Development of Mn-Cr-(C-N) Corrosion Resistant Twinning

Induced Plasticity Steels: Thermodynamic and Diffusion

Calculations, Production, and Characterization
LAÍS MÚJICA RONCERY, SEBASTIAN WEBER, and WERNER THEISEN

In this work, the development of corrosion-resistant twinning induced plasticity steels is pre-
sented, supported by thermodynamic and diffusion calculations within the (Fe-Mn-Cr)-(C-N)
alloy system. For the calculations, ambient pressure and primary austenitic solidification were
considered as necessary to avoid nitrogen degassing in all processing steps. Manganese is used as
an austenite stabilizer, chromium is used to increase nitrogen solubility and provide corrosion
resistance, and carbon and nitrogen are used as interstitial elements to provide mechanical
strength. Isopleths of the different elements vs temperature as well as isothermal sections were
calculated to determine the proper amount of Mn, Cr, total interstitial content, and the C/N
ratio. Scheil and diffusion calculations were used to predict the extent of microsegregations and
additionally to evaluate the effect of diffusion annealing treatments. The materials were pro-
duced in laboratory scale, being followed by thermomechanical processing and the character-
ization of the microstructure. Tensile tests were performed with three different alloys, exhibiting
yield strengths of 460 Mpa to 480 MPa and elongations to fracture between 85 pct and 100 pct.

DOI: 10.1007/s11661-010-0334-z
 The Minerals, Metals & Materials Society and ASM International 2010

I. INTRODUCTION This work deals with the development of Fe-Mn-Cr-

C-N TWIP steels, starting with thermodynamic equilib-
IN general, transformation induced plasticity (TRIP) rium and diffusion calculations to obtain appropriate
and twinning induced plasticity (TWIP) steels are chemical compositions and parameters for the produc-
materials characterized by an exceptional combination tion and thermal treatment. The calculations comprise
of strength, toughness, and ductility. Nevertheless, the phase diagrams, isothermal sections, stacking fault
corrosion resistance is poor because of the chemical energy, solid fraction, and manganese content during
constitution of such alloys, for instance, Fe-22Mn-0.6C solidification, as well as element distribution during
and Fe-15Mn-3Al-3Si.[1,2] diffusion annealing. After casting of three proposed
In an attempt to produce a TWIP steel not only with compositions, mechanical, and thermal treatments, the
excellent mechanical properties but also with enhanced work continues with the characterization of the pro-
resistance to wet corrosion, a new series of steels has duced steels, covering the composition, microstructure,
been developed. Manganese is incorporated to provide strength, and corrosion.
the plasticity effect. Carbon and nitrogen are used as Within this context, several key issues had to be taken
interstitial elements to obtain a stable austenitic struc- into account, namely induced plasticity effects, stacking
ture. Besides, nitrogen improves the resistance to pitting fault energy and the concept of C+N alloying.
corrosion. Additionally, the incorporation of chromium,
on one hand, promotes the formation of a passivation
layer and improves the corrosion resistance, and, on the
other hand, would increase the solubility of nitrogen in A. TWIP and TRIP Steels
the melt during casting. The induced plasticity effects occurring in TWIP and
TRIP steels during deformation are responsible for the
mechanical response of those materials. Two kinds of
LAÍS MÚJICA RONCERY, Scientific Assistant, is with effects are known so far, TRIP and TWIP. TRIP steels
Ruhr-Universität Bochum, Insitut für Werkstoffe, Lehrstuhl consist of a ferritic-bainitic matrix with retained aus-
Werkstofftechnix, D-44780, Bochum, Germany, and Ph.D. Student tenite and show strain induced c fi e transformation.[3]
with the Max Planck Institute für Eisenforschung, D-40237, This transformation also can be found in high manga-
Düsseldorf, Germany. Contact e-mail: [email protected]
SEBASTIAN WEBER, Group Leader, is with Ruhr-Universität
nese steels with relatively low carbon contents
Bochum, Insitut für Werkstoffe, Lehrstuhl Werkstofftechnix, and is (Fe-17Mn-0.2C), which have austenitic-martensitic
with the Joint Research Group, Hemholtz-Zentrum Berlin für structure.[1]
Materialien und Energie GmbH, D-14109 Berlin, Germany. The second kind of steel consists of a fully austenitic
WERNER THIESEN, Head, is with Ruhr-Universität Bochum, matrix, showing intensive mechanical twinning under
Insitut für Werkstoffe, Lehrstuhl Werkstofftechnix.
Manuscript submitted December 4, 2009. strain deformation. Well-known TWIP steels result
Article published online June 23, 2010 from the alloy system Fe-Mn-Si-Al, in which the

