Int J Fract

https://doi.org/10.1007/s10704-020-00457-z

ORIGINAL PAPER

The differences of damage initiation and accumulation of DP

steels: a numerical and experimental analysis
Felix Pütz · Fuhui Shen · Markus Könemann ·
Sebastian Münstermann

Received: 10 September 2019 / Accepted: 21 May 2020

© The Author(s) 2020

Abstract Many studies have examined the damage 1 Introduction

behaviour of dual-phase steels already. It is a topic of
high interest, since understanding the mechanisms of The usage of dual-phase (DP) steels has been on the
damage during forming processes enables the produc- rise in recent years. For example, the automotive indus-
tion of steels with improved properties and damage tol- try shows high interest in the development of these
erance. However, the focus was rarely on the compari- advanced high strength steels (AHSS), since DP steels
son between representatives of this steel class, and the show high strength values while still maintaining good
numerical simulation for the quantification of damage formability. Thus, a lightweight component design can
states was not thoroughly used. Therefore, this study be achieved by reducing component thickness while
compares the damage initiation and accumulation of still keeping the identical safety conditions (Davies and
two dual-phase steels (DP800 and DP1000), which are Magee 1979). These specific properties result from a
used in the automotive industry. Additionally, param- distinct microstructure, that is composed of a soft fer-
eter sets of a phenomenological damage mechanics ritic phase with hard martensite islands on the grain
model with coupled damage evolution are calibrated boundaries and triple points of ferrite grains. Due to
for each material. The combined analysis reveals an the difference in mechanical properties of the two
earlier initiation of damage for the DP800, where the phases, plastic behaviour of DP steels shows a relatively
damage accumulation phase is prolonged. For DP1000 low yield to tensile ratio, pronounced strain hardening
the damage nucleates only shortly before material fail- and excellent global formability. The reason for this
ure. The material model is able to correctly predict the extraordinary property profile lies in the partitioning of
behaviour, while experimental analysis confirms the stress and strain between the involved phases, allow-
prediction via light optical and SEM metallography. ing for multiple degrees of freedom for microstructural
design (Bieler et al. 2009) .
Keywords Steel · Dual-phase · Ductile damage · The strain partitioning between ferrite and marten-
Damage model · FEM simulation · Damage tolerance site depends vastly on the specific microstructure.
Marteau et al. reported that the local microstructural
neighbourhood is the critical factor for strain het-
erogeneity (Marteau et al. 2013). Strain accumulates
F. Pütz (B)· F. Shen · M. Könemann · S. Münstermann mostly in the ferrite forming localized bands with an
Steel Institute - Integrity of Materials and Structures,
angle of 45–50◦ with respect to the loading direction
RWTH Aachen University, Intzestraße 1, 52072 Aachen,
Germany (Ghadbeigi et al. 2010; Tasan et al. 2014a), whereas
e-mail: [email protected] the martensite carries the majority of the applied stress

123
 Felix et al.

(Tasan et al. 2014b). Therefore, martensite is elasti- et al. 2012). Alternatively, phenomenological, coupled
cally deformed for materials with low martensite con- models are used to describe the damage in materials
tent, while its deformation behaviour is plastic for high numerically. In contrast to the micromechanical mod-
contents (Shen et al. 1986). The local microstructure els, damage evolution is treated in a macroscopic way,
especially determines the strain distribution, e.g. aver- where a number of effects are described by a mathe-
age size of martensite islands and global distribution matical formulation. A good example for this type of
(Park et al. 2014; Saai et al. 2014). model is the Lemaitre model (Lemaitre 1985, 1992),
Due to this inhomogeneity in the material con- which describes damage as an irreversible process.
stituents’ behaviour, the microscopic damage modes of Contrary to that, uncoupled models describe the
dual-phase steels differ quite significantly to those of material behaviour including fracture without taking
common structural steels. Where for structural steels damage into account. Both the Johnson–Cook (Johnson
the inclusions play the major role for void initiation, and Cook 1985), as well as the Bai–Wierzbicki (BW)
in DP steels damage incidents occur in relation to the model are good examples for this type of model (Bai
two phases, martensite and ferrite (Tasan et al. 2010). and Wierzbicki 2008). Further development has been
Mechanisms for damage initiation in dual-phase steels applied by Lian et al. who combined the advantages
are mostly decohesion of the martensite/ferrite inter- of uncoupled and coupled models into a hybrid formu-
face, cracking of the martensite phase, or a localization lation, making it the modified Bai–Wierzbicki model
of plastic strain in the ferrite phase, which results in (MBW) (Lian et al. 2013). The model therefore holds
debonding of the ferrite grain boundaries (Ahmad et al. an easy formulation and combines it with the influ-
2000). The mode for the damage initiation depends ence of damage onto material behaviour. The model
on the microstructure and the resulting strain hetero- has been developed further since its inception. For the
geneity (Kadkhodapour et al. 2011). Therefore, grain first version a locus for the damage initiation point,
size and martensite content do play an important role which was dependent on both stress triaxiality and Lode
(Maire et al. 2008; Ramazani et al. 2013; Tasan et al. angle was utilized. Additionally, a set of critical values
2015). Additionally, martensite morphology influences for the damage variable was applied, at which mate-
the early damage nucleation (Ghadbeigi et al. 2013; rial fracture was assumed in the numerical simulation.
He et al. 1984). Besides, observations have shown, that Wu et al. changed that considerably by implementing
for banded martensite, cracking is far more likely than a locus for the fracture, as well as considering non pro-
a decohesion of the interface boundary of ferrite and portional loading paths until the inception of ductile
martensite (Avramovic-Cingara et al. 2009). damage (Wu et al. 2017). A further development of the
To assess the material’s properties and predict MBW model was made by Shen et al. to characterise
the load bearing capabilities of structures, damage the influence of loading orientation, which was used
mechanics models are widely used for DP steels, e.g. to describe the anisotropic ductile damage and fracture
in the automotive industry. In the field of damage behaviour of pipeline steels (Shen et al. 2020). Since
mechanics, two different model types exist: Coupled the MBW damage mechanics model is easy to use and
and uncoupled models (Besson 2010). For the cou- calibrate, while also depicting the damage behaviour
pled damage mechanics models usually a damage vari- accurately, it is applied here for the characterisation of
able is employed to reduce the yield potential accord- damage behaviour in DP steels.
ing to the softening resulting from ductile damage in While many studies focused on the damage in dual-
the material during deformation. In case of the cou- phase steels from an experimental standpoint, it is hard
pled models, micromechanical models are very pop- to experimentally determine the evolution of damage
ular, for instance the Gurson–Tvergaard–Needleman during the tests. Therefore, this study aims to enhance
(GTN) model (Gurson 1977; Tvergaard 1981; Tver- the experimental investigation by performing finite ele-
gaard and Needleman 1984). Micromechanical mod- ment (FE) based numerical simulations that are utilized
els are characterized by the depiction of physical phe- to quantify the damage in the material during forming
nomena like void nucleation, growth and coalescence processes. Thus, in this study, two dual-phase steels,
through sets of parameters. Therefore, the parameters DP800 and DP1000 were compared. Their damage
are interdependent and thus, an extensive iteration pro- and fracture properties are distinctly different, while
cess is necessary for the parameter calibration (West their strengths are similar. To compare the materials

