the effects of elasticity, plasticity, pileup, sink-in, grain size,

Materials 2020, 13, 3126; doi:10.3390/ma13143126 www.mdpi.com/journal/materials

Materials 2020, 13, 3126 2 of 14

and indentation size on determining hardness by means of depth-sensing indentation techniques at

the micro- and nanoscale.
The applications of instrumented nanoindentation are not limited to the identification of
conventional material properties, such as elastic modulus and hardness; the ease of performing
this method has also attracted the attention of some authors who have employed it to estimate complex
material properties like fatigue life or even the famous S-N (Wöhler) curve [12–14]. The determination of
these quantities requires tedious and difficult conventional fatigue experiments; consequently, there are
some analytical relationships available that allow the determination of the S-N curve and fatigue strength
by using indentation hardness. For example, the Strzelecki model [12] has described the relationship
between fatigue life and material hardness. Similarly, by utilizing the Murakami [13] formulation,
Bandara [14] has presented full range S-N curves for six different medium steel grades. However,
this suggested formulation is material-specific and requires two input parameters, i.e., Vickers hardness
and ultimate tensile strength. Furthermore, Bandara [15] has used the Brinell hardness value in order
to determine material fatigue strength. The three [12,14,15] aforementioned analytical approaches
(mainly dependent on the hardness value of the material) have been comprehensively summarized
by Strzelecki and Tomaszewski in [16], where they have expounded the merits and demerits of each
individual approach and concluded that both models are good for predicting tensile strength and yield
strength, while the Strzelecki model [12] is superior in the prediction of fatigue life as compared to the
Bandara model [14,15].
In classical indentation experiments, only single loading and unloading on the specimen is
performed for determining the desired mechanical properties. However, to study the fatigue life in
materials, the cyclic indentation has attracted the attention of many authors [17–20]. For example,
Lyamkin et al. [17] have shown the potential of cyclic indentation for studying the fatigue properties.
As the cycles of indentations are repeated at the same location, the elastoplastic area of high cyclic
indentation demonstrates fatigue under these conditions. For instance, a nanoimpact fatigue test has
been studied by Faisal et al. [18] by using cyclic indentation, and they have concluded that the indenter
geometrical shape (Berkovich or conical) and the indentation loading history are crucial in inducing
film failure. Moreover, Haghshenas et al. [19] have employed cyclic nanoindentation in order to
determine the indentation size effects and the strain rate sensitivity for tantalum. Similarly, Prakash [20]
has demonstrated fatigue damage in materials with the help of two different nondestructive techniques,
of which one is spherical cyclic indentation. In addition, he has investigated fatigue properties with
a cyclic small punch test and with cyclic automated ball indentation and has concluded that the
stiffness in the weld region drops more quickly in comparison to the base metal [21]. Xu et al. [22] have
performed the numerical analyses of flat cylindrical indentation for polycrystalline copper. They have
concluded that strain accumulation reached a steady-state indentation depth rate in sinusoidal cyclic
loading just like in real fatigue experiments.
From this literature review, it is concluded that using instrumented indentation for determining the
fatigue life of materials would reveal significant advantages of indentation testing (quasi-nondestructive
technique, less time- and cost-intensive) over conventional fatigue experiments. Some analytical
relations are indeed available to predict the S-N curve for fatigue life based on the hardness value,
but these methods are mostly limited to steels and not applicable to softer materials, for instance,
copper. Furthermore, it is not possible to analyze the microstructural influences. Hence, there is a need
for a more general approach that can be applied to a larger range of materials. In addition, a method
that uses the indentation experimental data for predicting complete uniaxial stress–strain hysteresis
is still missing in the literature according to the authors’ knowledge. Yet, this uniaxial stress–strain
hysteresis has a key role in fatigue life determination. Although some attempts have already been
made, as discussed earlier, the literature still lacks a hybrid method that predicts the material fatigue
life and the complete uniaxial stress–strain hysteresis with a combination of the numerical analysis
and Vickers indentation.
Materials 2020, 13, x 3 of 14

2. Material and Experiment

Materials 2020, 13, 3126 3 of 14
2.1. Materials Specifications
In2.this study,
Material andwe used a conventional quenched and tempered martensitic high-strength steel
Experiment
SAE 4150 (German denomination 50CrMo4, Judenburg GmbH, Judenburg, Austria) that exhibits a
hardness2.1. of
Materials
38 HRC. Specifications
The cyclic properties of this material were analyzed in detail by Schäfer et al.
[23]. The second material
In this study, we used
used ainconventional
this study, quenched
technically
andpure copper,
tempered is a conventional
martensitic Cu-ETP
high-strength steel R290
SAE 4150
(Electrolytic (German denomination
Tough-Pitch, 50CrMo4,
Wieland-Werke AG,Judenburg GmbH, Judenburg, Austria) that exhibits a
Ulm, Germany).
hardness of 38 HRC. The cyclic properties of this material were analyzed in detail by Schäfer et al. [23].
The second
2.2. Indentation andmaterial
Fatigueused in this study, technically pure copper, is a conventional Cu-ETP R290
Testing
(Electrolytic Tough-Pitch, Wieland-Werke AG, Ulm, Germany).
Indentation testing was performed by using a small Zwick load indenter and was adapted from
2.2. Indentation and Fatigue Testing
Kramer et al. [24]. The test was performed in a cyclic fashion; therefore, the indenter tip was
repeatedlyIndentation
indented at the same
testing spot. To by
was performed hold usingthea sample in the
small Zwick loadtesting
indenterposition,
and wasaadapted
constant fromminimal
Kramer et al. [24]. The test was performed in a cyclic fashion; therefore,
load remained between the cycles. A minimum load of 2 N and a maximum load of 50 N were used the indenter tip was repeatedly
indented at the same spot. To hold the sample in the testing position, a constant minimal load remained
for martensitic steel. Further experiments were also performed for higher force amplitudes, i.e., at 75
between the cycles. A minimum load of 2 N and a maximum load of 50 N were used for martensitic
and 100 N, and for higher hardness, such as 47 HRC at 50 N. In addition, a cyclic indentation test was
steel. Further experiments were also performed for higher force amplitudes, i.e., at 75 and 100 N,
performed
and forforhigher
Cu with a maximum
hardness, such as 47load
HRCofat10 50 N
N. and a minimum
In addition, a cyclicload of 1 N.test
indentation Thewasload increased at a
performed
rate offor
5 N/s. As an
Cu with example of
a maximum loadindentation
of 10 N andtesting,
a minimum onlyload
the offirst complete
1 N. The loadcycle for 50CrMo4
increased at a rate ofand Cu
is demonstrated
5 N/s. As an inexample
Figure 1a,c. Fatigue experimental
of indentation testing, only the data
firstofcomplete
uniaxialcycle
cyclicforstress–strain
50CrMo4 andcurves Cu is were
demonstrated
obtained from the studyin Figureof1a,c. Fatigue
Schäfer experimental
et al. data of stress–strain
[23]. Similarly, uniaxial cyclic hystereses
stress–strainarecurves werefor the
shown
obtained from the study of Schäfer et al.
aforementioned two materials (50CrMo4 and Cu) in Figure 1b,d.[23]. Similarly, stress–strain hystereses are shown for the
aforementioned two materials (50CrMo4 and Cu) in Figure 1b,d.
The complete cycle of the indentation test for martensitic steel (50CrMo4) is demonstrated in
The complete cycle of the indentation test for martensitic steel (50CrMo4) is demonstrated
Figurein2 Figure
with three different colors. The first (blue) part of the curve indicates the loading, the green
2 with three different colors. The first (blue) part of the curve indicates the loading,
sectionthedisplays the unloading,
green section and the red
displays the unloading, and portion demonstrates
the red portion demonstratesthe the
reloading
reloading for
forthe
the force–
displacement. It can be seen
force–displacement. It canthat the unloading
be seen and reloading
that the unloading curvescurves
and reloading makemakea closed
a closedloop, which is
loop,
named as the force–displacement loop (FD loop), and the intersection point is named ΔF. This loop
which is named as the force–displacement loop (FD loop), and the intersection point is named ∆F.
will beThis
used loop
aswill be used
a target as a target
curve for thecurve for the parameter
parameter identification
identification in the following
in the following section.
section.
50 50CrMo4 at 50N Wcyc for 50CrMo4
1000
Stress in MPa