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, OCTOBER 2010—2471

amount of manganese leads to a fully austenitic struc- In case of the temperature, it has been shown for
ture, and the elements Al and Si are used to adjust the Cr-Ni[10] and Fe-Mn-C[11] austenitic steels that both
stacking fault energy (SFE) and reduce the specific DGc!e and SFE increase with temperature. Conse-
weight. These materials can reach elongations up to 0.6 quently, at lower temperatures martensitic transforma-
true strain and tensile strengths up to 1200 MPa true tions will be favored, whereas at higher temperatures,
stress.[2] Another system is Fe-22Mn-0.6C, which has a twinning and dislocation glide would be the main
higher strength than the alloys based on Fe-Mn-Si-Al deformation mechanisms.
because of the effect of carbon as interstitial.[4] It is also The influence of the amount of alloying elements is
less expensive and easier to manufacture. complex and should be analyzed separately. For Fe-Mn
The corrosion behavior of such steels has not been as a binary system, a parabolic behavior with a
studied widely. Works on different systems Fe-Mn-Al minimum at 14 at. pct manganese was found.[14] In
and Fe-Mn-Al-Si in different media[5–7] show that, in Fe-Mn-C systems, Allain et al.[11] have predicted that
general, this kind of steel has a poor corrosion the SFE increases with the contents of manganese and
resistant in acidic and saline aqueous environments carbon. However, measurements performed on the
in comparison with Cr/Ni stainless and interstitial free system Fe-22Mn-C[13,15] show that a local minimum at
steels. molar carbon content xC = 0.6 at. pct exists.
The effect of nitrogen differs significantly between
various austenitic steels grades. For example, in the case
of the system Fe-Mn-N, Kibey et al.[16] have shown,
B. SFE
that the SFE is increasing with the content of nitrogen.
It is known that the SFE is one of the properties Petrov[15] found a parabolic behavior for the steel
related to the predominant deformation mechanisms. Fe-15Cr-17Mn-N with a minimum at nitrogen molar
The martensitic transformation takes place with SFE content xN = 2.0 at. pct. In the case of austenitic steels
values < 20 mJ/m2, twinning at SFE > 20 mJ/m2.[8,9] containing Ni, the effect of nitrogen is parabolic as well
The relationship of SFE and the Gibbs free energy but has a maximum, for example, in the system
DGc!e is well known. Olson and Cohen[10] proposed a Fe-18Cr-16Ni-10Mn-N, at xN = 1.5 at. pct.
model in which the SFE is dependent on the Gibbs free
energy of the phase change, the molar density of the
face-centered cubic (fcc) {111} close-packed planes, and C. Carbon-Nitrogen Steels
the surface energy of the interface c ! e. Austenitic steels with carbon and nitrogen combined
as interstitial elements have been a matter of study by
C ¼ gqA ðDGc!e þ Estr Þ þ 2r ½1 Gavriljuk and Berns[17,18] on alloy systems such as Fe-
18Mn-18Cr-C-N, introducing the C + N content and
C ¼ 2qA ðDGc!e Þ þ 2r ½2 the C/N ratio as variables that determine the austenite
stability.
In Eqs. [1] and [2], C is the SFE, g is the number of fault The combination of C and N affects the density of free
planes, qA is the density of atoms in a close-packed electrons, atomic interactions, short-range atomic order-
plane, Estr is the strain energy, and r(g) is the surface ing, stability of the solid solution, and of course,
energy per unit area. The dependence of the SFE on the mechanical properties of the steel.[19–21]
Gibbs free energy implies a strong dependence of these C and N as interstitial elements in an austenitic matrix
two variables on the temperature and the composition. affect the atomic interactions by modifying the concen-
The second expression is the simplification of the tration of electrons in the conduction band, thereby
previous one,[11] neglecting Estr. The surface energy changing the character of the interatomic bonds. For
of the interface c/e of austenitic steels is assumed at nitrogen, a higher content produces an increase of the
10 mJ/m2.[10] concentration of free electrons, whereas carbon en-
Another thermodynamic approach for the calculation hances the covalent character of the bonds. The com-
and correlation of SFE and free Gibbs energy was bination of C + N has a synergetic effect, shifting the
considered by Ishida and Yakubtsov,[12,13] in which maximum of free electron concentration to higher
the SFE has a bulk contribution and a segregation interstitial contents and therefore enhancing the metallic
contribution as follows: character of the bonds.
In C + N alloyed austenitic steels, short-range atomic
C ¼ Cbulk þ Cseg ½3
ordering is found because of repulsive forces of C and N
atoms as nearest neighbors. This agrees with the fact
DGc!e that the concentration of free electrons favors short-
Cbulk ¼ ½4 range ordering also being associated with a greater
2V2=3 N0
stability of the solid solution toward phase transforma-
According to this approach, the segregation contri- tions. One of these transformations is c ! e. Another
bution depends on the elastic energy, surface energy, transformation is the precipitation of carbides promoted
and chemical-free energy caused by segregations in by the clustering of Cr and C atoms, an effect that is
the stacking fault being relatively small and usually reduced by the incorporation of N, extending the area of
neglected. the homogeneous austenitic solid solution.