123
The differences of damage initiation and accumulation of DP steels...

Fig. 1 Microstructures of
steels DP800 and DP100
revealed by HNO3 etching,
in light optical
metallography

behaviour, a damage mechanics model has been used the microstructure is noticeable. Since the martensite
that can describe both, damage initiation as well as bands run parallel to the rolling direction, there will
ductile material fracture, while also taking the changes be a significant influence on the mechanical properties.
of the stress state during deformation, due to non- A certain extent of failure anisotropy is expected due
proportional loading effects, into account . This allows to the banded microstructure, however, the anisotropic
a comparison of the damage initiation for different fracture properties are beyond the scope of this study
stress states between the materials. Additionally, by and all tensile specimens were manufactured perpen-
means of a calibrated fracture locus, the damage accu- dicular to the rolling direction of both DP steels. Both
mulation phase can be analysed and compared. Thus, steels were delivered with a thickness of 1.5 mm; their
tensile tests were conducted on flat specimens of dif- respective chemical compositions are given in Table 1.
ferent geometries to gather information about materi- While the alloying concepts show noticeable similari-
als deformations and damage properties under differ- ties, some minor differences are present.
ent stress states. On that basis, the material parameters The carbon content for DP1000 is lower compared
of the modified Bai Wierzbicki model were calibrated. to DP800, thus leading to higher carbon concentration
For the validation of the numerical results regarding the in the martensite phase for the DP800, since the phase
damage initiation and accumulation of the investigated fraction of martensite is higher for DP1000. Thus, it
material, interrupted tensile tests were conducted and a is to be expected that the strength of the martensite is
metallography analysis was performed using the light reduced for DP1000 due to the decreased carbon con-
optical microscopy and scanning electron microscopy tent, therefore leading to a bigger contrast of proper-
(SEM). ties between ferrite and martensite in the DP800. Fur-
thermore, manganese and chromium contents are dif-
ferent, which leads to slight disparities due to solid
2 Materials characterization solution hardening. Additionally, Si as well as Mn and
Cr reduce the critical cooling rate needed for form-
In the present study, two dual-phase steels were evalu- ing martensite, thus influencing the respective time–
ated for comparison purpose. Even though both materi- temperature–transformation graphs. On top of that the
als are dual-phase steels, vastly different properties are solubility of carbon in ferrite is reduced by silicon.
observable. These varying characteristics are obtained Therefore, both materials will have very distinct pro-
by distinct alloying concepts as well as heat treatment cessing routes tailored to the respective production pro-
processes. cess. For the characterization of the resulting mechan-
Figure 1 shows a comparison of the respective ical properties, isothermal, uniaxial quasistatic tensile
microstructures at a magnification of 1000. It is tests were carried out on flat specimens without a notch.
observed that the average grain size of DP1000 is sig- To ensure a proper depiction of the material’s proper-
nificantly smaller than that of DP800. Additionally, ties, three tests were carried out per DP steel. A video
DP1000 has increased martensite contents of approx- extensometer was used to capture the elongation of the
imately 38% while DP800 contains about 32%. For specimen during deformation, where the starting length
DP800 a pronounced banding of the martensite in of 40 mm was used. The necking took place inside the

123
 Felix et al.

Table 1 Chemical composition of dual-phase steels DP800 and DP1000, in mass-%

C Si Mn Cr Mo Cu

DP800 0.15 0.21 1.67 0.73 0.01 0.044

DP1000 0.14 0.32 1.97 0.40 0.05 0.023

development of the existing MBW model was required

to determine the material behaviour and damage accu-
mulation more realistically.

3 Extension of MBW model for non-proportional

loading paths (npMBW-19)

In the framework of continuum damage mechanics,

the modified Bai Wierzbicki (MBW) model has been
proposed by Lian et al. (2013) and widely applied to
describe the damage and fracture behaviours of vari-
ous grades of steels (Lian et al. 2013; Münstermann
Fig. 2 Engineering stress-strain curve of uniaxial tensile tests et al. 2017; Wu et al. 2017; Novokshanov et al. 2015;
of DP800 and DP1000 Shen et al. 2020; Liu et al. 2020). Like in other damage
mechanics models, the significant influences of stress
state on the ductile fracture are considered through
area tracked by the extensometer for all 6 tensile tests, defining a strain based criterion which is usually a
ensuring a good comparability. weighted function of two particularly important vari-
The results for both materials are shown in Fig. 2. ables, the stress triaxiality η and the Lode-angle param-
The scatter shown is the difference in fracture, result- eter θ that are related to the three stress invariants.
ing from the three tensile tests per material mentioned
before. From this figure it is obvious, that DP1000 I1 = tr [σ ] = (σ1 + σ2 + σ3 ) (1)
shows higher strength, while the strain at fracture of 1 1
J2 = [σ ]2 = [(σ1 − σ2 )2 + (σ2 − σ3 )2 + (σ3 − σ1 )2 ] (2)
DP800 is considerably higher. The higher strength is a 3 6
 I1   I1   I1 
result of both; the higher martensite phase fraction as J3 = σ1 − · σ2 − · σ3 − (3)
3 3 3
well as significantly refined grains in DP1000. In addi- I1 I1
η= √ =
tion, there is a clear scatter in the elongation values, 3 · 3 · J2 3σ
with elongation at fracture between 18 and 23 percent (σ1 + σ2 + σ3 ) (4)
= 
for DP800, while the scatter for the DP1000 is about 3·
1
[(σ1 − σ2 )2 + (σ2 − σ3 )2 + (σ3 − σ1 )2 ]
one percent. This variation in elongation at fracture can 2
 √ 
be explained by the clearly pronounced band structures, 1 3 · 3 · J3
θ = cos −1 3/2
(5)
which can lead to significant deviations depending on 3 2 · J2
the position of the bands in the specimen. Since DP1000 θ =1−
6θ
(6)
fractures shortly after uniform strain, the necking is far π
less pronounced than in DP800. During necking the where σ1 , σ2 and σ3 are principal stresses and σ is the
stress state in the sample can change quite significantly von Mises equivalent stress. For the material model, the
leading to non-proportional loading paths during the Lode-angle parameter θ was used, which has a linear
deformation of the sample. Thus, to describe the mate- relationship with the Lode-angle θ .
rials behaviour after the uniform strain, it is necessary The effects of stress state on plasticity in some metal-
for the material model to consider the effects of the lic materials have been reported, while steels typically
changes of stress state during deformation. Therefore, a show a negligible pressure sensitivity, therefore, only