40
500
Force in N

30
0
20
-500
10
-1000
0
0 4 8 12 16 20 24 -1.0 -0.5 0.0 0.5 1.0
a) Displacement in µm b) Strain in %
10 Cu at 10N 400 Wcyc for Cu
8
Stress in MPa

200
Force in N

6
0
4

2 -200

0 -400
0 4 8 12 16 20 -1.0 -0.5 0.0 0.5 1.0
Displacement in µm Strain in %
c) d)

FigureFigure
1. Experimental results
1. Experimental for for
results CuCu
and 50CrMo4:
and 50CrMo4:(a)
(a)indentation cycleatat5050NNforfor
indentation cycle 50CrMo4; (b)
50CrMo4;
(b) stress–strain from fatigue experiments of 50CrMo4; (c) indentation cycle
stress–strain from fatigue experiments of 50CrMo4; (c) indentation cycle at 10 N for Cu at 10 N for(d)
CuStress–
strain (d)
fromStress–strain from fatigue experiments of Cu.
fatigue experiments of Cu.
Materials 2020, 13, 3126 4 of 14
Materials
Materials2020,
2020,13,
13,xx 44 ofof 14
14

Loading
Loading
50
50 Unloading
Unloading
ΔF
ΔF
Reloading
Reloading

loop
Force-displacement loop
40
40
N
in N

Force-displacement
Force in 30
Force 30

20
20

10
10

00
00 55 10
10 15
15 20
20 25
25
Displacement
Displacementininµm
µm
2.2.Force–displacement
Figure2.
Figure Force–displacement
Force–displacementcurve curvefrom
fromthe cyclic
thecyclic indentation
cyclicindentation curve.
indentationcurve. Componentsof
curve.Components ofthe
thecycle
cycleare
are
shown in different
showninindifferent colors
differentcolors for
colorsfor clarity.
forclarity.
clarity.

2.3.
2.3. Numerical
2.3.Numerical
NumericalModels Models
Models
The
Thefinite
The finiteelement
finite elementmodel
element modelused
model usedin
used inthe
in theinvestigation
the investigationis
investigation isisdepicted
depictedin
depicted inFigure
in Figure3a.
Figure 3a.The
3a. Theindenter
The indenterwas
indenter was
was
modeled
modeled
modeledas as a rigid
asaarigid body.
rigidbody. The
body.The specimen
Thespecimen
specimenwas was fixed
wasfixed from
fixedfrom the
fromthe bottom,
thebottom,
bottom,andand a vertical
andaavertical load
verticalloadloadwaswas
wasapplied
applied
applied
at the center byusing
usingaaaVickers
Vickers indenter. The square pyramid ◦
at
atthe
thecenter
centerby by using Vickers indenter.
indenter. The
The square
square pyramid
pyramid hadhad
had an an opposite
anopposite
opposite faceface
face angle
angle angle
of
of136° of (DIN
136° 136
(DIN
(DIN
EN
EN ISO EN6507-2:
ISO ISO 6507-2:
6507-2: 2005)2005)
2005) [25].[25].
[25]. For For
For example,example,
example, the the
the applied applied
applied force
force force amplitude
amplitude
amplitude of of
of 50 50
50 N N
N inin in
the
thethe simulation
simulation
simulation isis
is explained
explained
explainedas as follows:
asfollows:
follows:The The specimen
Thespecimen
specimenwas was indented
wasindented
indentedwith with
withthe the preselected
thepreselected
preselectedapplied applied
appliedforce force
forceamplitude amplitude
amplitude (i.e.,
(i.e.,
(i.e.,
50
50N), 50which
N), N), which
which was was
wasunloadedunloaded
unloaded untiluntil
until the the force
theforce
force of of approximately
ofapproximately
approximately 22N Nand 2 Nreloaded
and and reloaded
reloaded to theto
tothe the maximum
maximum
maximum force.
force.
force.
Similarly, Similarly,
Similarly,the the simulations
thesimulations
simulationswere were
wererun runat run at
atother other
otherindentation indentation
indentationforce force
forceamplitudesamplitudes
amplitudesof of25, of
25,75, 25,
75,and 75,
and100and
100N. 100
N.The N.
The
The force–time
force–time
force–timehistory history
historyfrom from
fromthe the experiments
theexperiments
experimentswas wasused was
usedas used
asan as
aninput an
inputfor input
forthe for the
thesimulations simulations
simulationsso sothat so
thatthe that
theloading the
loading
loading
in
inexperiments in
experimentscouldexperiments
couldbe could
befully be fully
fullydepicted
depictedin depicted
inthe in the
thesimulations. simulations.
simulations.
It
ItItisisisour
ourgoal
our to identify
goal
goal to the material
to identify
identify the parameters,
the material
material which requires
parameters,
parameters, which many simulations
which requires
requires many to be performed
many simulations
simulations to
to bebe
and is
performed computationally very costly. Therefore, the size of the simulation
performed and is computationally very costly. Therefore, the size of the simulation model (2 mm ×22
and is computationally very costly. Therefore, the size of the model
simulation (2 mm
model × 2(2mmmm and
×
extruded
mm
mmand to
andextruded 1.5 mm)to
extruded towas
1.5optimized
1.5 mm)
mm)was with
wasoptimized a mesh
optimized sizeaaof
with
with 4 µmsize
mesh
mesh (Figure
size of443b);
of µmµmafter the3b);
(Figure
(Figure mesh
3b);after convergence
after the
themesh
mesh
study,
convergence C3D8
convergencestudy, linear elements
study,C3D8
C3D8linear with a full
linearelements integration
elementswith withaafull scheme
fullintegrationwere
integrationscheme chosen
schemewere for this
werechosenpaper.
chosenfor The
forthis
thisfriction
paper.
paper.
effect
The between
Thefriction
friction the between
effect
effect indenter
betweenthe and the specimen
theindenter
indenter and
andthe was
the also studied,
specimen
specimen was
wasalsoand itstudied,
did notand
alsostudied, show
and a considerable
ititdid
did not
notshow
showaa
effect on
considerable the
considerableeffect simulation
effectonontheresults in
thesimulation the scope
simulationresults of
resultsin this
inthe work.
thescope
scopeof ofthis
thiswork.
work.

