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

net/publication/275847306

Precipitation strengthening in high manganese austenitic TWIP steels

Article in International Journal of Materials Research (formerly Zeitschrift fuer Metallkunde) · May 2011
DOI: 10.3139/146.110508

CITATIONS READS

78 928

8 authors, including:

C. P. Scott Frederic Danoix

CanmetMATERIALS French National Centre for Scientific Research
62 PUBLICATIONS 2,373 CITATIONS 182 PUBLICATIONS 2,905 CITATIONS

SEE PROFILE SEE PROFILE

Benoit Malard Caroline Curfs

Institut National Polytechnique de Toulouse Helmholtz-Zentrum Geesthacht
62 PUBLICATIONS 665 CITATIONS 81 PUBLICATIONS 1,552 CITATIONS

SEE PROFILE SEE PROFILE

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

PhD thesis View project

Spinodal decomposition View project

All content following this page was uploaded by Benoit Malard on 15 May 2017.

The user has requested enhancement of the downloaded file.

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

Colin Scotta, Blandine Remya, Jean-Louis Colletb, Aurelie Caela, Cuimin Baod,
Frederic Danoixd, Benoît Malardc, Caroline Curfse
a
ArcelorMittal Research, Automotive Product Development, Maizières-lès-Metz, France
b
CRM, Product Technology, Zwijnaarde, Belgium
c
SIMAP, University of Grenoble, Grenoble, France
d
GPM, University of Rouen, Rouen, France
e
ESRF, Grenoble, France
Not for use in internet or intranet sites. Not for electronic distribution.

Precipitation strengthening in high manganese

austenitic TWIP steels
hardening mechanisms and mechanical properties of Fe–
At low strains (e < 0.25), the precipitation of Ti, Nb, V car- Mn–C TWIP steels, under industrially relevant processing
bides in austenitic Fe – Mn – C steels which exhibit twinning conditions. No systematic study of precipitation strengthen-
induced plasticity has no significant effect on work harden- ing in cold rolled and annealed TWIP steels exists in the lit-
ing. Strengthening follows the Orowan mechanism and can erature; however several authors have reported on the preci-
be described by a simple yield stress offset. However, at pitation of transition metal carbides in high manganese
higher strains (e > 0.3) the work hardening rate decreased austenitic steels [4 – 21]. Most of the earlier data are related
slightly. Synchrotron X-ray diffraction analysis suggested to attempts at precipitation strengthening in Hadfield steels
that this was due to a reduction in the kinetics of twin for- but recently there has been interest in hardening Fe – Mn
mation. The highest strengthening coefficient in cold strips based shape memory alloys. Out of eighteen studies identi-
was obtained with Ti additions £ 0.1 wt.% (+1 380 MPa/ fied eleven concern V precipitation [6, 7, 9 – 11, 13, 15 –
wt.% Ti) but the effect quickly saturated after an increase 18, 20]; five Nb precipitation [4, 8, 10, 12, 14]; two
of *+150 MPa. With Nb additions only modest hardening (Ti + V) precipitation [5, 19] and one (Ti + W) precipita-
(+187 MPa/wt.%) could be achieved. The strengthening tion [21]. Most of the articles describe experiments carried
due to V was > 530 MPa/wt.% for V additions £ 0.4 wt.% out on ingots or hot rolled plates. Only three reported on
Saturation effects are less critical with V additions and cold strips [4, 13, 18] but the precipitation heat treatments
yield stress increases of +375 MPa were demonstrated. described were all incompatible with modern flat carbon
production parameters. None of the authors determined the
www.ijmr.de

Keywords: TWIP steels; Precipitation strengthening; effect of high manganese concentrations on precipitate sol-
Work hardening; Twinning kinetics ubility products and, on the rare occasions where any preci-
pitation calculations were made, standard literature values
were taken. Few of the studies provide chemical extraction
1. Introduction data for precipitate volume fractions and only one system-
atic TEM investigation of precipitate size histograms and
W 2011 Carl Hanser Verlag, Munich, Germany