2472—VOLUME 41A, OCTOBER 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

 Concerning its corrosion resistance, it is shown that Table I. Composition of the Reference TWIP and TRIP
Fe-18Mn-18Cr-C-N steels are superior compared Steels; Values in Mass Pct with Iron Being the Dependent
with high manganese Hadfield steel in saline aqueous Substitutional Element
solutions.[22,23]
Steel Mn Al Si C
TWIP-ref 23.9 0.003 0.280 0.506
II. EXPERIMENTAL TRIP-ref 1.67 0.042 1.734 0.207
A. Thermodynamic and Diffusion Calculations
The thermodynamic calculations including the
formed during mechanical preparation, carried out with
isopleths and Scheil-Gulliver solidification paths were
Struers A2 solution (perchloric acid/ ethanol/ 2-Butoxy-
performed by Thermo-Calc software version R.
ethanol/ water).
Considering iron as the main constituent of the alloy,
manganese, chromium, carbon and nitrogen as the main
alloying elements and taking into account that the gas E. XRD
phase is required for the analysis, the database TCFE5
was chosen for all the calculations. The diffusion XRD measurements were carried out on the deformed
calculations were performed using the Dictra 2.4 soft- tensile specimens to check for the formation of a or e
ware with the TCFE5 thermodynamic database and the martensite caused by the deformation. A Co-Ka labo-
MOB2 mobility database. ratory source was used with a beam diameter of 1.5 mm.

