See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/316526885

Micromechanical modeling of dual phase steel using representative volume

elements based on actual microstructure

Conference Paper · February 2014

CITATIONS READS

0 233

4 authors:

Majid R. Ayatollahi Ali Cheloee Darabi

Iran University of Science and Technology Universität Stuttgart
379 PUBLICATIONS   6,764 CITATIONS    11 PUBLICATIONS   133 CITATIONS   

SEE PROFILE SEE PROFILE

Hamidreza Chamani Mohammad Javad Razavi

Iran University of Science and Technology Norwegian University of Science and Technology
26 PUBLICATIONS   136 CITATIONS    96 PUBLICATIONS   797 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

BOOK SERIES ON STRUCTURAL INTEGRITY (ISSN: 2522-560X) View project

fracture of polymeric nanocomposite View project

All content following this page was uploaded by Ali Cheloee Darabi on 27 April 2017.

The user has requested enhancement of the downloaded file.

 The Bi-Annual International Conference on Experimental
Solid Mechanics and Dynamics (X-Mech-2014)
Feb. 18-19, 2014, Tehran, Iran
Center of Excellence in Experimental Solid Mechanics and Dynamics,
School of Mechanical Engineering, Iran University of Science and Technology

Micromechanical modeling of dual phase steel using representative volume

elements based on actual microstructure

M. R. Ayatollahi1 *, A. Cheloee Darabi2, H.R. Chamani3, S.M.J. Razavi2

1
Professor, 2M.Sc. Student, 3Ph.D. Student

Fatigue and Fracture Lab., Center of Excellence in Experimental Solid Mechanic and Dynamics,
School of Mechanical Engineering, Iran University of Science and Technology, Narmak,
16846, Tehran, Iran.
*
(Tel: (98) 21 77240201, e-mail: m.ayat@iust.ac.ir)

Abstract
Dual phase (DP) steels are among the advanced high strength steel (AHSS) products recently developed
for the automotive industry. Due to their high strength and good formability, this group of steels is used
in the automotive industry for complex and high strength parts. The microstructure of DP steel basically
consists of a matrix of ferrite embedded by martensitic islands.
In the past, many methods have been used to analyze dual phase steels, but these theories are based on
the assumption of uniform deformation throughout the constituents, neglecting the local strain gradients.
This assumption contradicts the experimental observations and reduces the understanding of the
mechanics and mechanism of deformation of such materials.
In this paper, the C–Mn steel specimens with 0.2% carbon were put under intercritical annealing
treatment (ICT) at 765°C and quenched in water to obtain 25% martensite. After metallographic analysis
an actual microstructure of the dual phase steel is obtained by optical microscopy.
A 2D Representative Volume Element (RVE) was generated by FE code Ansys on the basis of an actual
microstructure which was obtained by image processing code in Matlab software. The individual single
phases flow curves was obtained based on dislocation theory and local chemical composition of
constituent. Both of the periodic and symmetric boundary conditions were imposed on RVE. The
micromechanical results were compared with the experimentally obtained data. It is shown that the
micromechanical modeling can predict both strength and ductility of the investigated dual phase steel.
The micromechanical modeling may then be used to portray the local stress and strain evolution of the
individual phases in the DP microstructures.

Keywords: Dual phase steel, microstructure, micromechanical modeling, mechanical behavior