Fe – Mn – C austenitic steels which exhibit the TWinning number densities was found [18]. In conclusion, the litera-
Induced Plasticity (TWIP) effect are currently being devel- ture results on precipitation in high manganese austenitic
oped for weight reduction and energy absorption applica- steels are open to qualitative interpretations only.
tions in the automotive sector [1 – 3]. Like the Hadfield . The hardening effect of V and (Ti + V) is well estab-
steels to which they are related, TWIP alloys possess re- lished [19, 20]
markable strain hardening properties which promote high . The hardening effect of Nb appears to be rather low at
strength and high formability. One factor limiting their room temperature [4, 8, 10]
immediate implementation is their relatively low yield . There are insufficient data on Ti additions to draw any
strength, especially when compared with existing advanced firm conclusions.
high strength (AHSS) steels. This is particularly important It is clear that the literature is of limited use in determining
for anti-intrusion (crash) assemblies. Increasing the elastic the optimum carbide forming element and thermomechani-
limit by grain refinement is not practical as optimised cold cal treatment. The main concerns are the impact of high Mn
rolled and annealed TWIP alloys are already naturally contents on the solubility products for transition metal car-
fine-grained (dc < 3 lm) and further significant reductions bides and nitrides in austenite [22] and the effect on the var-
are beyond the capabilities of current production tools. ious diffusion coefficients. A few studies do exist concern-
Pre-straining by rolling is another possibility; however our ing the influence of Mn on the established solubility
experience shows that this increases anisotropy and can products of VC [23, 24]. Both authors found that increasing
strongly reduce formability. Thus there is a strong interest the Mn content increased the solubility of VC in austenite,
in precipitation strengthening by microalloying. In this but the change appears to be rather small for Mn composi-
paper we describe the effects of three common microalloy- tions up to 3 wt.%. No information is available at higher
ing elements, Ti, Nb and V on the microstructure, strain Mn contents. Concerning NbC, Koyama et al. [25] found

538 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

that at temperatures above 1 050 8C, Mn additions up to tated weight fractions were quantified using a plasma torch
10 wt.% increased the activity of Nb in austenite but de- coupled with an optical emission spectrometer (ICP-OES).
creased the activity of C by a greater amount. The overall For Ti, Nb and V the measurement error is evaluated to be
result was to strongly increase the solubility of NbC in aus-
5 ppm wt. in the measurement range 10 – 100 ppm wt.
tenite at higher temperatures. A similar effect was noted by and
10 ppm wt. for values above 100 ppm wt. Mechani-
Abken et al. [26] at lower temperatures (875 8C). No infor- cal testing was carried out at a constant strain rate of
mation was found on the effect of Mn on the solubility of 0.008 s – 1 using ISO 12.5 · 50 mm tensile test specimens
TiC in austenite. which were machined using a high speed mill equipped
Not for use in internet or intranet sites. Not for electronic distribution.

with carbide cutting tools. The tensile axis was cut perpen-
2. Experimental procedure dicular to the sheet rolling direction.

2.1. Specimen preparation 3. Results

All the steel compositions used in this study were cast as Figure 1 shows true stress – true strain curves for a variety
15 kg ingots in a laboratory vacuum induction furnace and of Fe – Mn – C TWIP steels in large grain (dc > 10 lm) for-
solidified under still air conditions. Regarding reheating, mat. Three base compositions have been extensively inves-
the relatively low solidus temperature of high-Mn alloys tigated: Fe-22Mn-0.6C, Fe-18Mn-0.6C and Fe-17Mn-0.9C.
precludes the application of a homogenisation treatment to In this paper the nomenclature Fe-22Mn-0.6C represents a
dissolve large TiC or NbC inclusions. For example, the base ternary alloy containing 22 wt.% Mn, 0.6 wt.% C and the
alloy Fe-22Mn-0.6C has a start melting temperature of balance as Fe. The first two are (relatively) low carbon
1 240 8C
20 8C. In consequence, the reheating and hot TWIP grades which are spot weldable with ferritic steels
rolling parameters were chosen to be the following:
. Reheating at 1 180 8C for 15 min in an argon atmo-
sphere.
. Reduction from approximately 70 mm to a final thick-
ness of 3 mm.
. End rolling temperature > 900 8C.
. Fast water cooling (30 K s – 1).
Coiling simulations were done at low temperatures
(< 200 8C) to avoid the formation of Mn-rich intergranular
cementite which strongly reduces ductility [2]. A pilot labo-
ratory cold rolling mill was used to reduce the hot strip
thickness from 3 mm to 1.5 mm (50 % reduction). Recrys-
tallisation annealing was carried out in a high vacuum
AET 1 450 fast annealing simulator under a back filled high
purity Ar atmosphere (dew point < – 60 8C).
www.ijmr.de