B. Materials F. Corrosion
Based on the results of the thermodynamic and The corrosion behavior of the alloys was analyzed by
diffusion calculations, melting, thermal, and mechanical means of electrochemical measurements. Linear polar-
processing of the alloys was performed. Electrolytic ization curves were obtained using a PGP201 galvano-
manganese, chromium nitride, graphite, and ARMCO stat/potentiostat, saturated calo-mel as reference
iron were used as raw materials for the melt. Ingots of electrode and platinum as counter electrode in a 0.5 M
about 4 kg and 54 9 30 mm cross section were pro- H2SO4 electrolyte. For comparison, two reference TWIP
duced at a nitrogen partial pressure of 0.9 bar. The and TRIP steels were analyzed with the same conditions.
ingots were diffusion annealed in argon atmosphere for The compositions of the reference materials are given in
16 hours, quenched, hot rolled to 16 9 40 mm, solution Table I.
annealed at 1473.15 K (1200 C) in argon atmosphere
for 45 minutes, and finally quenched in water. The bulk
composition of the materials was measured by optical
emission spectrometry. III. RESULTS AND DISCUSSION
A. Thermodynamic Calculations
C. Electron Microscopy TWIP steels are characterized by a fully austenitic
The microstructure of the ingots in the as-cast state structure in which the SFE controls if twinning is the
was characterized using a scanning electron microscope predominant deformation mechanism. As mentioned,
(SEM) equipped with a tungsten filament. Samples were the composition is the main factor that influences the
prepared by grinding and mechanical polishing to free Gibbs energy associated with the austenite-mar-
submicron size. Electron backscattered contrast as well tensite transformation and therefore the SFE.
as secondary electron imaging was used to reveal the Consequently, with the intention to obtain a fully
microstructure. Energy dispersive X-ray spectrometry stable austenitic structure that could be corrosion
(EDX) line scans were taken with an acceleration resistant, the combination Fe-Mn-Cr-C-N is a potential
voltage of 25 kV and 3.5 lm step size to a total length candidate to fulfill these two main requirements.
of 3 mm. Manganese is an austenite former and promotes plas-
ticity induced effects such as twinning. Nevertheless, it
leads to poor corrosion resistance. Carbon acts as an
D. Tensile Tests interstitial and stabilizes the austenitic structure, being
Tensile tests were performed according to the stan- an even more effective austenite former in combination
dard EN 10002 with a constant elongation rate of with nitrogen. Nitrogen not only stabilizes the austenite
1 9 103 at room temperature. Additional specimens like carbon but also provides corrosion resistance.
were deformed up to 10 pct and 20 pct engineering Chromium enhances the solubility for nitrogen in the
strain for further X-ray diffraction (XRD) analyses. For liquid and provides corrosion resistance to the alloy but
these analyses, samples in the direction of the load and assists the sensitization at high carbon contents and/or
in transverse direction were taken with electro discharge improper heat treatment. As chromium is stabilizing
machining. Afterwards, the samples were ground and the ferrite, its amount has to be controlled care-
mechanically polished to submicron size. Electropolish- fully to prevent the formation of delta ferrite during
ing was used to remove any possible surface distortion solidification.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, OCTOBER 2010—2473

 Calculations of isopleths support the estimation of a to a C/N ratio of 1.0 are illustrated in Figure 2(a),
suitable amount of each element. As a parameter, the whereas the ones for 0.75 and 0.5 are depicted in
pressure is fixed to 1 bar. The chromium content is set to Figures 2(b) and (c), respectively. In the first case
12 mass pct, being the minimum value that could result (C/N = 1.0) the carbides M7C3 and M23C6 are
in a steel showing passivity. A content of manganese present for (C + N) exceeding 1.2 mass pct, whereas a
between 20 pct and 30 mass pct is required to stabilize
the austenite and provide twinning as a predominant
deformation mechanism.
The appropriate carbon and nitrogen contents are
more difficult to obtain. Their effect can not be treated
separately, because both occupy the interstitial sites, and
the austenite stability is not linearly dependent on them
as single elements. Thus, two variables are considered—
the total interstitial content (C + N) and the C/N ratio.

1. Temperature vs Total Interstitial Content

For the system Fe-20Mn-12Cr-C-N, Figure 1 shows a
(C + N) isopleth at C/N equal to 0.75 mass/mass. One
of the main features of the system is the presence of the
phase field liquid and austenite (L + A) in the range of
(C + N) between 0.5 mass pct and 0.75 mass pct.
According to the equilibrium calculation, solidification
should take place primarily austenitic without the
formation of a bcc phase (delta ferrite). Because of the
higher solubility of nitrogen in austenite compared with
ferrite, this will ensure that all nitrogen dissolved in the
liquid phase is incorporated into the solid state. The
coexistence of liquid and gas in equilibrium must be
avoided because of defects such as pores that may occur
during casting. A partly ferritic solidification likewise
could promote defects because of the release of nitrogen
from the freshly formed crystalline bcc structure. This
effect is related to the low nitrogen solubility in ferrite.
In this specific case, a value of (C + N) of 0.55 mass pct
to 0.6 mass pct avoids the inconveniences mentioned
previously.

2. Isothermal Sections
The determination of an appropriate C/N ratio was
performed by analyzing isothermal sections at solution
annealing temperature. In Figure 2 the isothermal
section of the system Fe-Mn-12Cr-C-N at 1273.15 K
(1000 C) is depicted. The phase equilibria corresponding

Fig. 2—Isothermal section of the system Fe-Mn-12Cr-C-N at

Fig. 1—Phase diagram of the system Fe-20Mn-12C-0.75(C/N) at 1273.15 K (1000 C), with C/N ratio of 1.0 (a), 0.75 (b), and 0.5 (c)
pressure of 1 bar. mass/mass.