(a)
(a)Geometry
Geometrywith
withboundary
boundaryconditions
conditions (b)
(b)Meshed
Meshedspecimen
specimen
Figure 3.3.Details
Figure3.
Figure Details of
Detailsof the
ofthe numerical
thenumerical model
numericalmodel used.
used.(a)
modelused. (a)Specimen
(a) Specimenis
Specimen isisheld
heldfixed
held from
fixedfrom
fixed the
fromthe bottom
thebottom and
bottomand the
andthe
the
indenter is placed
indenterisisplaced
indenter in the
placedininthe center
thecenter of the
centerofofthe specimen
thespecimen that
specimenthat moves
thatmoves in
movesin and
inand
andoutout during
outduring loading
duringloading and
loadingand unloading,
andunloading,
unloading,
respectively. (b)
respectively.(b)
respectively. The
(b)The fine
Thefine meshing
finemeshing
meshingis isisperformed
performed
performedatatatthe
the center
thecenter of
centerof the
ofthe specimen.
thespecimen.
specimen.
Materials 2020, 13, x 5 of 14

2.4. Material Model

In 2020,
Materials this 13,
investigation,
3126 the Chaboche material model was used, because it is one of the 5most of 14
efficient and convenient constitutive models incorporating the cyclic plasticity behavior of materials
during cyclic loading. Furthermore, the ratcheting behavior of the material, which rises under cyclic
2.4. Material Model
loading, can also be analyzed with it. Details of this model for J2 plasticity are given in the subsequent
section.
In this investigation, the Chaboche material model was used, because it is one of the most efficient
According to
and convenient the von Mises
constitutive yielding
models criterion [26],
incorporating the yielding
the cyclic of material
plasticity behavior starts once the during
of materials second
cyclic loading. Furthermore, the ratcheting behavior of the material, which rises under cyclic loading,f
deviatoric stress invariant J2 reaches a critical value. The common formula of this yielding criterion
is asalso
can follows:
be analyzed with it. Details of this model for J2 plasticity are given in the subsequent section.
According to the von Mises yielding criterion [26], the yielding of material starts once the second
deviatoric stress invariant J2 reaches 3
a critical
𝑓= √ (S −value.
𝛼): (SThe
− 𝛼)common
− (𝜎0 + formula
R) of this yielding criterion(1)f is
as follows: r
2
3
where 𝜎0, 𝛼, and S represent thef initial = (S − αstress,
yield α) − (σ0 + R
) : (S −backstress, ) deviatoric stress, respectively,
and (1)
2
and R represents the isotropic hardening (i.e., constant growth of yield surface) [27]. The increase in
plasticσstrain
where 0 , α, and S represent
with respect the initial
to the yield stress,
gradient of the backstress, and deviatoric
yield surface will leadstress,
to therespectively,
definition ofandthe
R
represents the isotropic hardening
associative flow rule used in this study. (i.e., constant growth of yield surface) [27]. The increase in plastic
strain with respect to the gradient of the yield surface will lead to the definition of the associative flow
rule used in this study. 𝑅 = 𝑄(1 − 𝑒 −𝑏𝜀𝑒𝑞 ) (2)
 
−bεeq
where 𝜀𝑒𝑞 is the equivalent plastic strain, R𝑏 = Q 1 − e the rate of isotropic hardening, and 𝑄 is the
determines (2)
maximum
where εeq ischange in the size
the equivalent of thestrain,
plastic yield surface [28,29].the rate of isotropic hardening, and Q is the
b determines
In order
maximum to model
change in thethe cyclic
size of thebehavior of materials,
yield surface [28,29]. a nonlinear kinematic hardening model was
proposed by Armstrong and Frederick [30]. This
In order to model the cyclic behavior of materials, kinematic hardening
a nonlinear model
kinematic containsmodel
hardening only was
one
backstress term (𝛼) and was extended by Chaboche by decomposing the single
proposed by Armstrong and Frederick [30]. This kinematic hardening model contains only one backstress term into
several backstress
backstress term (α)terms, making
and was the Chaboche
extended material
by Chaboche model [28] capable
by decomposing of capturing
the single theterm
backstress complex
into
kinematic hardening behavior.
several backstress terms, making the Chaboche material model [28] capable of capturing the complex
The decomposed
kinematic backstress terms of Chaboche kinematic hardening [28] model are described
hardening behavior.
in theThe
following equation:
decomposed backstress terms of Chaboche kinematic hardening [28] model are described in
the following equation: 𝑛
n 𝛼 ; 𝑑𝛼 =
2
𝛼=∑ X 𝑖 𝑖 23 𝐶𝑖 𝑑𝜀𝑝 − 𝛾𝑖 𝛼𝑖 𝑑𝜀𝑒𝑞 , (3)
α = 𝑖 αi ; dαi = Ci dεp − γi αi dεeq , (3)
3
i
where 𝛾𝑖 describes the reduction rate of the related modulus with respect to the plastic strain 𝑑𝜀𝑝 ,
where γi describes the reduction rate of the related modulus with respect to the plastic strain dεp , while Ci
while Ci represents the kinematic hardening moduli. The change in the yield surface of the combined
represents the kinematic hardening moduli. The change in the yield surface of the combined hardening
hardening evolution for monotonic tension and in the stress space is graphically presented in Figure
evolution for monotonic tension and in the stress space is graphically presented in Figure 4 [27].
4 [27].

 2
Bounding yield
surface  2


0 0 + C/ + R
R + 0
C
 
 1  pl

Subsequent yield
surface
Limiting value
of 
a) Stress space b) Monotonic tension
Figure 4.
Figure 4. Graphical depiction of the combined hardening growth in the (a) stress
stress space
space shown
shown by
by the
the
yield surface
yield surfaceand
and(b)
(b)under
undermonotonic
monotonic tension
tension presented
presented as aasstress–strain
a stress–strain diagram,
diagram, redrawn
redrawn fromfrom
[27]
[27] under
under the CC-BY
the CC-BY license.
license.

In the present study, three backstress terms, which comprise six unknowns, and isotropic softening
with two unknown parameters are used initially, leading to a total of eight unknowns which are
Materials 2020, 13, x 6 of 14

In the present study, three backstress terms, which comprise six unknowns, and isotropic
Materials 2020, 13, 3126 6 of 14
softening with two unknown parameters are used initially, leading to a total of eight unknowns
which are identified by using an inverse modeling technique. Furthermore, an effort has been made
to use only
identified bytwo
usingbackstress terms,
an inverse whichtechnique.
modeling reduces the unknown an
Furthermore, terms to has
effort six been
with made
the almost
to usesame
only
quality
two of results
backstress in our
terms, case reduces
which [31]. Therefore, in thisterms
the unknown study,toonly the results
six with of two
the almost backstress
same quality terms are
of results
shown.
in our case [31]. Therefore, in this study, only the results of two backstress terms are shown.

3. Inverse Parameter Identification

capture the
In order to capture the experimental
experimental force–displacement
force–displacement loop by simulation, the commercially
available LS-Opt optimizer (DYNAmore GmbH, Stuttgart, Stuttgart, Germany)
Germany) [32]
[32] was
was used.
used. The inverse
identification technique
identification techniquewaswasapplied,
applied, where
where the the difference
difference between
between experimental
experimental and simulated
and simulated values
values of the force–displacement loop, shown in Figure 2, was minimized by varying
of the force–displacement loop, shown in Figure 2, was minimized by varying the material parameters the material
parameters
in in procedure,
an iterative an iterativeasprocedure, as shown schematically
shown schematically in Figure 5. The in Figure
quality5.ofThe quality of
fit between thefitsimulated
between
the simulated
and and the force–displacement
the experimental experimental force–displacement loops was
loops was evaluated evaluated
by using by using themean
the normalized normalized
square
mean square
error (NMSE): error (NMSE):
11 X(𝐸 (Ei−−𝑆Si))22
NMSE =
NMSE = N ∑
𝑖 𝑖
,, (4)
(4)
𝑁 i 𝐸̅ES
𝑆̅
𝑖
an
an objective
objectivefunction,
function,where E and
where S represent
𝐸̅ and the averages
𝑆̅ represent of the experimental
the averages values Ei and
of the experimental simulation
values 𝐸𝑖 and
values S , respectively, for the displacement at the same force, and N is the total number of
simulation values 𝑆𝑖 , respectively, for the displacement at the same force, and N is the total number
i data
points (N = 75).
of data points (N = 75).

Start

Generate
iteration
No

Yes Numerical
Stop Converged
simulation
?

Evaluation
of Fitness Experimental
curve

Figure 5. Optimization
Optimization loop used to identify
identify the material
material parameters by fitting the experimental curve
with the
the simulation
simulation curve.
curve.