Fig. 1. Experimental true stress–true strain curves for Fe–Mn–C TWIP

steels. All concentrations are given in wt.%.
2.2. Analysis techniques

Due to the fine grain size of these alloys, the specimens [2] whilst the latter shows one of the highest work harden-
were examined using backscattered electron imaging (ori- ing rates of any of the tested compositions [28, 29]. While
entation contrast mode) in a JEOL JSM-7001F field emis- the strain hardening coefficients of these steels are quite
W 2011 Carl Hanser Verlag, Munich, Germany

sion gun scanning electron microscope (FEG-SEM). Mea- remarkable, the yield stress (YS) is rather low, typically
surements of grain sizes and recrystallised fractions were < *400 MPa. Decreasing the grain size is an effective
carried out using an Aphelion quantitative image analysis method for increasing the elastic limit and Fig. 2a shows
system. The precipitation state was characterised using the variation in engineering stress-engineering strain re-
three complementary techniques; a Philips CM20 200 kV sponse for Fe-22Mn-0.6C cold strips as the mean grain size
field emission gun transmission electron microscope is reduced. In Fig. 2b we have plotted the variation in YS as
(FEG-TEM) was used to analyse the composition and size a function of grain size for the same alloy. The data can be
histograms of precipitates extracted on carbon extraction well fitted using the standard Hall–Petch type relation
replicas, 3D-Tomographic Atom Probe (3D-TAP) studies (shown as a solid curve):
were carried out on instruments developed at the University
of Rouen [27] to determine the composition of the smallest ky
YS ¼ r0 þ pffiffiffi ð1Þ
particles, and precipitate volume fractions were determined d
by selective chemical dissolution. In the latter technique
samples were etched in hydrochloric acid and rinsed in where r0 is the lattice friction stress, ky is the strengthening
methanol and then placed in an electrolytic bath containing coefficient and d is the grain size. The fitted values for r0
lithium chloride and salicylic acid diluted in methanol. Ap- and ky are given in Table 1 which is contained in Fig. 2b.
proximately 1.5 g of metal was dissolved at an applied cur- Unfortunately, industrial process limitations in the cold
rent density of 20 mA cm – 2. The precipitates were recov- rolling and annealing steps limit the minimum achievable
ered by passing the electrolyte through a polycarbonate grain size to *2.5 lm for this grade. Thus the maximum
filter of porosity 200 nm and the filter and precipitate resi- YS of cold strips is of the order of 450 MPa, whereas a
due were subsequently mineralised. Finally, the precipi- reasonable target for automotive applications is 600 –

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 539

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

700 MPa. The following sections describe the results of ex- the highest attainable industrial reduction (50 %). Next the
periments on the three base compositions microalloyed annealing parameters were adjusted so as to assure com-
with Ti, V and Nb. The hot strips were cold rolled using plete recrystallisation and, as far as possible, optimise the
Not for use in internet or intranet sites. Not for electronic distribution.

Table 1.

(a) (b)

Fig. 2. (a) Engineering stress – strain curves for Fe-22Mn-0.6C as a function of grain size. (b) Variation of YS with grain size for
Fe-22Mn-0.6C. The data has been fitted with a Hall–Petch curve (solid line). The Hall–Petch parameters are given in the inset Table 1.

Table 2a. Fe-22Mn-0.6C microalloyed with Ti – heat compositions in wt.%.