1. Introduction

Modern high strength steels have been created for reduction of car body weight in automotive industry, improved
passive safety features and considerations of energy saving. These types of steels like DP and triple phase (TRIP)
include different phases in their microstructure [1].
Combination of high strength with good formability exhibits high energy absorption in the case of a crash for these
groups of steels. Their magnificent mechanical properties are because of the presence of different phases in the
microstructures with distinct differences in ductility and strength. Consequently, the favorable properties related
to each of the phases of multiphase steels can be combined into one material, consisting a purely ferrite matrix and
about 5–35% martensite scattered in attached as a second phase. Their behavior is like composite materials such
that the ferrite matrix certifies good cold formability, while the martensite particles are the strengthening elements
[2].
Studying about dual phase steels will help us to optimize the processing parameters to obtain the best mechanical
properties. Thus, to design an optimized the microstructure for the dual phase steel, comprehensive understanding
of the local deformation behavior is necessary. Shen et al [4] investigated the distribution of strain between ferrite
and martensite phases. Kang et al. [5] observed microscopic strain distribution in dual phase steels using coupled
SEM and Digital Image Correlation (DIC) techniques.
1
Sun et al [6] studied the failure modes and ductility of dual phase steel in a 2D RVE where the plastic work
hardening properties were determined by the in-situ synchrotron-based high-energy X-ray diffraction technique.
Ramazani et al [7] investigated the effect of microstructure banding on mechanical behavior of dual phase steels
and compared mechanical properties obtained from the equiaxial and banded microstructure by a 2D RVE.
Uthaisangsuk et al [3] have created real 3D RVEs slice wise from optical micrographs that attempts to describe the
influence of the multiphase microstructure on the failure mechanism by approaching the problem using
representative volume elements (RVE) within the framework of continuum damage mechanics.
In this work, DP 800 was produced by use of intercritical annealing process. Afterwards, the metallographic
investigations and tensile tests were carried out for characterizing the produced DP steel. To investigate the
influence of the multiphase microstructures on the deformation behavior, micrographs taken from optical
microscopy were applied for generating 2D Representative Volume Elements (RVEs) with symmetric and periodic
boundary conditions. The ferritic and martensitic phases in the RVE model were considered particularly, that a
flow curves based on dislocation theory and chemical compositions were given for individual phases. It is
necessary to investigate the effect of mesh size in order to the model validate.

2. Experiment

2.1 Materials

A low carbon steel sheet with 3.00 mm thickness was used for investigating DP800 steel. Table 1 lists the chemical
compositions of low carbon steel sheet used. The intercritical annealing process at 760 °c was accomplished to
obtain steel sample with DP microstructures containing a specific martensite phase fraction. This temperature lies
between its AC1 and AC2 temperatures. The holding time after reaching the intercritical temperature was 10min.
An overview of the accomplished intercritical annealing process is presented in Figure 1. The samples were finally
quenched quickly in water to room temperature, in which the phase transformation from austenite to martensite
occurred.

Table 1: Chemical compositions of low carbon steel heat-treated in this study

Element C Mn Si P S Cr Ni Al Mo V Cu Co
Mass 0.2 0.5 0.213 0.004 0.02 0.157 0.083 0.011 0.046 0.008 0.121 0.019
contents%.

Figure 1: The intercritical annealing process

The sheet specimen after heat treatment was prepared for the metallographic analysis. The sample was abraded
by silicon carbide papers with grit sizes of 100, 240, 400, 600, 800, 1000, 1500 and 2000 in sequence, and was
then polished by 0.5 m alumina solution. After polishing, the specimens were etched with 2% nital for 3–5 s.
The microstructure of DP800 was obtained by optical microscope as shown in Figure 2.

2
 Figure 2: Optical micrograph of dual phase steels, microstructure of DP800 Vm=25%

In this micrograph the light gray zones are ferrite phases and the dark regions are martensitic phases. The volume
fractions of martensite, Vm , was found to be about 25% by image analysis software.

2.2 Mechanical testing

After the metallographic analysis, tensile specimens from the heat-treated steel sheets were prepared for tensile
testing. Tensile sample of 140 mm gage length and width of 20 mm was used (see Figure 3). The specimens were
loaded under uniaxial condition in a universal testing machine equipped with an automatic controller using
displacement control mode at a strain rate of 1.65 × 10 . A test specimen inside the test machine subjected
to tensile loading is shown in Figure 4. During the tensile experiments, displacement was recorded using an
extensometer. Tensile tests were repeated 6 times for DP800 samples.