123
The differences of damage initiation and accumulation of DP steels...

the effects of Lode-angle parameter are considered in the stress state for non-proportional loading paths (Wu
the yield criterion of the MBW model. et al. 2017; Mu et al. 2020).

 εp
Φ = σ (σ ) − (1 − D) · σ y (ε p , θ ) ≤ 0 (7) 1
  ηavg = p η(ε p )dε p (12)
γ m+1  ε
σ y (ε p , θ) = σ y (ε p ) · cθs + (cθax − cθs ) · γ − (8) 0
m+1
√
3 θ ·π
γ = √ · sec −1 (9)
2− 3 6  εp
1
cθt , for θ ≥ 0. θ avg = θ (ε p )dε p (13)
cθax = (10) εp 0
cθc , for θ < 0.
Since the damage is dependent on stress state it is
where D is a scalar variable to quantify the damage necessary to define equations for the initiation of dam-
effects, yield stress σ y is determined by the equiva- age, as well as the fracture, that represent this depen-
lent plastic strain ε p and Lode-angle parameter θ , and dency. Therefore, the damage initiation locus (DIL)
σ y (ε p ) corresponds to the flow stress at given equiva- and ductile fracture locus (DFL) have been defined as
lent plastic strain ε p un‘s state. cθs , cθt , cθc are the nor- two individual equations f di and f d f with the stress
malised strength under shear, tension and compression triaxiality and the Lode-angle parameter as indepen-
state and m is a material parameter with positive inte- dent variables. The instantaneous and average values
gral values that describes the Lode-angle sensitivity. γ of the independent stress state variables have been used
is another stress state parameter with unique correla- in the damage and fracture criteria under proportional
tion to the Lode-angle parameter θ. According to the and non-proportional loading conditions, respectively.
derivation of Lian et al. (2013), the yield locus of MBW Under non-proportional loading conditions, these two
model is convex if the material parameters are located equations describe the critical equivalent plastic strains
√ at the moment of damage initiation and ductile fracture,
cs
within the specific range of 2 ≤ caxθ ≤ 1. The con-
3
θ respectively.
ventional normality rule is applied in the MBW model
1
and the plastic strain components are updated accord- f di (ηavg , θ avg ) = (D1 e−D2 ηavg + D5 e−D6 ηavg )
2
ing to the following equation and dλ is a non-negative  1
−D3 e−D4 ηavg θ avg + (D1 e−D2 ηavg
2
plastic multiplier. 2
−D5 e−D6 ηavg )θ avg + D3 e−D4 ηavg (14)
1
f d f (ηavg , θ avg ) = (F1 e−F2 ηavg + F5 e−F6 ηavg )
2
δΦ 
dε p = dλ · (11) 1
−F3 e−F4 ηavg θ avg + (F1 e−F2 ηavg
2
δσ 2
−F5 e−F6 ηavg )θ avg + F3 e−F4 ηavg (15)
In the coupled damage mechanics model, two indi- where D1 –D6 and F1 –F6 define the damage initiation
vidual criteria have been defined to identify the duc- locus and ductile fracture locus. Under the condition
tile damage initiation (DDI) and ductile fracture (DF), that D1 = D5 , D2 = D6 and F1 = F5 , F2 = F6 , the
which corresponds to the initiation of degradation on DIL and DFL are symmetric with respect to the Lode-
microscopic scale in the material and the loss of load angle parameter and four independent parameters are
carrying capacity on the macroscopic scale. Numeri- enough to define the corresponding loci. Based on pre-
cally, damage initiation, in this study, is defined as the vious experimental observations, a cut-off value of the
onset of macroscopic softening due to damage, which stress triaxiality ηc exists, below which the initiation
must be taken into account by the numerical represen- and evolution of ductile damage cannot be triggered
tation of the material behaviour. In order to consider due to pressure effects. ηc = −1 3 as a reasonable esti-
the change of stress state during plastic deformation, mation has been adopted in the MBW model (Wu et al.
the average values of the stress triaxiality ηavg and the 2017). Therefore, when the stress triaxiality is lower
Lode-angle parameter θ avg have been used to describe than ηc , the equations f di and f d f are set to be infinite.