Heat C Mn Si Ti N

Fe-22Mn-0.6C 0.66 21.6 0.19 0 < 0.001

Fe-22Mn-0.6C-0.1Ti 0.65 21.7 0.19 0.097 < 0.001
Fe-22Mn-0.6C-0.2Ti 0.65 21.6 0.19 0.197 < 0.001
Fe-22Mn-0.6C-0.3Ti 0.64 21.5 0.19 0.305 < 0.001

Table 2b. Fe-22Mn-0.6C microalloyed with Ti – hot strip microstructure and mechanical properties.

Heat Grain size Ti precipitated Ti in solid YS UTS A50mm

in hot strip solution
www.ijmr.de

(lm) (wt.%) (wt.%) (MPa) (MPa) (%)

Fe-22Mn-0.6C 9.4 – – 345 947 72

Fe-22Mn-0.6C-0.1Ti 10.2 0.073 0.024 341 997 64
Fe-22Mn-0.6C-0.2Ti 8.0 0.174 0.023 351 993 61
Fe-22Mn-0.6C-0.3Ti 9.1 0.277 0.028 335 973 60
W 2011 Carl Hanser Verlag, Munich, Germany

Table 2c. Fe-22Mn-0.6C microalloyed with Ti – cold strip annealing conditions, microstructure, mechanical properties and calculated
Hall-Petch hardening.

Heat Grain size YS UTS A50mm Annealing DYS DYSH – P

(mm) (MPa) (MPa) (%) conditions (MPa) (MPa)

Fe-22Mn-0.6C 2.5 471 1 102 64 740 8C/120 s – –

Fe-22Mn-0.6C-0.1Ti 1.6 609 1 135 57 740 8C/120 s 138 71
Fe-22Mn-0.6C-0.2Ti 1.6 615 1 131 51 740 8C/120 s 144 71
Fe-22Mn-0.6C-0.3Ti 1.5 635 1 145 46 740 8C/120 s 164 83

Table 2d. Fe-22Mn-0.6C microalloyed with Ti – cold strip precipitation parameters and calculated precipitation strengthening.

Heat Total (primary and Ti remaining Ti precipitated Volume fraction Secondary DYSP
secondary) in solid in secondary of secondary TiC mean (Ashby–Orowan)
precipitated Ti solution TiC TiC radius
(wt.%) (wt.%) (wt.%) (f) (X/2) (MPa)

Fe-22Mn-0.6C – – – – – –
Fe-22Mn-0.6C-0.1Ti 0.089 0.008 0.0163 0.000341 1.9 nm 74
Fe-22Mn-0.6C-0.2Ti 0.192 0.005 0.0168 0.000351 – 75
Fe-22Mn-0.6C-0.3Ti 0.294 0.011 0.0165 0.000345 – 74

540 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

precipitation state (volume fraction and mean radius). All is very high; between 75 % and 91 % of the nominal value.
trial parameters were maintained within the time/tempera- Most of this can be considered as “lost” as the primary pre-
ture limits set by current continuous annealing production cipitates are in the form of large inclusions which can attain
lines. In order to make sensible comparisons of the intrinsic 10 lm in size (Fig. 3a). Energy dispersive X-ray analysis
strengthening effect of different microalloying elements, (EDX) and electron diffraction measurements established
the final grain size should ideally be kept constant. Unfortu- that these particles were TiC, almost certainly formed dur-
nately this was not possible, especially for Ti and V addi- ing cooling after casting and not re-dissolved during reheat-
tions which showed strong grain refining effects. For the ing. There are some smaller TiC precipitates formed during
Not for use in internet or intranet sites. Not for electronic distribution.