Figure 3: Tensile specimen prepared from dual phase steel

Figure 4: The tensile specimen subjected to tensile loading

3
2. Micromechanical Modeling

3.1 Representative volume element (RVE)

The DP material on the macro-scale is generally treated as a continuum, while on the microstructure size, the
strength and elongation of steel are clearly discontinuous due to its multiphase character. The assumptions of
homogeneity and continuity are valid on the macro-scale analysis. So a RVE is generated based on an actual
microstructure to describe the inhomogeneous structure of the DP steel. The RVE model could demonstrate both
phase combination and microstructural configuration. This method provides a realistic description of the material
deformation on the micro size giving insight into the stress and strain evolution and distribution in phases and the
interface of them. The RVE should be large enough to represent the most important microstructure traits of the
investigated DP steel. In a RVE, some parameters such as shape, morphology, size and randomness of constituent
phases can be considered to be representative of the microstructure under investigation. In this work, a 2D RVE
with both of symmetric and periodic boundary conditions was analyzed in which the volume fraction of the
martensite phase (Vm) is about 25%.

3.2 2D RVE based on real microstructure

2D RVE models were generated by FE code Ansys on the basis of actual microstructures which were obtained
by image processing code. According to this method, 2D RVEs from the actual microstructure of DP800 steel
with 25% martensite phase is created. The 2D RVE generation is visualized in Figure 5, where dark elements
refer to martensite and white regions refer to ferrite. The size for this 2-D RVE consisting of 90,000 elements is
100×100 µm and 300×300 pixels. The dimensions of RVE are small relative to the thickness of test specimen,
thus plane strain elements were used in this study (PLANE182).

(a) (b)
Figure 5: 2D RVE model studied (a) without mesh, (b) with mesh

3.3 Boundary condition of unit cell

Two types of boundary conditions were applied in the micromechanical model in this study: symmetric and
periodic boundary conditions as shown in Figure 6. For periodic boundary condition, the multipoint constraints
need to implement Eq. (1), especially in three-dimensional problems. Eq. (1) represents periodicity boundary
conditions for the mechanical material in the small strain range as [11]:

u DC ( X ) u AB ( X ) uD u B C (Y ) u A D (Y ) uB (1)

where the point A is assumed to be fixed and (X, Y) are local coordinates that are used to denote homologous
points on pairs of face. Eq. (2) directly implies that:
u u u (2)
C D B

The formulation for the symmetric boundary condition can be expressed as (Eq. 3),
u B C (Y ) uB u C D (X ) uD u AD (Y ) 0 u AB (X ) 0 (3)

4
 (a) (b)
Figure 6: Deformation of (a) periodic, (b) symmetric boundary conditions under uniaxial loading

3.4 Material behavior

Many works have been done about accurate measurement of mechanical properties in each phase of different
materials [7, 8]. Rodriguez and Gutierrez [8] studied the accuracy of predictions obtained from an empirical
formulation with nano-hardness.
In this paper, a dislocation theory based on strain hardening method is used to predict the flow curve of each
phase which is computed empirically according to the local chemical composition in DP800. This method is for
estimating the stress-strain relation of the single phases. Thus, the Peierl’s stress is calculated empirically
according to this method [8–10]. The approach can be expressed as [8]:

1 exp( M.k. )
0 .M . . b. ( (4)
k.L
° is the Peierl’s stress, which depends on the chemical constituents of each phase in the material, and is presented
for DP800 as Eq. (5).
0 77 80 (%Mn) 750 (%p) 80 (%Cu)60 (%Si)
( (5)
45 (%Ni) 11 (%Mo) 60 (%Cr) 5000 (Nss )

In Eq. (6) estimates material strengthening by carbon content. For ferrite it is calculated by:
(MPa) 5000 (%Cssf ) ( (6)
and for martensite it is estimated as:
(MPa) 3065 (%Cmss ) 161 ( (7)

where %Cssf and %Cmss denote the wt % carbon content in ferrite and martensite, respectively. The third term
in Eq. (4) describes the effects of cumulated dislocation strengthening in addition to work softening due to
recovery [10], where is a constant, M is a Taylor factor, is the shear modulus, b is the Burger’s vector, L is
the dislocation mean free path, k is the recovery rate. For ferrite and martensite, the last two parameters and
are different in each phases of steel [8-10]. Table 2 show all constants used for individual phases.

Table 2: Summary of all model constants for flow curves prediction of individual phases.