123
 Felix et al.

The damage initiation specified by this model is unre- In summary, the damage evolution is determined by
lated to the materials mechanisms of damage initiation, the two independent damage initiation and ductile frac-
e.g. micro crack formation, void formation. Instead, it ture criteria. After a certain damage nucleation period,
aims to describe the aggregative accumulation of the which is controlled by plastic deformation, damage
defects and their influence on the load bearing capabil- evolution takes place. Once the indicator of the ductile
ities. For this step a plasticity model is no longer able fracture Id f reaches unity, the final crack propagation
to describe the materials mechanical behaviour (Keim is triggered and failure occurs. Therefore, the model,
et al. 2019). For the non-proportional loading, two indi- hereafter called npMBW-19, is capable of represent-
cators have been applied to describe the ductile dam- ing the influence of the necking, and thus the change
age initiation Idd and ductile fracture Id f respectively of stress state, during deformation.
to consider the effects of stress state evolution.
 εp
dε p
Idd = p with 4 Calibration of the new model for materials
0 ε di (ηavg , θ avg )
DP800 and DP1000
p +∞, ηavg ≤ ηc
ε di (ηavg , θ avg ) = (16)
f di (ηavg , θ avg ), ηavg > ηc . The calibration approach for the material models for
 εp
dε p both steels follows roughly the approach of (Lian et al.
Id f = p with
εdi
p,c
εd f (ηavg , θ avg ) (2013, 2014)). Since the calibrated npMBW-19 model
needs to be able to account for various stress states,
p +∞, ηavg ≤ ηc
ε d f (ηavg , θ avg ) = (17) the calibration of the material model is carried out on
f d f (ηavg , θ avg ), ηavg > ηc .
a variety of sample geometries. By varying the sample
The values of equivalent plastic strain and equivalent geometries in tensile tests, different stress states can be
stress at the moment of damage initiation (Idd = 1) are accomplished. In this study three differently notched
p,c
defined as two characteristic variables εdi and σ cdi , specimen types were applied for the calibration of the
respectively: material model in addition to the uniaxial tensile test.
p,c
εdi = ε p (Idd = 1) (18) Used specimen types were: Notched dogbone samples
(varying notch geometries at the edge of the sample),
σ cdi =σ (Idd = 1) (19)
central hole samples (round, as well as elliptical holes
After the damage initiation criterion is fulfilled, in the center of the specimen) and plane strain samples
damage evolution is controlled according to the energy (notch with different radii over the thickness of the sam-
dissipation theory. Depending on the shape of dam- ple). The applied specimens for each material can be
age initiation locus and ductile fracture locus, when seen in Figs. 3 and 4 The type of notch of the sample
the indicator of the ductile fracture Id f reaches unity, is abbreviated with an r continuing with the radius, for
the damage variable D does not necessarily reach unity. the notched dog bone samples.
Therefore, a critical value of the damage variable Dcr The corresponding stress states, characterised by the
exists, at which the material point will fail regardless Lode-angle parameter and the stress triaxiality in the
of the value of the D variable: applied samples are delineated in Table 2. To achieve
multiple stress states, notches were modified with var-
σ cdi  p p ious radii to gain geometries of different stress states
Dcr = ε d f − ε di (20)
Gf within one sample type. Per specimen geometry, three
tensile tests were performed in accordance with the
Where G f is a material parameter which controls the procedure described earlier for the uniaxial tensile
damage evolution rate. Linear damage evolution is test. Afterwards, simulations of the experiments were
assumed in the MBW model, which is expressed as: conducted, using ABAQUS, to achieve a comparison
between the force–displacement curves of experimen-
⎧ tally determined values and simulated ones.
⎪
⎨0, Idd < 1 For the determination of the base flow curve, the uni-
D = Dcr · Id f , Idd ≥ 1 ∧ Id f < 1 (21) axial tensile tests (T), presented in the previous chapter,
⎪
⎩
1, Idd ≥ 1 ∧ Id f ≥ 1 were utilised. From the determined engineering stress-

123
The differences of damage initiation and accumulation of DP steels...

Fig. 3 Applied tensile specimens for steel DP800

Fig. 4 Applied tensile specimens for steel DP1000

Table 2 Stress states of utilised sample geometries

Sample geometry Lode-angle parameter θ Stress triaxiality η
1
Uniaxial tensile (UT) 1 3
Notched dog bone (DB) 0.3–0.8 0.4–0.6
Central hole (CH) –1 0.3–0.4
Plane strain (PS) 0 0.5–0.7

strain curve, the true stress-true strain curve was calcu- tool.
lated until the uniform elongation point. This data was
then used to fit the Hollomon–Voce hardening model σ = α · (K εnp ) + (1 − α) · (A − B · e−Cε p ) (22)
to the material’s flow curve via the Matlab curve fitting

123
 Felix et al.

Table 3 Hollomon–Voce fitting parameters for steels DP800 and DP1000

α K n A B C

DP800 0.5138 1843 0.44 1167 820.4 100

DP1000 0.5879 2000 0.1127 725.5 300 57.2

This specific hardening model was chosen, since points. For non-proportional loading paths it is neces-
it shows a good compromise between accurate repre- sary to average the stress state of the critical element,
sentation at low plastic strains and realistic hardening where damage happens first, over the simulated steps
behaviour for higher strains. In Table 3 the parame- (Wu et al. 2017).
ters for the Hollomon–Voce models are given for both, After determining the locus for the onset of damage
DP800 and DP1000. (DIL), the effect that damage has on the component
After the calibration of the flow curve the basic needs to be adjusted. In the npMBW-19 model parame-
parameters of the MBW model were determined (cθs , ter G f is calibrated to adjust the speed at which damage
cθt , cθc , m). This was done by iterating over multiple accumulates in the simulated material. G f is defined as
simulations using a range of different sample geome- the energy dissipation between damage initiation and
tries. complete fracture. When the softening is specified the
Subsequently the damage and fracture parameters of fracture locus can be determined. The approach used
the npMBW-19 model were determined. Damage and for this determination follows the one from the dam-
fracture criteria in this material model are described by age initiation locus closely. This time the point for the
Eqs 14 and 15. Therefore, the specified locus needs to experimental fracture is compared to the simulation.
be calibrated for both events, damage initiation and The step where the fracture should occur is identified
fracture (Lian et al. 2013; Wu et al. 2017). For the and Lode-angle parameter, stress triaxiality and equiva-
damage initiation locus, a comparison of force and dis- lent plastic strain are extracted for the critical elements.
placement curve between simulation and experimental Again, the stress states are averaged from the point of
results was used. Since the damage described in this damage initiation to the presumed fracture of the sam-
model is related to the accumulated damage incidents, ple. After gathering the data for all sample geometries
a threshold method has been utilised to find the numer- the locus is fitted in regards to the obtained points using
ical damage initiation. For that reason, the numerical the Matlab curve fitting tool.
onset of damage was determined as the point where the In Fig. 5 the final results of this calibration process
deviation between simulated and experimental force are depicted for steel DP800. From this figure it is obvi-
and displacement curves was apparent. Similar meth- ous that a good match between experimental data and
ods have been used by other authors within the damage simulations was obtained. The scatter for the experi-
mechanics field (Børvik et al. 2001; Bouchard et al. mental testing can be seen in the shaded areas.
2011). Likewise, the material model for DP1000 was cali-
At this step, the Lode-angle parameter and stress tri- brated. The applied flow curve can be seen in Table 3.
axiality as well as the equivalent plastic strain (PEEQ) Additionally, a damage initiation, as well as a ductile
are taken from the simulation. Since these parameters fracture locus were calibrated using the same approach
may vary locally, the element is chosen that shows the as described above for the DP800. By duplicating the
most critical state of stress and thus is most likely to approach stated above, good agreement with the exper-
encounter damage first. By extracting the Lode-angle imental data was reached (Fig. 6). Contrary to the
parameter, stress triaxiality and equivalent plastic strain DP800 almost no scatter could be found during the
for a multitude of different tensile geometries, data tests of the DP1000 material, which also shows no sig-
points are gathered in the space defined by these three nificant banding in its microstructure. The applied set
variables. Applying the curve fitting tool of Matlab, of parameters can be found in Table 4. Interestingly, the
a function can be defined that describes the desired calibrated G f parameter for DP1000 is higher which
surface while using the obtained results as supporting results in a slower development of the damage variable.