16 different alloys processed in this study the final cold hot rolling and coiling (Fig. 3b) but none of these are less
strip grain sizes varied between 1.2 – 2.5 lm, the median than *20 nm in diameter.
grain size was 1.8 lm and the standard deviation was
0.39 lm. Where possible, we have tried to separate pure
precipitation strengthening from solid solution hardening 3.1.2. Cold strips
and grain refinement effects.
In contrast to the hot strip results, a noticeable grain refine-
ment effect is seen in all Ti microalloyed cold strips and Ti
3.1. Precipitation of Ti additions promote a strong increase in the cold strip mechan-
ical properties. From Table 2c the measured YS increment,
3.1.1. Hot strips DYS is + 138 MPa for an addition of only *0.1 wt.% Ti.
However, the effect quickly saturates as the improvement is
The basic Fe-22Mn-0.6C matrix was microalloyed with Ti only + 164 MPa for the heat containing 0.3 wt.% Ti. In Ta-
additions of 0.1 wt.%, 0.2 wt.% and 0.3 wt.%. The compo- ble 2c we have calculated the contribution DYSH–P due to
sitions are given in Table 2a together with the measured mi- grain refinement, using the Hall–Petch data of Fig. 2b. We
crostructural and mechanical parameters in the hot rolled found between + 71 MPa and + 83 MPa of strengthening due
(Table 2b) and cold-rolled and annealed (Table 2c) states. to this effect. The remainder, DYS – DYSH–P must be due to
Clearly the Ti additions do not have a strong effect on either strengthening from secondary precipitates formed during
the grain size or the mechanical properties of the hot bands continuous annealing and/or from solid solution effects. The
(apart from introducing a monotonic decrease in the elonga- latter can be estimated using the data from Irvine et al. [30]
tion), in spite of the fact that the fraction of precipitated Ti who measured a solid solution strengthening coefficient of
www.ijmr.de

Fig. 3. (a) TEM extraction re-

plica from heat Fe-22Mn-
0.6C-0.3Ti hot strip showing
large TiC inclusions formed
W 2011 Carl Hanser Verlag, Munich, Germany

during solidification.
(b) TEM extraction replica
from heat Fe-22Mn-0.6C-0.2Ti
hot strip showing TiC precipi-
tates > 20 nm.
(c) TEM extraction replica
(a) (b) from heat Fe-22Mn-0.6C-0.1Ti
cold strip showing secondary
TiC precipitates formed during
continuous annealing (marked
with arrows).
(d) Size distribution histogram
of secondary TiC in Fe-22Mn-
0.6C-0.1Ti cold strips (636 pre-
cipitates analysed).

(c) (d)

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 541

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

23.4 MPa/wt.% for Ti in austenitic stainless steel. Thus in comes from very small TiC precipitates formed during the
our case the maximum solid solution contribution is negligi- cold strip annealing cycle. However, this hardening effect
ble; *0.2 MPa. The chemical extraction data in Table 2d rapidly saturates at Ti contents > 0.1 wt.% for two reasons –
confirms that most of the Ti remaining in solid solution in . No further significant grain refinement is seen for Ti ad-
the hot band precipitates during continuous annealing after ditions above 0.1 wt.%.
cold rolling, producing a bimodal distribution of particle . Essentially all of the precipitation strengthening occurs
sizes. The TEM image in Fig. 3c shows a clear example of in cold strips – however for Ti additions between
this; one large primary TiC particle is surrounded by much 0.1 wt.% and 0.3 wt.% the amount of Ti present in solid
Not for use in internet or intranet sites. Not for electronic distribution.