2474—VOLUME 41A, OCTOBER 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

cubic carbo-nitride of M(C,N) type is found for less the model described in Eqs. [3] and [4] are given. It is
than 20 mass pct manganese. In the second case shown that the alloy exhibiting a higher SFE is
(C/N = 0.75), the carbide transition is shifted signifi- Fe-30Mn-12Cr-0.3C-0.4N, the one with a higher man-
cantly toward a higher total interstitial content of ganese content. The alloy Fe-20Mn-12Cr-0.24C-0.32N
2.5 mass pct, widening the austenitic phase field. At exhibits a because of its lower manganese and total
last, a value of C/N equal to 0.5 promotes the existence interstitial content. The results obtained by the second
of a gas + austenite phase field, considered negative for model[12] are in the range of TWIP as a predominant
casting purposes. It is shown, that a C/N ratio of 0.75 deformation mechanism. However, using the first
would be more convenient in terms of avoiding carbide model,[10,11] the range of SFE is in the order of
precipitation and/or degassing. dislocation glide as a predominant deformation mech-
Based on these considerations, three different compo- anism.
sitions were proposed, shown in Tables II and III. All
alloys have a C/N ratio of 0.75 with a total interstitial
B. Solidification and Diffusion Calculations
content adjusted to obtain a fully austenitic solidifica-
tion and avoid gas formation at a temperature of 1. Solidification
1873.15 K (1600 C), the upper limit of the melt The solidification of the material was evaluated based
temperature before casting. on the Scheil–Gulliver assumptions, implying that the
substitutional components have infinite diffusivity in
3. Stacking Fault Energy the liquid and zero diffusivity in the solid phase. The
Based on models described earlier, the SFE was interstitial elements carbon and nitrogen were consid-
calculated from thermodynamic data. The DGc!e was ered as fast diffusing species, meaning that their con-
calculated by computational thermodynamics at centration in the one-dimensional cell always is
298.15 K and 0.1 MPa for each phase separately, setting according to the local equilibrium during the Scheil
the following as references states: graphite for carbon, calculation. Taking that into account, the degree of
bcc for iron and chromium, gas for nitrogen, and fcc for segregation for the different elements was calculated.
manganese. In Table IV, the results of the calculations In Figure 3, the solid fraction vs the temperature of
are given. In the first column, the estimation according the system Fe-20Mn-12Cr-0.24C-0.32N at 0.1 MPa is
to Eqs. [1] and [2] is shown, assuming r is equal to depicted. The dashed line corresponds to the equilib-
10 mJ/m2.[10] In the second column, the results applying rium calculation, whereas the continuous line is the
result of the Scheil–Gulliver calculation. The solidifica-
tion proceeds totally in the field of liquid + austenite,
but the solidus temperature significantly decreases from
Table II. Intended Content of the Main Alloying Elements 1598.15 K (1325 C) to 1503.15 K (1230 C).
of the Investigated TWIP C + N Steels; Values in Mass Pct
with Iron Being the Dependent Substitutional Element 2. Diffusion Annealing
With the intention of analyzing the segregation
C N Mn Cr C+N C/N phenomena in these steels, the manganese content in
A 0.240 0.320 20.0 12.0 0.560 0.75 the austenite and in the liquid phase was obtained from
B 0.300 0.400 25.0 12.0 0.700 0.75 the calculations based on the Scheil–Gulliver model. For
C 0.300 0.400 30.0 12.0 0.700 0.75 alloy A (Fe-20Mn-12Cr-0.24C-0.32N), the austenite
starts to solidify with a manganese content of 17
mass pct, whereas at the end of solidification, it reaches

Table III. Measured Composition of the Main Alloying

Elements of the Investigated TWIP C + N Steels; Values
in Mass Pct with Iron Being the Dependent Substitutional
Element

C N Mn Cr C+N C/N
A 0.240 0.330 21.085 11.89 0.570 0.728
B 0.319 0.447 25.367 11.99 0.766 0.713
C 0.332 0.440 29.941 11.77 0.772 0.754

Table IV. SFE Estimation Based on the Approaches

of Allain et al.[11] and Ishida[12]

SFE1,2 [mJ/m2] SFE3,4 [mJ/m2] DGc!e [J/mol]

A 53.7 33.8 744.0
B 72.4 44.0 1064.52
C 78.0 47.1 1165.54 Fig. 3—Solid fraction during solidification calculated with the
Scheil–Gulliver model.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, OCTOBER 2010—2475

Fig. 4—Manganese weight fraction in the solid and in the liquid
phase during solidification, calculated with the Scheil–Gulliver
model.