123
The differences of damage initiation and accumulation of DP steels...

capacity (Lemaitre 1992), it is not to be equated with

component failure. Damage occurs on a microscale and
is usually described as the development of voids inside
the microstructure, while on a macroscale damage usu-
ally equates to cracks in the component and therefore
can be seen as component failure. It is therefore highly
relevant to differentiate between micro and macroscale
(Tekkaya et al. 2017). For numerical analysis, damage
is defined as the macroscopic reduction of the stress
during loading, that cannot be described by basic plas-
ticity modelling. Thus, Lemaitre introduced a factor for
Fig. 5 Comparison between experimentally obtained data
(background and lighter color) and simulation results for DP800 damage in a microstructure, which results in a reduc-
tion of the flow potential by the term (1 − D), where D
is the damage variable (Lemaitre 1985). The damage
variable adopted by Lian et al. shows some differences
to the one postulated by Lemaitre. While Lemaitre’s
damage variable is calculated based on the area frac-
tion of defects, Lian et al. refer to the stress at damage
initiation, divided by the energy required to create new
surfaces in a volume of the material (see G f ), an adap-
tation of the damage evolution law used by ABAQUS
finite element code (Lian et al. 2013). Accordingly,
both damage variables are scalar, but there are quite
pronounced differences between both numerical dam-
Fig. 6 Comparison between experimentally obtained data age rules. These differences between the damage mod-
(background and lighter color) and simulation results for DP1000
els must be distinguished, as well as the differences
between micro- and macroscopic damage phenomena.
This results in a fairly slow accumulation of damage Because the damage law used in the MBW model
after the initiation. does not refer to a physical material characteristic,
like the area fraction of voids, except for the energy
for cracks, it is a more phenomenological approach to
model the influence of damage on the materials flow
5 Damage behaviour prediction
potential. The damage evolution, as discussed before,
starts when a specific equivalent plastic strain (PEEQ)
For the scope of this study, it is important to differenti-
locally exceeds a certain threshold, which changes with
ate between failure and damage of a material or compo-
stress states. The respective value for PEEQ is deter-
nent. Because damage is the deterioration of materials
mined by the damage initiation locus. After this point,
properties before failure, especially the load bearing

Table 4 Calibrated npMBW-19 parameter sets for both materials

cθs cθt cθc m D1 D2 D3

DP800 0.95 1 0.9 6 0.5 2 0.365

DP1000 0.95 1 0.97 6 0.4 1 0.1
 J

D4 Gf mm 3
F1 F2 F3 F4

DP800 3 1.2 0.7 1 0.366 2

DP1000 1.5 6.5 0.58 0.76 0.443 1.57

123
 Felix et al.

Fig. 7 Comparison of the ductile damage initiation locus and the ductile fracture locus for DP800 (left) and DP1000 (right)

softening occurs in the simulation, which leads to a

direct reduction of the resulting stress compared to
virgin materials. The length of this following phase
where damage accumulates depends on the stress state
which is considered in the ductile fracture locus. The
comparisons of the ductile damage initiation locus and
the ductile fracture locus for each respective material
are shown in Fig. 7. The shape of the loci for DP800
and DP1000 are different, as was to be expected. The
distance between the plots is higher for the DP 800
material which leads to a longer damage accumulation
phase. Merely for higher triaxialities and Lode-angle
Fig. 8 Flow curves from uniaxial tensile test of DP800 and
parameters around zero, the differences between the DP1000 with numerically determined points of damage initia-
loci of DP800 and DP1000 is minimal. Some research tion and fracture under uniaxial tension condition
suggests a different shape for the ductile damage initia-
tion locus and the ductile fracture loci, especially for the
area around a stress triaxiality of 0, namely shear stress strain for both materials, while fracture is delayed sig-
state (Papasidero et al. 2015). Nevertheless, based on nificantly for DP800.
the experimental and numerical results in this study, the To further examine precision of the numerical
loci in Fig. 7 constructed for both steels using the corre- results, interrupted tensile tests were conducted for both
sponding calibrated damage and fracture parameters is materials. For each material a sample was therefore first
validated within the range of investigated stress states. tested until failure and subsequent specimens of iden-
In the case of an application of the calibrated material tical geometry were stopped after a distinct strain was
model for even lower or higher stress triaxialities, the reached. The lowest elongation used in this investiga-
loci would have to be revisited to confirm or adapt their tion was the uniform elongation, as no or little dam-
shape. age is expected below this. This way a metallographic
Due to the differences in the distance of the DIL damage analysis could be carried out to investigate the
and the DFL the damage accumulation phase is signif- average amount of damage that could be observed in a
icantly different between the two steels. Fig. 8 shows sample. For both materials unnotched dog bone speci-
this difference utilising the flow curves obtained from mens were utilised to ensure a good comparability. For
the uniaxial tensile test of both materials, as well as the analysis of the damage in the material, light opti-
calculating the points for damage initiation and frac- cal microscopy was chosen, since a bigger area can
ture under uniaxial tension condition (η = 13 , θ = 1) be investigated by light optical analysis, where scan-
based on calibrated material parameters. The point for ning electro microscopy (SEM) pictures resolve only
the damage initiation takes place at roughly the same smaller areas of the samples. However, it is not easily
possible to differentiate between voids and inclusions