smaller secondary precipitates (marked with arrows). A size solution before recrystallisation annealing is indepen-
histogram based on measurements of 636 s TiC precipitates dent of the nominal concentration.
shown in Fig. 3d confirms that the secondary particles are
extremely small; their mean radius is only 1.9 nm. The cold 3.2. Precipitation of Nb
strip precipitation data summarised in Table 2d indicate that,
in the composition range and process conditions tested, the 3.2.1. Hot strips
amount of Ti which precipitates as secondary TiC during
continuous annealing is essentially independent of the The compositions of the Nb microalloyed TWIP steel and
initial Ti composition, varying between 0.0163 wt.% and reference matrix are given in Table 3a together with the
0.0168 wt.% for the three heats. We can compute the ex- measured hot strip and cold strip mechanical and micro-
pected YS increment, DYSP due to secondary TiC precipita- structural parameters (Table 3b – d). The matrix is different
tion using the modified Ashby-Orowan relation proposed by from the previous case in that the Mn content has decreased
Gladman [31]: from 22 % to 18 % and small amounts of Al and Cu were
 1=2    present. However, these changes do not significantly influ-
f X ence the precipitation behaviour of Nb. Unlike in the case
DYSP ¼ 0:538 Gb ln ð2Þ of Ti additions, we do not observe any large inclusions in
X 2b
the hot band, and there is a useful grain refinement effect
where G is the shear modulus in MPa (62 000 MPa at 20 8C as the microalloyed hot strip grain size is reduced from
for Fe-22Mn-0.6C [32]), b is the Burgers vector in lm 8.6 lm to 5.5 lm. Chemical extraction data (Table 3b) indi-
(2.5 · 10 – 4 lm), f is the volume fraction of secondary TiC cated that almost all the Nb had already precipitated as NbC
precipitates and X is the mean particle diameter expressed at this point. We note a strength increase DYS in the hot
in mm. The calculated values are given in Table 2d. Here band of +61 MPa of which we estimate that *38 MPa is
we made the assumption that the mean radius of the second- due to grain refinement (extrapolating from the data of
ary precipitation in the heats with 0.2 wt.% and 0.3 wt.% Ti Fig. 2b). The contribution from Nb solid solution hardening
additions is the same as that measured by TEM for the is negligible, (*1 MPa) [30] so that DYSP due to NbC pre-
0.1 wt.% Ti heat. This is reasonable as the amount of Ti cipitation in hot strips is approximately *23 MPa. An in-
present in solid solution before annealing is very similar, verse calculation using the modified Ashby–Orowan equa-
tion indicates that the primary NbC precipitate radius
www.ijmr.de

the C is in excess and the applied thermal treatment is iden-

tical. This hypothesis appears to be justified as can be seen should therefore be very large; *50 nm.
in the graphical summary shown in Fig. 4, where the differ-
ence between the experimentally observed YS increase and 3.2.2. Cold strips
the sum of the calculated Hall–Petch and modified Ashby–
Orowan contributions (DYSH–P + DYSP) is shown to be less After cold rolling and annealing at 800 8C there was no
than 5 % for all three alloys. In summary, for Ti additions longer any difference in grain size between the microal-
W 2011 Carl Hanser Verlag, Munich, Germany

£ 0.1 wt.% the total strengthening is very high. Roughly loyed and reference heats (Table 3c). In fact the grade con-
half of this is due to grain refinement and the other half taining Nb appeared to have a slightly coarser cold strip
grain size than the reference (hence the small negative val-
ues of DYSH–P in Table 3c). Selective chemical extraction
indicated that there was virtually no change in the amount
of Nb precipitated which is in agreement with the mechani-
cal tests as the measured DYS was only + 15 MPa. TEM ex-
traction replicas (Fig. 5a) showed the presence of large
NbC precipitates almost certainly inherited from the hot
band. Their mean radius was 36 nm and the minimum ra-
dius was 18 nm (Fig. 5b). The predicted hardening increase
DYSP from this precipitate distribution using the modified
Ashby-Orowan equation is only + 32 MPa. Clearly the pre-
cipitates are too large to have any appreciable strengthening
effect. A second annealing cycle at the lowest temperature
compatible with complete recrystallisation (750 8C/120 s)
was tested. After this treatment the grain size was similar
to the reference heat and an increment of 0.023 wt.% Nb
had precipitated as secondary NbC. A modest strength in-
Fig. 4. Contributions to the YS increase from Ti additions to Fe-22Mn- crease (DYS = + 58 MPa) was achieved. As there is no grain
0.6C cold strips. refinement and practically all the Nb is precipitated, we as-

542 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

sume that this is entirely due to NbC precipitation strength- tate radius of 3 nm would be sufficient to generate this in-
ening. An inverse calculation using the modified Ashby- crease. However, due to the small effect on the mechanical
Orowan equation indicates that a secondary NbC precipi- properties this was not checked.