Hence, the force–displacement loop resulting from the partial unloading and reloading during
indentation was was used
usedasasthe
thetarget
targetfor for the
the optimization.
optimization. ForFor determining
determining the the material
material parameters
parameters that
that meet the given objective (i.e., that minimize the value of NMSE), a genetic
meet the given objective (i.e., that minimize the value of NMSE), a genetic algorithm [33] was used algorithm [33] was
used
becausebecause
it doesitnot
does not require
require a goodainitial goodguess
initialfor
guess for theparameters.
the target target parameters. This algorithm
This algorithm generates
generates the firstwhich
the first iteration, iteration, whichcertain
contains contains certain
material material sets
parameter parameter
(one setsets (one set
contains contains
four four
parameters
parameters for kinematic hardening and two for isotropic hardening) for
for kinematic hardening and two for isotropic hardening) for the identification process. The force–the identification process.
The force–displacement
displacement loop fromloop from the simulation
the simulation was takenwas outtaken
usingout using a postprocessing
a postprocessing script in
script written written
Pythonin
Python
3. Based3. on
Based
the on the fitness
fitness resultsresults obtained
obtained at end
at the the end of the
of the first
first iteration,
iteration, thealgorithm
the algorithmupdates
updates the
material parameters in the subsequent iterations as the optimization loop continues. In each each iteration,
iteration,
the algorithm calculates the fitness of the obtained obtained force–displacement loop with the experimental experimental
force–displacement loop by using NMSE. This optimization loop continues until the convergence
criterion
criterion (i.e., NMSE == 3 ×
(i.e., NMSE 10−5−5) )isismet
× 10 metor orthe
themaximum
maximumallowed
allowediterations
iterations are
are reached.
reached. The
The yield stress
and Young’s modulus are kept constant at 1060 MPa and 204 GPa, respectively, based on monotonic
stress–strain experimental data. It It is
is known that the yield strength cannot be uniquely uniquely determined
determined
only based on force–displacement curves with sharp indenters. Hence, this material parameter must
Materials 2020, 13, 3126 7 of 14
Materials 2020, 13, x 7 of 14

be
be assumed
assumed as as known
known and
and can be determined
determined by otherother methods,
methods, e.g., e.g., tensile
tensile tests
tests or
or other
other inverse
inverse
methods
methods based on indentation (e.g., see [1–11]).
Two kinds of optimization procedures were studied in this work: “objective “objective function
function 1”
1” includes
includes
aa free
free optimization
optimizationof ofthe
theobjective
objectivefunction
functiondefined
definedininEquation
Equation (4)(4)
until the
until convergence
the convergence criterion is
criterion
reached; “objective
is reached; function
“objective 2” 2”
function is isalso
alsobased
basedononthetheobjective
objectivefunction
functionofof Equation
Equation (4),
(4), but
but the
the
minimization
minimization occurs
occurs under
under the
the side
side condition
condition that
that the
the height
height of of the force–displacement
force–displacement loop loop is
restricted
restricted to the experimentally found value, i.e., ΔF ∆Fsim
sim ==ΔF∆Fexp.
exp .

4.
4. Results
Results and
and Discussion
Discussion

4.1. Method
4.1. Method Development
Development
The experimental
The experimental force–displacement
force–displacement looploop of
of the
the first
first indentation
indentation cycle
cycle is
is used
used as
as aa target
target curve
curve
along with ∆F. The material parameters (see Table 1) obtained after the optimization
along with ΔF. The material parameters (see Table 1) obtained after the optimization with this with this strategy
show a good
strategy showagreement for the complete
a good agreement for theindentation cycle: The normalized
complete indentation cycle: The mean square mean
normalized error (NMSE)
square
between the simulated force–displacement loop and the experimental
error (NMSE) between the simulated force–displacement loop and the experimental force– force–displacement loop is
−5
2.0 × 10 . loop is 2.0 × 10−5.
displacement
To achieve
To achieve aa comparison
comparison withwith experimentally determined hysteresis
experimentally determined hysteresis from
from fatigue
fatigue tests,
tests, the
the
hysteresis under a tensile–compressive load is predicted in the next step by using
hysteresis under a tensile–compressive load is predicted in the next step by using the parameters the parameters from
TableTable
from 1. Comparing
1. Comparingthis prediction to the
this prediction to experimental
the experimental results reveals,
results with
reveals, a plastic
with work
a plastic workerror of
error
2.5%, a quite good accuracy, as can be seen in Figure 6. In the scope of this study, the identified
of 2.5%, a quite good accuracy, as can be seen in Figure 6. In the scope of this study, the identified material
parameters
material of cyclic macroindentation
parameters are used to
of cyclic macroindentation arepredict
used to the complete
predict uniaxial stress–strain
the complete hysteresis
uniaxial stress–strain
for the firstfor
hysteresis time.
the first time.

50 Optimization Prediction
Experiment Experiment
1000
40 Simulation Simulation
Stress in MPa

500
Force in N

30
0
20
-500
10
-1000
0
0 4 8 12 16 20 24 -1.0 -0.5 0.0 0.5 1.0
a) Displacement in µm b) Strain in %
Figure
Figure 6.
6. (a)
(a)Complete
Completecycle
cycleofofthe
theforce–displacement
force–displacement curve
curve from
from indentation,
indentation, with
with normalized
normalized mean
mean
square
square error (NMSE) = 2.0 × 10 . (b) Predicted uniaxial stress–strain hysteresis with aa plastic
error (NMSE) = 2.0 × 10 −5.
−5 (b) Predicted uniaxial stress–strain hysteresis with plastic work
work
error
error of
of 2.5%.
2.5%.

There isisaaclear
clearrelationship
relationship between
between ∆F value
the the ΔF value measured
measured by indentation
by cyclic cyclic indentation
and the and the
uniaxial
uniaxial stress–strain
stress–strain hysteresis.
hysteresis. This relationship
This relationship is qualitatively
is qualitatively investigated
investigated in this
in this study.
study. In In Figure
Figure 7,
7, the results from the two objective functions can
the results from the two objective functions can be seen. be seen.
By using the “objective function 1”, the comparison of the simulated and experimental
Table 1. Identified
force–displacement material
loops parameters
seems to be infor
an 50CrMo4
acceptable (38range
HRC) after fitting of force–displacement
of agreement (see Figure 7, solidat blue
50 N.
FD loop), with NMSE 3.0 × 10 ; nevertheless, the value of ∆F from the simulation is lower than
−5

the experimental ∆F, which has a direct impact Symbol on Value

the uniaxial stress–strain hysteresis prediction.
The inclusion of the ∆F into “objective function 2” leads
C1 (MPa) to a better prediction of the uniaxial stress–strain
262,197
γ 373
hysteresis, which is shown by the dotted blue line hysteresis in Figure 7.
1

C2 (MPa) 4714
γ2 0.25
Q (MPa) –575
b 262
 Materials 2020, 13, 3126 8 of 14

Table 1. Identified material parameters for 50CrMo4 (38 HRC) after fitting of force–displacement at
Materials 2020, 13, x
50 N. 8 of 14

By using the “objective function 1”,Symbol Value of the simulated and experimental force–
the comparison
C1 (MPa)range262,197
displacement loops seems to be in an acceptable of agreement (see Figure 7, solid blue FD loop),
γ1 373
with NMSE 3.0 × 10 ; nevertheless, the value
−5 of ΔF from
C2 (MPa)
the simulation is lower than the experimental
4714
ΔF, which has a direct impact on the uniaxial
γ2 stress–strain
0.25 hysteresis prediction. The inclusion of the
Q (MPa) −575
ΔF into “objective function 2” leads to a better prediction of the uniaxial stress–strain hysteresis,
b 262
which is shown by the dotted blue line hysteresis in Figure 7.