123
The differences of damage initiation and accumulation of DP steels...

Fig. 9 Comparison of light optically detected area fraction for

DP800 and DP1000 for different strains

Fig. 10 Evolution of damage in DP1000. Single martensite

in the material. Thus the area fraction that is detected crack with magnification of 5000 for 0.7 of fracture strain
is not quantitatively representative of the actual void
fraction. To find out about the area fraction for each
sample, multiple pictures were taken to gather infor-
mation about the scatter band where the actual values
lie. For this analysis, the light optical pictures were con-
verted to greyscale images, which were subsequently
evaluated by a threshold method, with which a differ-
entiation between matrix material and voids/inclusion
could be made. For these steps Fiji was used as image
analysis software (Rueden et al. 2017; Schindelin et al.
2012).
The results of this analysis are depicted in Fig. 9. To
better compare the values for both materials, a normal-
isation was carried out, where the current strain was
divided by the respective fracture strain. A comparison
Fig. 11 Evolution of damage in DP1000. Many voids have
of the values for the detected area fraction reveals a formed and grown, magnification of 2000, 0.95 of fracture strain
gradual increase for DP800, while for DP1000 no sig-
nificant rise in fraction can be observed until just before
fracture of the sample. The large scatter, especially at (Fig. 10). While the amount of these initiation locations
the beginning, can be explained by the lack of neck- increases with the strain, growth is very limited (Fig.
ing, which means that the region of interest cannot be 11).
identified accurately. A comparison of the samples that are at 95% of frac-
Thus, the damage accumulation phase for DP800 ture strain reveals that the voids in DP1000 (Fig. 11) are
starts at lower strains relative to the fracture strain of the significantly smaller than in DP800 (Fig. 12). It is there-
material. By contrast, the damage accumulation phase fore concluded, that the damage accumulation phase
for DP1000 starts very late and just before fracture. for DP1000 is indeed significantly shortened. In par-
Therefore, the damage in the material behaves exactly ticular, it is noticeable that no void in DP1000 exceeds
as predicted using the npMBW-19 model. To assess a length of 1 µm , while the DP800 features multiple
the damage state in the microstructure, pictures were larger voids. Additionally, voids for DP800 are more
taken in the SEM. Especially for DP1000, an analy- circular, while they are shaped like cracks for DP1000
sis for higher magnifications was necessary to reveal if again leading to the conclusion that there has been no
damage forms earlier than shortly before fracture. For time for growth after initiation. This is in line with the
70% of the fracture strain, only very few events of dam- results demonstrated in Fig. 8, where a shorter dam-
age initiation could be found under high magnification age accumulation phase is present in DP1000 and thus

123
 Felix et al.

The found differences during the experimental test-

ing and analysis can be attributed to the differences
in the microstructure. Microstructural features, espe-
cially grain size and martensite content, but also the
pronounced banding in the DP800 play an important
role for mechanical properties, as well as damage initi-
ation and accumulation. In this study, it was shown, that
the damage in both dual-phase steels initiates at sim-
ilar equivalent plastic strains. Oppositely, the fracture
happens at vastly different equivalent plastic, as well as
global strains. This leads to completely different dam-
age accumulation phases in the material. The numerical
simulations showed an exceedingly different length of
Fig. 12 Damage shortly before fracture in DP800 the damage accumulation phase for the two steels. This
difference was subsequently verified by experimental
tests, where the amount of damage in the material after
a lower decrease of load bearing capabilities is to be an interrupted tensile test was examined. For these tests
expected. it could be shown that DP800 exhibits a pronounced
Besides, the damage initiation modes were investi- damage accumulation phase, while DP1000 fractures
gated. For both materials, the prevalent modus for dam- shortly after a critical amount of voids forms in the
age initiation was the cracking of martensite islands. material. Thus fracture occurs with almost no dam-
For DP800, the martensite bands, especially, were sites age accumulation and the void growth phase is nearly
for damage initiation. Furthermore, decohesion of fer- skipped. Therefore, the different microstructures lead
rite and martensite islands was found in the DP800 after to specific damage characteristics, which in turn influ-
about 80% of fracture strain. ence and change the specific properties of the material.
Additionally, the contrast in the mechanical properties
between the two phases for both materials reinforce this
effect. Since the carbon content in martensite is rela-
6 Conclusions tively higher in DP800 than in DP1000, the martensite
fractures earlier leading to a relatively early damage
This study showed significant differences between two initiation and longer damage accumulation phase. For
industrially produced dual-phase steels. Starting with the DP1000 the the contrasts are not so distinct, which
the experimental results, the difference in fracture leads to a comparatively late initiation of damage and
strain was found to be significant with large scatter failure shortly afterwards.
for the DP800. This scatter was attributed to the pro- During the comparison of experimental and numeri-
nounced banding found within the material. For the cal results, it was quite obvious that the presented mate-
numerical analysis it was found that the change in the rial model is able to accurately represent the experimen-
stress state during necking needs to be considered by tal tensile test curves, both uniaxial and notched spec-
the material model for proper simulation results. There- imens. The stress states do, however, show only minor
fore, the MBW model was extended to account for variance. Thus, for higher deviations, an adjustment of
non-proportional loading paths. The parameters of the the averaging scheme for determining stress triaxial-
model were then fitted for both materials to reveal the ity and Lode angle parameter for the ductile damage
disparities in the material behaviour numerically. Sub- and fracture loci might be appropriate. Furthermore,
sequently, the damage initiation and fracture loci were the model shows a discontinuity around the value of
calibrated. The comparison of simulation results to ηc = −1 3 ; as a result it will be difficult to deal with, when
the experimentally obtained force-displacement curves getting close to it (e.g. η = −0.32). However, since the
reveals a high agreement for both materials. In particu- loading paths in this study are exclusively above this
lar, the differences in damage behaviour were modelled value, this is a promising and important concern for
precisely. future development.