33 mass pct, having a degree of segregation (S = CmaxMn /

Cmin
Mn ) of approximately 2 (Figure 4).
For the estimation of proper diffusion annealing
parameters, thermodynamic and diffusion processes
should be considered. Concerning the temperature, at
least 30 K (30 C) beneath the solidus temperature
should be considered for diffusion annealing. In agree-
ment with the Scheil–Gulliver solidification, the appro-
priate temperature to carry out the heat treatment of
the system Fe-20Mn-12Cr-0.24C-0.32N should be
1473.15 K (1200 C).
Concerning homogenization, the evolution of the
element distribution during diffusion annealing at
1473.15 K (1200 C) was calculated. An EDX compo-
sition profile of a sample of Fe-20Mn-12Cr-0.24C-0.32N
was measured from the center to the outer part and used
as an input for the diffusion calculation. Only the
chromium and manganese levels were measured, as
carbon and nitrogen cannot be quantified by EDX. At
time t = 0, a minimum weight percentage of Mn = 19
mass pct and a maximum of Mn = 25 mass pct is
encountered, corresponding to a degree of segregation
of 1.3 besides the large scattering of the measured
values. The maximum and minimum values for chro-
mium are Cr = 12.1 mass pct and Cr = 13.9 mass pct,
respectively. Carbon and nitrogen are fast diffusing
species at this temperature; they were assumed to have a
linear distribution at time t = 0 with a value equal to
that of the bulk materials, meaning C = 0.24 mass pct
and N = 0.32 mass pct. The composition profiles for
each element are shown in Figure 5, which the values
are presented in mol fraction.
One interesting result form the diffusion calculations
is the tendency to the form substitutional interstitial
decomposition. It is known that carbon and nitrogen
have more affinity to the elements chromium and
manganese rather than iron, because the chemical Fig. 5—Calculated distribution of nitrogen, carbon, chromium and
potential of nitrogen in iron is decreased.[18] Even manganese for time t = 0 and t = 30 min.
though at time t = 0 the fast diffusing components
(carbon and nitrogen) were considered with a linear the element profiles at time t = 30 minutes, where the
distribution along the sample, they rapidly migrate after carbon and nitrogen profiles match those of chromium
the firsts seconds toward points of high manganese and and manganese (Figure 5). That reflects a strong coseg-
chromium content. This trend is shown by comparing regation of the carbon and nitrogen with chromium.

2476—VOLUME 41A, OCTOBER 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

 Table V. Yield Strength, Ultimate Strength, True Strength,
and Elongation to Fracture of the Developed Alloys

Rp0.2 [MPa] R [MPa] Rm [MPa] A [%]

A 455.4 ± 9.2 864.6 ± 2.4 1571.2 ± 58.0 89.2 ± 6.6
B 484.8 ± 12.0 877.5 ± 3.5 1658.7 ± 23.5 98.0 ± 1.6
C 488.3 ± 10.0 858.3 ± 9.4 1520.9 ± 21.8 88.2 ± 0.1

Fig. 6—Manganese segregation simulation for a diffusion annealing

treatment at 1473.15 K (1200 C).

Fig. 8—Engineering stress–strain curve of the developed alloys.