Table 3a. Fe-18Mn-0.6C microalloyed with Nb – heat compositions in wt.%.

Not for use in internet or intranet sites. Not for electronic distribution.

Heat C Mn Si Nb N

Fe-18Mn-0.6C 0.65 18.0 0.2 0 < 0.001

Fe-18Mn-0.6C-0.31Nb 0.65 18.0 0.2 0.31 < 0.001

Table 3b. Fe-18Mn-0.6C microalloyed with Nb – hot strip microstructure and mechanical properties.

Heat Grain size Precipitated Nb YS UTS A50mm DYS DYSH – P

(lm) (wt.%) (MPa) (MPa) (%) (MPa) (MPa)

Fe-18Mn-0.6C 8.6 – 366 827 62 – –

Fe-18Mn-0.6C-0.31Nb 5.5 0.2733 427 868 55 61 38

Table 3c. Fe-18Mn-0.6C microalloyed with Nb – cold strip annealing conditions, microstructure, mechanical properties and calculated
Hall – Petch hardening.

Heat Grain size YS UTS A50mm Annealing DYS DYSH – P

(lm) (MPa) (MPa) (%) conditions (MPa) (MPa)

Fe-18Mn-0.6C 2.2 474 989 64 800 8C/180 s – 0

Fe-18Mn-0.6C-0.31Nb 2.5 489 937 63 800 8C/180 s 15 – 19
Fe-18Mn-0.6C-0.31Nb 2.3 532 942 50 750 8C/120 s 58 –7

Table 3d. Fe-18Mn-0.6C microalloyed with Nb – cold strip precipitation parameters and calculated precipitation strengthening.

Heat Total (primary Nb remaining Nb precipitated Volume fraction Primary DYSP

and secondary) in solid in secondary of primary and NbC mean (Ashby–Orowan)
precipitated Nb solution NbC secondary NbC radius
(wt.%) (wt.%) (wt.%) (f) (X/2) (MPa)
www.ijmr.de

Fe-18Mn-0.6C – – – – – –
Fe-18Mn-0.6C-0.31Nb 0.2744 0.0356 0.001 0.00322 36 nm 32
(800 8C)
Fe-18Mn-0.6C-0.31Nb 0.2958 0.0142 0.0225 0.00347 – –
(750 8C)
W 2011 Carl Hanser Verlag, Munich, Germany

(a) (b)

Fig. 5. (a) TEM extraction replica from heat Fe-18Mn-0.6C-0.31Nb cold strip annealed at 800 8C. (b) Size distribution histogram of NbC
Fe-18Mn-0.6C-0.31Nb cold strips annealed at 800 8C (397 precipitates analysed).

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 543

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

Table 4a. Fe-22Mn-0.6C and Fe-17Mn-0.9C microalloyed with V – heat compositions in wt.%.

Heat C Mn Si V N

Fe-22Mn-0.6C 0.66 21.6 0.2 0 < 0.001

Fe-22Mn-0.6C-0.16V 0.64 21.8 0.01 0.16 < 0.001
Fe-22Mn-0.6C-0.21V 0.66 22.0 0.19 0.21 0.011
Fe-22Mn-0.6C- 0.4V 0.59 21.7 0.18 0.40 0.028
Fe-22Mn-0.6C- 0.9V 0.63 21.6 0.01 0.87 < 0.001
Not for use in internet or intranet sites. Not for electronic distribution.

Fe-17Mn-0.9C 0.9 16.7 0.5 0 0.031

Fe-17Mn-0.9C-0.1V 0.9 17.3 0.5 0.09 0.024
Fe-17Mn-0.9C-0.17V 0.9 17.2 0.5 0.17 0.028
Fe-17Mn-0.9C-0.45V 0.9 16.6 0.5 0.45 0.038
Fe-17Mn-0.9C-1.1V 0.9 16.4 0.5 1.09 0.048

Table 4b. Fe-22Mn-0.6C and Fe-17Mn-0.9C microalloyed with V – hot strip microstructure and mechanical properties (NR = not 100 %
recrystallised).