123
The differences of damage initiation and accumulation of DP steels...

The comparison of numerical and experimental duc- author(s) and the source, provide a link to the Creative Com-
tile damage showed that the presented material model mons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s
is able to accurately predict the damage initiation, dam- Creative Commons licence, unless indicated otherwise in a credit
age accumulation and fracture of both materials. Nev- line to the material. If material is not included in the article’s Cre-
ertheless, the accuracy of the damage initiation point ative Commons licence and your intended use is not permitted by
in the material model is still an important topic for statutory regulation or exceeds the permitted use, you will need
to obtain permission directly from the copyright holder. To view
further investigation. Since the location of the fracture a copy of this licence, visit http://creativecommons.org/licenses/
locus strongly depends on the damage initiation locus, by/4.0/.
a high precision for the DIL is desirable. However, the
commonly used method of direct current potential drop
(DCPD) is not feasible for DP steel, since its void vol-
ume is comparably low. Therefore, an improvement of References
the method to determine aforementioned damage initi-
ation locus is necessary and currently examined. One Ahmad E, Manzoor T, Ali KL, Akhter JI (2000) Effect of
microvoid formation on the tensile properties of dual-phase
possibility is to measure the density of the material to
steel. J Mater Eng Perform 9(3):306–310
determine the time of damage initiation (Hering et al. Avramovic-Cingara G, Ososkov Y, Jain MK, Wilkinson DS
2019; Schowtjak et al. 2019; Meya et al. 2019). (2009) Effect of martensite distribution on damage
The analysis of damage initiation point and dam- behaviour in dp600 dual phase steels. Mater Sci Eng A
516:7–16
age accumulation by light optical microscopy is rather Bai Y, Wierzbicki T (2008) A new model of metal plasticity
qualitative than quantitative. Since statistical represen- and fracture with pressure and lode dependence. Int J Plast
tativeness and accuracy have to be balanced for this 24:1071–1096
type of examination, micro voids are not detected in Besson J (2010) Continuum models of ductile fracture: a review.
Int J Damage Mech 19(1):3–52. https://doi.org/10.1177/
the pictures. Thus, the values received are not the void 1056789509103482
area fraction. On top of that, light optical pictures show Bieler TR, Eisenlohr P, Roters F, Kumar D, Mason DE, Crimp
inclusions in a similar colour to voids; isolation of voids MA, Raabe D (2009) The role of heterogeneous deforma-
for analysis purpose is rather difficult for light optical tion on damage nucleation at grain boundaries in single
phase metals. Int J Plast 25(9):1655–1683. https://doi.org/
microscopy. For a more quantitative result of void area 10.1016/j.ijplas.2008.09.002
fraction SEM pictures with high resolution over a big Børvik T, Hopperstad O, Berstad T, Langseth M (2001) A com-
area seem to be more promising. putational model of viscoplasticity and ductile damage for
The presented results suggest that the damage, and impact and penetration. Eur J Mech A Solids 20(5):685–
712. https://doi.org/10.1016/S0997-7538(01)01157-3
therefore the materials mechanical properties depend Bouchard PO, Bourgeon L, Fayolle S, Mocellin K (2011) An
on the microstructure of the respective steel. It is there- enhanced lemaitre model formulation for materials process-
fore of high interest to investigate the influence of ing damage computation. Int J Mater Form 4(3):299–315.
each microstructural parameter on the damage char- https://doi.org/10.1007/s12289-010-0996-5
Davies RG, Magee CL (1979) Physical metallurgy of automotive
acteristics as well as the mechanical properties of high-strength steels. J Metals 31(11):17–23. https://doi.org/
the material. Three-dimensional representative volume 10.1007/BF03354565
elements seem to be a promising approach to inves- Ghadbeigi H, Pinna C, Celotto S, Yates JR (2010) Local plastic
tigate the influence of different microstructural char- strain evolution in a high strength dual-phase steel. Mater
Sci Eng A 527:5026–5032
acteristics, like martensite volume content, martensite Ghadbeigi H, Pinna C, Celotto S (2013) Failure mechanisms in
morphology and grain size. dp600 steel: initiation, evolution and fracture. Mater Sci
Eng A 588:420–431. https://doi.org/10.1016/j.msea.2013.
Acknowledgements Open Access funding provided by Pro- 09.048
jekt DEAL. The funding of this research by the Deutsche Gurson AL (1977) Continuum theory of ductile rupture by void
Forschungsgemeinschaft (DFG, German Research Foundation)— nucleation and growth: part I—yield criteria and flow rules
Projektnummer 278868966—TRR 188, is gratefully acknowl- for porous ductile media. J Eng Mater Technol 99(1):2.
edged. https://doi.org/10.1115/1.3443401
He XJ, Terao N, Berghezan A (1984) Influence of marten-
Open Access This article is licensed under a Creative Com- site morphology and its dispersion on mechanical prop-
mons Attribution 4.0 International License, which permits use, erties and fracture mechanisms of fe-mn-c dual phase
sharing, adaptation, distribution and reproduction in any medium steels. Metal Sci 18(7):367–373. https://doi.org/10.1179/
or format, as long as you give appropriate credit to the original 030634584790419953