Fraction C (%) [10] M [10] b [10] µ [10] L [10]

Steel Phase (%)
Ferrite 75 0.0469 254.6 234.5 0.33 3 2.5×10-10 80,000 15×10-6
DP800
Martens
ite 25 0.659 254.6 1858.8 0.33 3 2.5×10-10 80,000 3.8×10-6

5
 4. Result and discussion
In this paper, the mechanical behavior, such as stress - strain curve and micromechanical modeling of complex
deformation in 2D RVE investigated for dual phase steels ferritic– martensitic. Two boundary conditions of
symmetric and periodic have been used for micromechanical modelling. In Figure 7, the behavior of 2D
micromechanical models under tension in direction of Y and under periodic and symmetric boundary conditions
are compared with the experiment data.

Figure 7: Comparison between experimental data and 2D micromechanical modeling under

periodic and symmetric boundary conditions.

In Figure 8, the equivalent plastic strain within 2D and 3D models under both boundary conditions for DP9800
steel are shown and shear failure bands being displayed in all of models. In both 2D models, the propagation
directions of shear failure bands are similar. In out of plane direction (Z direction) they are again similar leading
to similar flow curves. Failure bands occur in the direction of 45° with respect to the direction of tensile loading.

(a) (b)
Figure 8: Local equivalent plastic strain distribution in the 2D micromechanical models of DP800, under (a)
symmetric and (b) periodic boundary conditions.

No prescribed damage law/failure criteria are used for various constituent phases in 2D, and the initial
microstructure inhomogeneity among the phases serves as the imperfection condition triggering the plastic
instability. The ultimate ductile failure of DP steel is predicted in the form of plastic strain localization resulting
from the incompatible deformation between the harder martensite and softer ferrite phases.

6
 5.Conclusions
The predicted stress–strain curves under periodic and symmetric boundary conditions were fairly agreed with the
experimentally determined curves, especially the yield stresses. The local stress and strain distributions in the DP
structure were characterized. Under tensile deformation several short interrupted shear bands were found in the
DP structure with small martensite contents, but long continuously pronounced localizing bands appeared in the
DP structure with high martensite fractions. The mechanical properties of DP steel were strongly influenced by
morphology of the dispersed martensitic phase. The modeling approach can be further developed and used for
describing deformation and failure behavior of the multiphase high strength steels.
In 2D models, the analysis was based on plane strain assumption. In this case, martensite phases in out of plane
(Z direction) have not got scattered partitioning, but are uniform similar to the surface. In other words, phase
arrangement in out of plane (Z direction) is similar to the surface, so it leads to less ultimate strength with respect
to the experimental result.

References
[1] Bleck W, Papaefthymiou S, Frehn A. (2004), Steel Res Int 75, 704–10.
[2] T. Furukawa, M. Tanino, H. Morikawa, M. Endo. (1984), Trans ISIJ 24 113–121.
[3] Uthaisangsuk V, Prahl U, Bleck W. (2008), Computational Materials Science 4, 27–35.
[4] Shen Hp, Lei Tc, Liu Jz. (1986). Mater Sci Technol 2, 28–33.
[5] Kang Jd, Ososkov Y, Embury Jd, Wilkinson Ds. (2007), Scr Mater 56, 999–1002.
[6] Sun X, Choi K.S, Liu W.N, Khaleel M.A. (2009), International Journal Of Plasticity 25, 1888–1909.
[7] Ramazani A, Mukherjee K, Prahl U, Bleck W. (2012), Computational Materials Science 52, 46–54.
[7] Lawson R.D, Matlock D.K, Krauss G. (1981) Fundamentals Of Dual-Phase Steels, The Metallurgical Society
Of Amie.
[8] Rodriguez R.M, Gutierrez I. (2003), Mater Sci Forum 2003;426:4525– 4530.
[9] Rodriguez R.M, Gutierrez I. (2004), Proceedings of the 2nd International Conference On Thermomechanical
Processing Of Steels, Liege.
[10]Reisner G, Werner E.A, Fischer F.D. (1988), Int. J. Solids Struct;1998;35:2457– 2473.
[11] B¨Ohm H J. A Short Introduction To Basic Aspects Of Continuum Micromechanics. Institute Of Lightweight
Design And Structural Biomechanics (Ilsb) Vienna University Of Technology;1998.

View publication stats