D. Tensile tests
The engineering stress–strain curves of the materials
indicate that the three developed alloys exhibit an
exceptional plasticity combined with a comparatively
high yield strength. Some interesting points of the results
are worth being mentioned. The first one is that the alloy
Fig. 7—SEM micrograph of the austenitic structure in the nonde- A, having 0.56 mass pct C + N (Fe-20Mn-12Cr-0.24C-
formed state. 0.32N)— the material with the lowest interstitial content
of all—is the one with the lowest yield strength. The
alloys B and C (Fe-25Mn-12Cr-0.3C-0.4N and Fe-
The scatter of the Cr and Mn signals of the originally 30Mn-12Cr-0.3C-0.4N), with a C + N content of 0.7
measured sample is diminished; however, 30 days of mass pct, exhibit higher yield strengths, as presented in
exposure of such a material would be necessary to Table V and in Figure 8. The increment of yield
obtain an even composition with a degree of segregation strength with total content of carbon and nitrogen can
of 1.06, a condition that cannot be implemented in be explained by the classic theory of strengthening by
practice. A diffusion annealing of 16 hours at 1473.15 K interstitial elements in solid solution. Nevertheless, it is
(1200 C) was used in the real procedure instead difficult to determine whether it is the carbon or nitrogen
(Figure 6). that strongly influences this mechanism or whether it is
the contribution given by the interaction of these two
elements. Berns and Gavriljuk[18] mention that nitrogen
C. Microstructure
is 1.5 times more effective than carbon as a strengthen-
The microstructure of the material Fe-20Mn-12Cr- ing element in austenitic steels, associated with the
0.24C-0.32N obtained after diffusion annealing, hot interaction of dislocations and interstitials reflected in
rolling, solution annealing, and quenching in water is the behavior of the enthalpy of binding between
shown in Figure 7. A fully austenitic structure without nitrogen atoms and dislocations, which is higher than
d-ferrite or martensite was obtained. Because of the that of carbon atoms and dislocations.
diffusion annealing process, the grain size is relatively Concerning the plastic deformation, alloy C having a
coarse, approximately 80 lm. Recrystallization twins higher manganese and C + N content exhibits a lower
from the annealing processes also can be recognized elongation to fracture as well as a lower extent of cold
within the grains. No pores were formed, but manganese work hardening. In fact, this is correlated directly with
oxide (dark points in the microstructure) is present. the transition from twinning to dislocation glide as a

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, OCTOBER 2010—2477

Fig. 9—SEM micrograph of the structure of alloy A after 20 pct of Fig. 11—Potentiodynamic curve of the developed alloys and two ref-
plastic elongation. erence TWIP and TRIP steels in 0.5 M H2SO4 electrolyte, potential
in mV compared with the standard hydrogen electrode.

located between –300 mV to 0 mV SHE, followed by a

passive region where the current density is in the order
of 1–20 lA/cm2 and a subsequent transpassive region.
For the three developed alloys in the active area, the
corrosion potential and the critical current, no signifi-
cant differences are found. The passive area could be
associated with the formation of chromium-rich layers.
For the A and B alloys, the passivation current seems to
be more stable than that of the C alloy because
manganese forms more unstable films.[6]
In comparison with the developed alloys, the refer-
ence steels show significantly larger active areas con-
cerning the extent of the potentials and the height of
current density. Passivation occurs, however, when the
Fig. 10—XRD of the alloy A after 20 pct and 30 pct of plastic current densities are of about 1 9 105 lA/cm2, values
deformation. that are 4 or 5 orders of magnitude higher. The
corrosion potential of the TRIP steel is more noble
because of the lower content of manganese. Its passive
plasticity mechanism, with the latter one being more area compared with that of the TWIP reference is wider
preponderant compared with the work hardening of the because of the influence of aluminium and silicon ending
alloys A and C. Moreover, the result can be interpreted to form stable oxide layers.
as an effect of the change in the stacking fault energy;
according to the predicted SFE, the material C exhibits
an SFE of 47.1 mJ/m2 (Table IV), being the highest of IV. SUMMARY AND CONCLUSIONS
the calculated values for the three different compositions
and above the expected value for twinning as the main Based on thermodynamic calculations, a new series of
deformation mechanism. The twinning is evidenced in Fe-Mn-Cr-C-N steels was developed. The results are
Figure 9, corresponding to a micrograph of the micro- fully austenitic microstructures without d-ferrite or
structure of alloy A after 20 pct of plastic elongation. pores. The diffusion calculations made for the prediction
XRDs taken from the samples of alloy A after 20 pct of the diffusion annealing process indicated the tendency
and 30 pct of plastic elongation, show no formation of of C and N to cosegregate with Cr and Mn. Also, it was
e-martensite or a-martensite, as shown in Figure 10. shown that even several days of heat treatment would
Only the reflections corresponding to the fcc crystal not lead to a fully flat composition profile of manganese.
structure of the austenite are detected. The yield strength of the materials is increased with
the total content of carbon and nitrogen, being in
agreement with the classic theory of strengthening by
E. Corrosion
interstitial elements in solid solution.
The electrochemical analysis performed in 0.5 M The cold work hardening of the alloys during plastic
H2SO4 electrolyte is shown in Figure 11. It is shown, deformation at room temperature is related to twinning
that the active range of the developed materials is as a deformation mechanism, which is in agreement with