Experiment 50
1000 Objective function 1
Objective function 2 EXP
SIM
Stress in MPa

500

Force in N
0

FD Loop

FD Loop
-500
Prediction
-1000
0
-1.0 -0.5 0.0 0.5 1.0
Displacement in µm
Strain in %
Figure
Figure 7. Effect ∆F on
7. Effectofof ΔF onthethe
uniaxial stress–strain
uniaxial hysteresis
stress–strain prediction. prediction.
hysteresis The solid blueThe
force-displacement
solid blue force-
(FD) loop displays the fitting of the FD loop to the blue experimental FD loop by
displacement (FD) loop displays the fitting of the FD loop to the blue experimental FD loop using objective function
by using
1, while the dotted solid stress–strain hysteresis is the prediction of stress–strain hysteresis.
objective function 1, while the dotted solid stress–strain hysteresis is the prediction of stress–strain Similarly,
the dotted blue FD loop shows the fitting of the FD loop by using objective function 2, while the dotted
hysteresis. Similarly, the dotted blue FD loop shows the fitting of the FD loop by using objective
blue stress–strain hysteresis represents the prediction of the stress–strain hysteresis.
function 2, while the dotted blue stress–strain hysteresis represents the prediction of the stress–strain
It can be observed that if the simulated ∆F has a lower value than the experimental ∆F, the prediction
hysteresis.
of the uniaxial stress–strain hysteresis reveals a larger disagreement (plastic work error = 20%) between
It can
the be observed
experimental that
and the if the simulated
simulated stress–strainΔF has a lower
hysteresis. On thevalue than the
other hand, the value of ∆FΔF,
whenexperimental is the
prediction of thetouniaxial
comparable the experimental ∆F value,
stress–strain hysteresis reveals
the uniaxial a larger disagreement
stress–strain (plastic
hysteresis provides work error =
an acceptable
20%)prediction
between (plastic work error =and
the experimental 2.0%),
theas demonstrated
simulated by the dotted
stress–strain solid curve
hysteresis. On intheFigure
other7.hand, when
the valueFrom nowison,
of ΔF we will onlyto
comparable present the results obtained
the experimental by “objective
ΔF value, function stress–strain
the uniaxial 2” after optimization.
hysteresis
As the Chaboche material model is also capable to capture the ratcheting behavior
provides an acceptable prediction (plastic work error = 2.0%), as demonstrated by the dotted solid in cyclic loading,
further simulations are performed with multiple cycles of indentation to compare the experimental
curve in Figure 7.
ratcheting effect of force–displacement by using the identified parameters from the complete indentation
From now on, we will only present the results obtained by “objective function 2” after
cycle. The ratcheting observed in simulation and experiment is slightly overestimated. In Figure 8b,
optimization. As the Chaboche material model is also capable to capture the ratcheting behavior in
force–displacement curves for 13 consecutive cycles are compared with the experiment.
cyclic loading, further simulations are performed with multiple cycles of indentation to compare the
4.2. Validation
experimental ratcheting effect of force–displacement by using the identified parameters from the
completeThe indentation cycle. The
material parameters, ratcheting
which have beenobserved in asimulation
identified at and experiment
50 N force amplitude, is slightly
are also tested at
overestimated.
higher forcesInofFigure
75 and 8b,
100 force–displacement
N to check the validitycurves for 13 consecutive
of the obtained parameters.cycles are compared
The comparison of thewith
the experiment.
simulation curve and the experimental curve at higher force amplitudes reveals that they are also in
good agreement at 75 N force amplitude (NMSE = 3.3 × 10−4 ; Figure 9a) and 100 N force amplitude
(NMSE = 1.6 × 10−4 ; Figure 9b).
 50 Prediction
Experiment
40 40 Simulation

Force in N

Force in N
30 30
Materials 2020, 13, 3126 9 of 14
Materials 2020, 13,20
x 9 of 14
20
50 Optimization 50 Prediction
10 10
Experiment Experiment
40 Simulation 400 Simulation
Force in N 0

Force in N
a) 30 Displacement in µm b) 30 Displacement in µm
Figure 8. (a) Simulated force–displacement loop. (b) Predicted force–displacement for 13 cycles.
20 20
4.2. Validation Optimization
10 10
Experiment
The material parameters, which have been identified at a 50 N force amplitude, are also tested
Simulation
at higher forces0 of 75 and 100 N to check the validity of the 0 obtained parameters. The comparison of
the simulation curve and the experimental curve at higher force amplitudes reveals that they are also
a)
in good agreement Displacement in µm
at 75 N force amplitude (NMSE =b) Displacement
3.3 × 10−4; Figure in µm
9a) and 100 N force amplitude
(NMSEFigure
= 1.6 ×8.
Figure 8.10
(a) ;Simulated
−4
(a) Figure 9b).
Simulated force–displacement
force–displacement loop.
loop. (b)
(b) Predicted
Predicted force–displacement
force–displacement for
for 13
13 cycles.
cycles.

4.2. Validation Prediction 75 100 Prediction

Experiment Experiment
The material60parameters, which have been identified 80
at a 50 N force amplitude, are also tested
Simulation Simulation
at higher forces of 75 and 100 N to check the validity of the obtained parameters. The comparison of
the simulation curve
45 and the experimental curve at higher force
60 amplitudes reveals that they are also
Force in N

Force in N

in good agreement at 75 N force amplitude (NMSE = 3.3 × 10−4; Figure 9a) and 100 N force amplitude
(NMSE = 1.6 × 1030−4; Figure 9b).
40
75 Prediction 100 Prediction
15 20
Experiment Experiment
60 Simulation 80 Simulation
0 0

a) 45 Displacement in µm b) 60
Force in N

Force in N

Displacement in µm
Figure
Figure 9.
9. Validation
Validation of
ofthe
themethod
methodatathigher
higherforce
forceamplitudes:
amplitudes:(a) predicted force–displacement at 75
30 40 (a) predicted force–displacement at
N; (b) predicted force–displacement at 100 N.
75 N; (b) predicted force–displacement at 100 N.

15 20
Figure 10a demonstrates the comparison of experimental and simulated uniaxial stress–strain
hystereses for
for the
the 10th
10thcycle
cycleofofthe
thesame
samematerial.
material. The
Theredred
hysteresis, which is obtained by using the
0 0hysteresis, which is obtained by using
above-obtained
the above-obtained fitted material
fitted materialparameters
parametersfrom
fromthe
thecyclic
cyclicindentation
indentationforce–displacement
force–displacement curve,
curve,
shows aa quite
quitea)
goodagreement
good agreement
Displacement (plastic work
in work
(plastic µm error = 3.5%)
error = 3.5%) with
b)with the experimental
theDisplacement
experimental hysteresis.
in µm
hysteresis. Figure
Figure 10b
10b demonstrates
demonstrates the maximum
the maximum and minimum
and minimum stress
stress on the on the vertical
vertical axis,the
axis, while while the horizontal
horizontal axis
axis displays
Figure 9. Validation of the method at higher force amplitudes: (a) predicted force–displacement at 75
displays
the cycle the cycle number.
number.
N; (b)2020,
Materials predicted
13, x force–displacement at 100 N. 10 of 14

Prediction
Figure 10a demonstrates
Experimentthe comparison of experimental and simulated uniaxial stress–strain
1000 1000
hystereses for the 10th Simulation
cycle of the same material. The red hysteresis, which is obtained by using the
Stress in MPa

above-obtained500fitted material parameters from the500 cyclic indentation force–displacement curve,

Prediction
shows a quite good agreement (plastic work error = 3.5%) with the experimental
Max-Simulation hysteresis. Figure
0 0 Max-Experiment
10b demonstrates the maximum and minimum stress on the vertical axis, while the horizontal axis
Min-Simulation
Min-Experiment
displays the cycle
-500 number. -500

-1000 -1000

-1.0 -0.5 0.0 0.5 1.0 0 2 4 6 8 10

a) Strain in % b) Number of cylces

Figure
Figure 10.10.
(a)(a) Predicteduniaxial
Predicted uniaxialstress–strain
stress–strain hysteresis
hysteresis for
for the
the 10th
10thcycle.
cycle.(b)
(b)Stress
Stressamplitude
amplitudeover
over
the number of cycles for the first 10 cycles.
the number of cycles for the first 10 cycles.