123
 Felix et al.

Hering O, Dunlap A, Tekkaya A, Aretz A, Schwedt A Park K, Nishiyama M, Nakada N, Tsuchiyama T, Takaki S (2014)
(2019) Characterization of damage in forward rod Effect of the martensite distribution on the strain hardening
extruded parts. Int J Mater Form. https://doi.org/10.1007/ and ductile fracture behaviors in dual-phase steel. Mater Sci
s12289-019-01525-z Eng A 604:135–141. https://doi.org/10.1016/j.msea.2014.
Johnson GR, Cook WH (1985) Fracture characteristics of three 02.058
metals subjected to various strains, strain rates, temperatures Ramazani A, Schwedt A, Aretz A, Prahl U, Bleck W (2013)
and pressures. Eng Frac Mech 21(1):31–48. https://doi.org/ Characterization and modelling of failure initiation in dp
10.1016/0013-7944(85)90052-9 steel. Comput Mater Sci 75:35–44
Kadkhodapour J, Butz A, Ziaei Rad S (2011) Mechanisms of Rueden CT, Schindelin J, Hiner MC, DeZonia BE, Wal-
void formation during tensile testing in a commercial, dual- ter AE, Arena ET, Eliceiri KW (2017) Imagej2:
phase steel. Acta Mater 59:2575–2588 Imagej for the next generation of scientific image
Keim V, Nonn A, Münstermann S (2019) Application of the data. BMC Bioinform 18(1):529, https://doi.org/10.1186/
modified bai-wierzbicki model for the prediction of ductile s12859-017-1934-z, journal Article, arXiv:2918.7165
fracture in pipelines. Int J Press Vessels Piping 171:104– Saai A, Hopperstad OS, Granbom Y, Lademo OG (2014) Influ-
116. https://doi.org/10.1016/j.ijpvp.2019.02.010 ence of volume fraction and distribution of martensite
Lemaitre J (1985) A continuous damage mechanics model for phase on the strain localization in dual phase steels. Proce-
ductile fracture. J Eng Mater Technol 107:83–89 dia Mater Sci 3:900–905. https://doi.org/10.1016/j.mspro.
Lemaitre J (1992) A course on damage mechanics. Springer Ver- 2014.06.146
lag, Berlin Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M,
Lian J, Sharaf M, Archie F, Münstermann S (2013) A hybrid Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B,
approach for modelling of plasticity and failure behaviour Tinevez JY, White DJ, Hartenstein V, Eliceiri K, Toman-
of advanced high-strength steel sheets. Int J Damage Mech cak P, Cardona A (2012) Fiji: an open-source platform
22(2):188–218 for biological-image analysis. Nature methods 9(7):676–
Lian J, Yang H, Vajragupta N, Münstermann S, Bleck W (2014) 682, https://doi.org/10.1038/nmeth.2019, journal Article
A method to quantitatively upscale the damage initiation of Research Support, N.I.H., Extramural Research Support,
dual-phasesteels under various stress states from microscale Non-U.S. Gov’t, arXiv:2274.3772
to macroscale. Comput Mater Sci 94:245–257 Schowtjak A, Wang S, Hering O, Clausmeyer T, Lohmar J,
Liu W, Lian J, Münstermann S, Zeng C, Fang X (2020) Prediction Schulte R, Ostwald R, Hirt G, Tekkaya A (2019) Prediction
of crack formation in the progressive folding of square tubes and analysis of damage evolution during caliber rolling and
during dynamic axial crushing. Int J Mech Sci 176:105534. subsequent cold forward extrusion. Prod Eng. https://doi.
https://doi.org/10.1016/j.ijmecsci.2020.105534 org/10.1007/s11740-019-00935-x
Maire E, Bouaziz O, Di Michiel M, Verdu C (2008) Initiation and Shen F, Münstermann S, Lian J (2020) Investigation on the
growth of damage in a dual-phase steel observed by X-ray ductile fracture of high-strength pipeline steels using a
microtomography. Acta Mater 56:4954–4964 partial anisotropic damage mechanics model. Eng Frac
Marteau J, Haddadi H, Bouvier S (2013) Investigation of Mech 227:106900. https://doi.org/10.1016/j.engfracmech.
strain heterogeneities between grains in ferritic and ferritic- 2020.106900
martensitic steels. Exp Mech 53(3):427–439. https://doi. Shen HP, Lei TC, Liu JZ (1986) Microscopic deformation
org/10.1007/s11340-012-9657-6 behaviour of martensitic-ferritic dual-phase steels. Mater
Meya R, Kusche C, Löbbe C, Al-Samman T, Korte-Kerzel S, Sci Technol 2(1):28–33. https://doi.org/10.1179/mst.1986.
Tekkaya A (2019) Global and high-resolution damage quan- 2.1.28
tification in dual-phase steel bending samples with varying Tasan CC, Hoefnagels J, Geers M (2010) Microstructural band-
stress states. Metals. https://doi.org/10.3390/met9030319 ing effects clarified through micrographic digital image cor-
Mu L, Jia Z, Ma Z, Shen F, Sun Y, Yang Y (2020) A theoret- relation. Scripta Mater 62(11):835–838. https://doi.org/10.
ical prediction framework for the construction of a frac- 1016/j.scriptamat.2010.02.014
ture forming limit curve accounting for fracture pattern Tasan CC, Diehl M, Yan D, Zambaldi C, Shanthraj P, Roters F,
transition. Int J Plast. https://doi.org/10.1016/j.ijplas.2020. Raabe D (2014a) Integrated experimental-simulation anal-
102706 ysis of stress and strain partitioning in multiphase alloys.
Münstermann S, Lian J, Pütz F, Könemann M, Brinnel V (2017) Acta Mater 81:386–400. https://doi.org/10.1016/j.actamat.
Comparative study on damage evolution during sheet metal 2014.07.071
forming of steels dp600 and dp1000. J Phys: Conference Tasan CC, Hoefnagels J, Diehl M, Yan D, Roters F, Raabe D
Series 896 (2014b) Strain localization and damage in dual phase steels
Novokshanov D, Döbereiner B, Sharaf M, Münstermann S investigated by coupled in-situ deformation experiments
(2015) A new model for upper shelf impact toughness and crystal plasticity simulations. Int J Plast 63:198–210.
assessment with a computationally efficient parameter iden- https://doi.org/10.1016/j.ijplas.2014.06.004
tification algorithm. Eng Fract Mech 148:281–303 Tasan CC, Diehl M, Yan D, Bechtold M, Roters F, Schemmann
Papasidero J, Doquet V, Mohr D (2015) Ductile fracture of alu- L, Zheng C, Peranio N, Ponge D, Koyama M, Tsuzaki K,
minum 2024–t351 under proportional and non-proportional Raabe D (2015) An overview of dual-phase steels: advances
multi-axial loading: Bao-wierzbicki results revisited. Int in microstructure-oriented processing and micromechani-
J Solids Struct 69–70:459–474. https://doi.org/10.1016/j. cally guided design. Ann Rev Mater Res 45(1):391–431.
ijsolstr.2015.05.006 https://doi.org/10.1146/annurev-matsci-070214-021103

123
The differences of damage initiation and accumulation of DP steels...

Tekkaya AE, Ben Khalifa N, Hering O, Meya R, Myslicki S, Wu B, Li X, Di Y, Brinnel V, Lian J, Münstermann S (2017)
Walther F (2017) Forming-induced damage and its effects Extension of the modified bai-wierzbicki model for pre-
on product properties. CIRP Ann 66(1):281–284. https:// dicting ductile fracture under complex loading conditions.
doi.org/10.1016/j.cirp.2017.04.113 Fatigue Frac Eng Mater Struct 40(12):2152–2168
Tvergaard V (1981) Influence of voids on shear band instabili-
ties under plane strain conditions. Int J Frac 17(4):389–407.
https://doi.org/10.1007/BF00036191 Publisher’s Note Springer Nature remains neutral with regard
Tvergaard V, Needleman A (1984) Analysis of the cup-cone to jurisdictional claims in published maps and institutional affil-
fracture in a round tensile bar. Acta Metall 32(1):157–169. iations.
https://doi.org/10.1016/0001-6160(84)90213-X
West O, Lian J, Münstermann S, Bleck W (2012) Numerical
determination of the damage parameters of a dual-phase
sheet steel. ISIJ Int 52(4):743–752. https://doi.org/10.2355/
isijinternational.52.743

123