2478—VOLUME 41A, OCTOBER 2010 METALLURGICAL AND MATERIALS TRANSACTIONS A

the the predicted values of SFE based on thermody- 11. S. Allain, P.J. Chateau, O. Bouaziz, S. Migot, and N. Guelton:
namic modeling. Mater. Sci. Eng. A, 2004, vols. 387–389, pp. 158–62.
12. K. Ishida: Phys. Stat. Sol. A,, 1976, vol. 36, pp. 717–28.
13. A.I. Yakubtsov, A. Ariapour, and D.D. Perovic: Acta Mater.,
1999, vol. 47 (4), pp. 1271–79.
REFERENCES 14. Y. Lee and C. Choi: Metall. Mater. Trans A, 2000, vol. 31A,
1. J. Kim, D. Han, S. Baik, and Y. Lee: Mater. Sci. Eng A, 2004, pp. 355–60.
vol. 378, pp. 323–27. 15. N.Y. Petrov: Scripta Mater., 2005, vol. 53 (10), pp. 1201–06.
2. G. Frommeyer, U. Brux, and P. Neumann: ISIJ Int., 2003, vol. 43 16. S. Kibey, B.J. Liu, J.M. Curtis, D.D. Johnson, and H. Sehitoglu:
(3), pp. 438–46. Acta Mater., 2006, vol. 54 (11), pp. 2991–3001.
3. S. Zaefferer, J. Ohlert, and W. Bleck: Acta Mater., 2004, vol. 52, 17. H. Berns, G.V. Gavriljuk, S. Riedner, and I.A. Tyshchenko: Steel
pp. 2765–78. Res., 2007, vol. 78 (9), pp. 714–19.
4. D. Barbier, N. Gey, S. Allain, N. Bozzolo, and M. Humbert: 18. G.V. Gavriljuk and H. Berns: High Nitrogen Steels: Structure,
Mater. Sci. Eng. A, 2009, vol. 500, pp. 196–206. Properties, Manufacture, Applications, Springer, Berlin, Germany,
5. J.C. Altstetter, P.A. Bentley, J.W. Fourie, and N.A. Kirkbride: 1999.
Mater. Sci. Eng., 1986, vol. 82, pp. 13–25. 19. G.V. Gavriljuk, D.B. Shanina, and H. Berns: Mater. Sci. Eng. A,
6. S.Y. Zhang and X.M. Zhu: Corrosion Sci., 1999, vol. 41, pp. 1817– 2008, vols. 481–482, pp. 707–12.
1833. 20. B. Shanina, G.V. Gavriljuk, and H. Berns: Steel Res., 2007, vol. 78
7. M. Bobby Kannan, K.R. Singh Raman, and S. Khoddam: Cor- (9), pp. 724–28.
rosion Sci., 2008, vol. 50, pp. 2879–84. 21. D.B. Shanina and G.V. Gavriljuk: Mater. Sci. Forum, 2010,
8. K. Sato, M. Ichinose, Y. Hirotsu, and Y. Inoue: ISIJ Int., 1989, vols. 638–642, pp. 3015–20.
vol. 29, pp. 868–877. 22. G.V. Gavriljuk, I.A. Tyshchenko, N.O. Razumov, N.Y. Petrov,
9. O. Grassel, L. Kruger, G. Frommeyer, and L.W. Meyer: Int. J. D.B. Shanina, and H. Berns: Mater. Sci. Eng. A, 2006, vol. 420 (1–2),
Plast., 2000, vol. 16, pp. 1391–409. pp. 47–54.
10. G. Olson and M. Cohen: Metall. Trans. A, 1976, vol. 7A, 23. G.V. Gavriljuk, D.B. Shanina, and H. Berns: Acta Mater., 2008,
pp. 1897–904. vol. 56 (18), pp. 5071–82.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 41A, OCTOBER 2010—2479