It is evident from Figure 10 that the material parameters obtained from the cyclic indentation
force–displacement curve can predict higher uniaxial stress–strain hysteresis very accurately, and the
difference is less than 4%. This uniaxial stress–strain hysteresis is obtained without any initial input
from uniaxial stress–strain hysteresis. The uniaxial stress–strain hysteresis has a key role in
determining material fatigue life. As already mentioned, performing fatigue experiments is quite
 -1.0 -0.5 0.0 0.5 1.0 0 2 4 6 8 10
a) Strain in % b) Number of cylces

Figure 10. (a) Predicted uniaxial stress–strain hysteresis for the 10th cycle. (b) Stress amplitude over
the number of cycles for the first 10 cycles.
Materials 2020, 13, 3126 10 of 14

It is evident from Figure 10 that the material parameters obtained from the cyclic indentation
force–displacement
It is evident from curve can predict
Figure 10 that higher uniaxial
the material stress–strain
parameters hysteresis
obtained fromverythe accurately, and the
cyclic indentation
difference is less than
force–displacement 4%. can
curve Thispredict
uniaxial stress–strain
higher uniaxial hysteresis
stress–strainis obtained
hysteresiswithout any initial
very accurately, andinput
the
from uniaxial
difference is lessstress–strain
than 4%. This hysteresis. The uniaxial
uniaxial stress–strain stress–strain
hysteresis hysteresis
is obtained without hasanya key
initialrole in
input
determining
from uniaxial material
stress–strainfatigue life. AsThe
hysteresis. already mentioned,
uniaxial stress–strainperforming
hysteresisfatigue
has a keyexperiments is quite
role in determining
expensive
material both in
fatigue terms
life. of cost and
As already time, and
mentioned, performing
performing indentation
fatigue tests is
experiments is quite
quite easy and requires
expensive both in
fewer resources.
terms of cost andBy using
time, andthis technique indentation
performing of identifying kinematic
tests is quitehardening materialfewer
easy and requires parameters, the
resources.
need
By to perform
using fatigueof
this technique experiments
identifyingwill be required
kinematic only for
hardening the validation
material parameters, process.
the need to perform
fatigue experiments will be required only for the validation process.
4.3. Transferability of the Method
4.3. Transferability of the Method
4.3.1. Transferability to Higher Force Amplitude (75 N)
4.3.1. Transferability to Higher Force Amplitude (75 N)
Until now, we have used a 50 N indentation force amplitude for identifying material parameters
Untilpredicted
and then now, we the have used a stress–strain
uniaxial 50 N indentation force amplitude
hysteresis for identifying
by using these material parameters
identified parameters. To check
and then predicted the uniaxial stress–strain hysteresis by using these identified
the robustness and transferability of our method, we have decided to also fit the 75 N indentation parameters. To check
the
curverobustness
and to tryand transferability
to predict of our
the uniaxial method, we
stress–strain have decided
hysteresis by usingto also
thesefitparameters.
the 75 N indentation
The rest of
curve and to try to predict the uniaxial stress–strain hysteresis by using
the optimization setting and procedure were kept the same as explained before. Figure 11a these parameters. The rest of
displays
the
the optimization
comparison setting of theand procedure
cyclic were kept the same
force–displacement curve,aswhile
explained before.
Figure 11b Figure 11a displays
demonstrates the
the
comparison of the uniaxial stress–strain between simulation and experiment. The results comparison
comparison of the cyclic force–displacement curve, while Figure 11b demonstrates the are in good
of the uniaxial
agreement (NMSE stress–strain
= 4.0 × 10between
−5) betweensimulation and experiment.
the experimental and theThe results are
simulated in good
curves agreement
for both force–
(NMSE = 4.0 × 10 −5 ) between the experimental and the simulated curves for
displacement and uniaxial stress–strain hysteresis. The parameters obtained after the simulation are both force–displacement
and uniaxial
reported stress–strain
in Table 2 and hysteresis. The parameters
not much different from the obtained after the
parameters simulation
obtained are reported
for the 50 N force in
Table 2
amplitude. and not much different from the parameters obtained for the 50 N force amplitude.

80 Optimization Prediction
Experiment
1000
Experiment
Simulation
60
Stress in MPa

500 Simulation
Force in N

40 0

-500
20
-1000
0
-1.0 -0.5 0.0 0.5 1.0
a) Displacement in µm b) Strain in %

Figure 11. Transferability of method: (a) indentation of 38 HRC at 75 N; (b) prediction of stress–strain
hysteresis of 38 HRC.

Table 2. Identified material parameters for 50CrMo4 (38 HRC) after fitting of force–displacement at
75 N force amplitude.

Symbol Value
C1 (MPa) 257,503
γ1 354
C2 (MPa) 3663
γ2 0.2837
Q (MPa) −611
b 163

4.3.2. Transferability to Higher Hardness (47 HRC)

Similarly, the force–displacement loop at 50 N force amplitude of 47 HRC hardness is used for
material parameter identification by the inverse method. The value of yield strength used (1400 MPa)
was obtained from the monotonic loading experiments. The results of the force–displacement curve after
 b 163

4.3.2. Transferability to Higher Hardness (47 HRC)

Similarly, the force–displacement loop at 50 N force amplitude of 47 HRC hardness is used for
material
Materials parameter
2020, 13, 3126 identification by the inverse method. The value of yield strength used (1400 MPa)
11 of 14
was obtained from the monotonic loading experiments. The results of the force–displacement curve
after the optimization show a good agreement between the experimental and simulated force–
the optimization show a good agreement between the experimental and simulated force–displacement
displacement loops, as shown in Figure 12. The prediction of the uniaxial stress–strain curve (by
loops,
usingas shown
the in Figure
material 12. The
parameters prediction
from Table 3)of thedepicts
also uniaxial stress–strain
a sufficient curve (by
agreement using simulation
between the material
parameters from Table
and experiment. 3) also depicts
The difference a sufficient
in energy agreement
dissipation between between simulation
the experiment and
and the experiment.
prediction is
The difference
4.5%. in energy dissipation between the experiment and the prediction is 4.5%.

1500
50 Prediction
1000 Experiment

Stress in MPa
40 Simulation
Force in N

500
30
0
20 -500
Optimization
10 Experiment -1000
0 Simulation
-1500
-1.0 -0.5 0 0.5 1.0
a) Displacement in µm b) Strain in %
Figure12.
Figure 12.Transferability
Transferability
of of method:
method: (a)(a) simulated
simulated indentation
indentation force–displacement
force–displacement at N
at 50 50for
N 47
forHRC;
47
HRC; (b) prediction of uniaxial stress–strain hysteresis of
(b) prediction of uniaxial stress–strain hysteresis of 47 HRC. 47 HRC.

Table Identified material

Table3.3. Identified materialparameters
parametersforfor 50CrMo4
50CrMo4 (47 (47
HRC)HRC) of force–displacement
of force–displacement at 50 Natforce
50 N
force amplitude.
amplitude.

Symbol
Symbol Value
Value
C1 (MPa) 337,885
C1 (MPa) 337,885
γ1 374
γ1 374
C2 (MPa) 6681
C2 (MPa) 6681
γ2γ2 2.3
2.3
Q (MPa)
Q (MPa) –724
−724
bb 273
273

4.3.3. Transferability for Other Material (Cu)

4.3.3. Transferability for Other Material (Cu)
The extensive study of using different force amplitudes and different hardnesses has been done
The extensive study of using different force amplitudes and different hardnesses has been done
with our method for martensitic steel in the previous section. In this section, the aim is to test our
with our method for martensitic steel in the previous section. In this section, the aim is to test our
methodology on a different metallic material. For this purpose, copper (Cu), a relatively softer
methodology on a different metallic material. For this purpose, copper (Cu), a relatively softer material
material is selected, and therefore, instead of using a 50 N force amplitude, a smaller force amplitude
isof
selected, and therefore,
10 N is applied instead
for cyclic of using
indentation. a 50
The N force amplitude,
optimization a smaller
is performed force
by using theamplitude
same setupofwith
10 N
istwo
applied for cyclic indentation. The optimization is performed by using the same setup with
backstress terms (Table 4), and the results after optimization are shown in Figure 13. Figure 13a two
backstress terms
demonstrates the(Table 4), and
comparison of the results after optimization
the force–displacement are shown
at 10 N between in Figure
simulation 13.experiment,
and Figure 13a
demonstrates the comparison of the force–displacement at 10 N between simulation and experiment,
while Figure 13b shows the predicted uniaxial stress–strain at 1% total strain amplitude. It can be
while Figure 13b shows the predicted uniaxial stress–strain at 1% total strain amplitude. It can be
seen from Figure 13a that the force–displacement has quite a good fit after optimization. The same is
true when we predict the uniaxial stress–strain hysteresis by using these identified parameters and
compare it with the experimental uniaxial stress–strain hysteresis. The difference in dissipated energy
and between predicted and experimental stress–strain hysteresis is only 3.5%.

Table 4. Identified material parameters of the force–displacement loop for Cu at 10 N force amplitude.

Symbol. Value
C1 (MPa) 154,790
γ1 2,257
C2 (MPa) 11,586
γ2 82
Q (MPa) −12
b 47
Materials 2020, 13, x 12 of 14

seen from Figure 13a that the force–displacement has quite a good fit after optimization. The same is
true when we predict the uniaxial stress–strain hysteresis by using these identified parameters and
compare
Materials it 13,
2020, with3126the experimental uniaxial stress–strain hysteresis. The difference in dissipated
12 of 14
energy and between predicted and experimental stress–strain hysteresis is only 3.5%.

10 400

8
200

Stress in MPa
Prediction
Force in N

6
Experiment
0 Simulation
4

2 Optimization -200
Experiment
0 Simulation -400
-1.0 -0.5 0.0 0.5 1.0
a) Displacement in µm b) Strain in %
Figure 13.
Figure 13. Transferability
Transferabilityof ofmethod:
method:(a)(a)simulated indentation
simulated force–displacement
indentation of Cu
force–displacement at 10
of Cu at N;
10 (b)
N;
prediction
(b) of uniaxial
prediction stress–strain
of uniaxial hysteresis.
stress–strain hysteresis.

5. Conclusions
Table 4. Identified material parameters of the force–displacement loop for Cu at 10 N force
A novel hybrid method for the inverse analysis amplitude.
of fatigue properties of metals has been introduced.
The method combines cyclic Vickers indentation Symbol. experiments
Value and finite element simulations in an
inverse method, by which the material parameters C1 (MPa)are 154,790
determined in an iterative way by an optimization
scheme. It has been demonstrated that this method γ1 2,257
can be used to determine the parameters of the
C (MPa) 11,586
Chaboche model for kinematic hardening. Based on these parameters, the model has been successfully
2

employed to predict the cyclic stress–strainγresponses 2 82 of a tempered martensitic steel, SAE 4150
Q (MPa) –12
(German denomination 50CrMo4), with different heat treatments and of technically pure copper.
b 47
The error in the parameters determined with the inverse method has been evaluated as less than 4%
on average. It has been observed that the difference between the maximum and minimum force of
5. Conclusions
the force–displacement loop obtained from cyclic indentation has a direct correlation with the stress
A novel
amplitude hybrid
of the method
hysteresis loopfor the inverse
measured analysis of fatigue
in strain-controlled properties
uniaxial of metals
fatigue tests; thus, has been
it plays a
introduced. The method combines cyclic Vickers indentation experiments
crucial role in predicting the uniaxial stress–strain hysteresis accurately. By applying the method to and finite element
simulations inmartensitic
high-strength an inversesteel,
method, by which
on which it hasthe material
been parameters
validated are determined
for different maximum forces,in an and
iterative
also
way
to by an optimization
technically pure copper, scheme. It has been
its validity demonstrated
has been demonstratedthat this
formethod
a wide can be used
variety to determine
of materials and
the parameters of the Chaboche model for kinematic hardening. Based on
process parameters. The prediction of a complete cyclic stress–strain curve by using data from cyclic these parameters, the
model has been
indentation successfully
has great potentialemployed
to reducetotime-predictand the cyclic stress–strain
cost-intensive responses ofand
fatigue experiments a tempered
can thus
martensitic
open a newsteel, SAE 4150 way
and economic (German denomination
to predict the fatigue 50CrMo4), with different
life of materials with aheat treatments and of
quasi-nondestructive
technically
test method.pure copper. The error in the parameters determined with the inverse method has been
evaluated as less than 4% on average. It has been observed that the difference between the maximum
Author force Conceptualization,
Contributions:
and minimum H.M.S., H.u.H.,
of the force–displacement loop M.K., and A.H.;
obtained frommethodology, H.M.S., has
cyclic indentation H.u.H., M.K.,
a direct
and A.H.; investigation, H.M.S., B.J.S., and P.S.-R.; data curation; B.J.S., P.S.-R., and M.K.; writing—original draft,
correlation
H.M.S.; with theH.M.S.;
visualization, stress writing—review
amplitude of the hysteresis
& editing, M.K.,loop measured
H.u.H., and A.H., insupervision,
strain-controlled
A.H. Alluniaxial
authors
fatigue
have tests;
read and thus,
agreedittoplays a crucialversion
the published role in of
predicting the uniaxial stress–strain hysteresis accurately.
the manuscript.
By applying
Funding: Thisthe method
research to high-strength
received martensitic steel, on which it has been validated for different
no external funding.
maximum forces, and also to technically pure copper, its validity has been demonstrated for a wide
Acknowledgments: The authors gratefully acknowledge the financial support of the Higher Education
variety of materials
Commission (HEC) of and process
Pakistan, DFGparameters.
Open AccessThe prediction
Publication Fundsof of
a complete cyclic stress–strain
the Ruhr-Universität Bochum, and curve
the
contributions
by using data of from
Robertcyclic
Boschindentation
GmbH in carrying out the
has great experiments.
potential to reduce time- and cost-intensive fatigue
experiments
Conflicts and can
of Interest: Thethus opendeclare
authors a newno and economic
conflict way to predict the fatigue life of materials with
of interest.
a quasi-nondestructive test method.
References
Author Contributions: Conceptualization, H.M.S., H.u.H., M.K., and A.H.; methodology, H.M.S., H.u.H., M.K.,
1. Huber, N.; Tsakmakis, C. Determination of constitutive properties from spherical indentation data using
and A.H.; investigation, H.M.S., B.J.S., and P.S.-R.; data curation; B.J.S., P.S.-R., and M.K.; writing—original draft,
neural networks. Part I: The case of pure kinematic hardening in plasticity laws. J. Mech. Phys. Solids 1999, 47,
1569–1588. [CrossRef]
Materials 2020, 13, 3126 13 of 14

2. Huber, N.; Tsakmakis, C. Determination of constitutive properties from spherical indentation data using
neural networks. Part II: Plasticity with nonlinear isotropic and kinematic hardening. J. Mech. Phys. Solids
1999, 47, 1589–1607. [CrossRef]
3. Wymysłowski, A.; Dowhań, Ł. Application of nanoindentation technique for investigation of elasto-plastic
properties of the selected thin film materials. Microelectron. Reliab. 2013, 53, 443–451. [CrossRef]
4. Oliver, W.C.; Pharr, G.M. Measurement of hardness and elastic modulus by instrumented indentation:
Advances in understanding and refinements to methodology. J. Mater. Res. 2004, 19, 3–20. [CrossRef]
5. Peirce, D.; Asaro, R.J.; Needleman, A. Material rate dependence and localized deformation in crystalline
solids. Acta Metall. 1983, 31, 1951–1976. [CrossRef]
6. Hyung-Yil, L. Ball Indenter Utilizing Fea Solutions for Property Evaluation. WO2003010515A1, 2002.
Available online: https://patents.google.com/patent/WO2003010515A1 (accessed on 10 July 2020).
7. Suresch, A.; Alcala, S.; Giannakopoulos, J. Depth Sensing Indentation and Methodology for Mechanical
Property Measurements. WO1997039333A2, 1996. Available online: https://patents.google.com/patent/
WO1997039333A2 (accessed on 10 July 2020).
8. Suresh, T.A.; Dao, S.; Chollacoop, M.; Van, N.; Venkatesh, K.V. Systems and Methods for Estimation and
Analysis of Mechanical Property Data. WO2002073162A2, 2002. Available online: https://patents.google.
com/patent/WO2002073162A2 (accessed on 10 July 2020).
9. Fontanari, V.; Beghini, M.; Bertini, L. Method and Apparatus for Determining Mechanical Features of
a Material with Comparison to Reference Database. WO2006013450A2, 2004. Available online: https:
//patents.google.com/patent/WO2006013450A2/en (accessed on July 10 2020).
10. Schmaling, B.; Hartmaier, A. Method for Testing Material, particularly for Hardness Testing, Involves
Producing Impression in to Be Tested Material in Experimental Manner with Test Body with Known
Geometry and with Known Test Load. DE102011115519A1, 2011. Available online: https://patents.google.
com/patent/DE102011115519A1/de (accessed on 10 July 2020).
11. Broitman, E. Indentation Hardness Measurements at Macro-, Micro-, and Nanoscale: A Critical Overview.
Tribol. Lett. 2017, 65, 23. [CrossRef]
12. Strzelecki, P. Analytical Method for Determining Fatigue Properties of Materials and Construction Elements in High
Cycle Life; Uniwersytet Technologiczno-Przyrodniczy w Bydgoszczy: Bydgoszcz, Poland, 2014.
13. Murakami, Y. Effects of small defects and nonmetallic inclusions on the fatigue strength of metals. JMSE Int. J.
1989, 32, 167–180. [CrossRef]
14. Bandara, C.S.; Siriwardane, S.C.; Dissanayake, U.I.; Dissanayake, R. Developing a full range S-N curve and
estimating cumulative fatigue damage of steel elements. Comput. Mater. Sci. 2015, 96, 96–101. [CrossRef]
15. Bandara, C.S.; Siriwardane, S.C.; Dissanayake, U.I.; Dissanayake, R. Full range S-N curves for fatigue life
evaluation of steels using hardness measurements. Int. J. Fatigue 2016, 82, 325–331. [CrossRef]
16. Strzelecki, P.; Tomaszewski, T. Analytical models of the S-N curve based on the hardness of the material.
Procedia Struct. Integr. 2017, 5, 832–839. [CrossRef]
17. Lyamkin, V.; Starke, P.; Boller, C. Cyclic indentation as an alternative to classic fatigue evaluation.
In Proceedings of the 7th International Symposium on Aircraft Materialsno, Compiegne, France, 24–26 April
2018.
18. Faisal, N.H.; Prathuru, A.K.; Goel, S.; Ahmed, R.; Droubi, M.G.; Beake, B.D.; Fu, Y.Q. Cyclic Nanoindentation
and Nano-Impact Fatigue Mechanisms of Functionally Graded TiN/TiNi Film. Shape Mem. Superelasticity
2017, 3, 149–167. [CrossRef]
19. Haghshenas, M.; Klassen, R.J.; Liu, S.F. Depth-sensing cyclic nanoindentation of tantalum. Int. J. Refract. Met.
Hard Mater. 2017, 66, 144–149. [CrossRef]
20. Prakash, R.V. Evaluation of fatigue damage in materials using indentation testing and infrared thermography.
Trans. Indian Inst. Met. 2010, 63, 173–179. [CrossRef]
21. Prakash, R.V. Study of Fatigue Properties of Materials through Cyclic Automated Ball Indentation and Cyclic
Small Punch Test Methods. Key Eng. Mater. 2017, 734, 273–284. [CrossRef]
22. Xu, B.X.; Yue, Z.F.; Chen, X. Numerical investigation of indentation fatigue on polycrystalline copper.
J. Mater. Res. 2009, 24, 1007–1015. [CrossRef]
23. Schäfer, B.; Song, X.; Sonnweber-Ribic, P.; Hassan, H.U.; Hartmaier, A. Micromechanical Modelling of the
Cyclic Deformation Behavior of Martensitic SAE 4150—A Comparison of Different Kinematic Hardening
Models. Metals (Basel) 2019, 9, 368. [CrossRef]
Materials 2020, 13, 3126 14 of 14

24. Kramer, H.S.; Starke, P.; Klein, M.; Eifler, D. Cyclic hardness test PHYBALCHT - Short-time procedure to
evaluate fatigue properties of metallic materials. Int. J. Fatigue 2014, 63, 78–84. [CrossRef]
25. DIN EN ISO 6507-2. Metallic Materials—Vickers Hardness Test—Part 2: Verification and Calibration of Testing
Machines; NSAI: Dublin, Ireland, 2005.
26. Mises, R.V. Mechanik der festen Körper im plastisch- deformablen Zustand. Nachrichten von der Gesellschaft
der Wissenschaften zu Göttingen, Mathematisch-Physikalische Klasse 1913, 1913, 582–592.
27. Srnec Nova, J.; Benasciutti, D.; De Bona, F.; Stanojević, A.; De Luca, A.; Raffaglio, Y. Estimation of Material
Parameters in Nonlinear Hardening Plasticity Models and Strain Life Curves for CuAg Alloy. IOP Conf. Ser.
Mater. Sci. Eng. 2016, 119, 12020. [CrossRef]
28. Chaboche, J.L.L. Constitutive equations for cyclic plasticity and cyclic viscoplasticity. Int. J. Plast. 1989, 5,
247–302. [CrossRef]
29. Lemaitre, J.; Chaboche, J.-L. Mechanics of Solid Materials; Cambridge University Press: Cambridge, UK, 1990.
30. Frederick, C.O.; Armstrong, P.J. A mathematical representation of the multiaxial Bauschinger effect.
Mater. High Temp. 2007, 24, 1–26. [CrossRef]
31. Sajjad, H.M.; Hanke, S.; Güler, S.; ul Hassan, H.; Fischer, A.; Hartmaier, A. Modelling cyclic behaviour of
martensitic steel with J2 plasticity and crystal plasticity. Materials 2019, 12, 1767. [CrossRef] [PubMed]
32. About LS-OPT—DYNAmore GmbH. Available online: https://www.dynamore.de/de/produkte/opt/ls-opt
(accessed on 26 April 2020).
33. Chaparro, B.M.; Thuillier, S.; Menezes, L.F.; Manach, P.Y.; Fernandes, J.V. Material parameters identification:
Gradient-based, genetic and hybrid optimization algorithms. Comput. Mater. Sci. 2008, 44, 339–346.
[CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).