Heat Grain size Precipitated V YS UTS A50mm DYS DYSH – P

(lm) (wt.%) (MPa) (MPa) (%) (MPa) (MPa)

Fe-22Mn-0.6C 9.4 – 345 947 72 – –

Fe-22Mn-0.6C-0.16V 9.9 0.0005 352 972 64 7 –4
Fe-22Mn-0.6C-0.21V 7.1 0.0042 412 992 62 67 26
Fe-22Mn-0.6C- 0.4V 9.2 0.0080 425 985 56 80 2
Fe-22Mn-0.6C- 0.9V NR N/A 586 1 051 44 – –
Fe-17Mn-0.9C 10.1 – 419 1 135 77 – –
Fe-17Mn-0.9C-0.1V 8.1 0.0027 401 1 099 67 – 18 –
Fe-17Mn-0.9C-0.17V 8.4 0.0085 435 1 104 67 16 –
Fe-17Mn-0.9C-0.45V NR N/A 667 1 163 35 – –
Fe-17Mn-0.9C-1.1V NR N/A 733 1 233 29 – –

Table 4c. Fe-22Mn-0.6C and Fe-17Mn-0.9C microalloyed with V – cold strip annealing conditions, microstructure, mechanical proper-
ties and calculated Hall – Petch hardening.

Heat Grain size YS UTS A50mm Annealing DYS DYSH – P

(lm) (MPa) (MPa) (%) conditions (MPa) (MPa)
www.ijmr.de

Fe-22Mn-0.6C 2.5 471 1 102 64 740 8C/120 s – –

Fe-22Mn-0.6C-0.16V 1.8 560 1 182 62 800 8C/180 s 89 50
Fe-22Mn-0.6C-0.21V 1.9 608 1 184 59 800 8C/180 s 137 42
Fe-22Mn-0.6C- 0.4V 1.5 686 1 142 46 800 8C/120 s 215 103
Fe-22Mn-0.6C- 0.9V 1.2 818 1 240 32 850 8C/ 60 s 347 125
Fe-17Mn-0.9C 2.2 514 1 216 50 740 8C/120 s – –
Fe-17Mn-0.9C-0.1V 2.1 590 1 264 49 800 8C/180 s 76 7
W 2011 Carl Hanser Verlag, Munich, Germany

Fe-17Mn-0.9C-0.17V 2.0 654 1 269 47 800 8C/180 s 140 15

Fe-17Mn-0.9C-0.45V 1.6 761 1 361 48 825 8C/120 s 247 52
Fe-17Mn-0.9C-1.1V 1.4 889 1 405 43 825 8C/120 s 375 77

Table 4d. Fe-22Mn-0.6C and Fe-17Mn-0.9C microalloyed with V – cold strip precipitation parameters and calculated precipitation
strengthening.

Heat Total V remaining V increment Volume fraction of V(C, N) DYSP

precipitated V in solid precipitated V(C, N) precipitated mean radius (Ashby–Orowan)
solution in the cold strip in cold strip
(wt.%) (wt.%) (wt.%) (f) (X/2) (MPa)

Fe-22Mn-0.6C – – – – – –
Fe-22Mn-0.6C-0.16V 0.0533 0.107 0.0528 0.0009 N/A N/A
Fe-22Mn-0.6C-0.21V 0.1207 0.089 0.1165 0.0019 3.5 nm 130
Fe-22Mn-0.6C- 0.4V 0.2957 0.108 0.2877 0.0049 N/A N/A
Fe-22Mn-0.6C- 0.9V N/A N/A N/A N/A N/A N/A
Fe-17Mn-0.9C – – – – – –
Fe-17Mn-0.9C-0.1V 0.037 0.053 0.0343 0.0005 N/A N/A
Fe-17Mn-0.9C-0.17V 0.090 0.080 0.0815 0.0014 N/A N/A
Fe-17Mn-0.9C-0.45V 0.429 0.021 0.429 0.0074 3.1 nm 208
Fe-17Mn-0.9C-1.1V 0.995 0.095 0.995 0.0172 N/A N/A

544 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

3.3. Precipitation of V (a)

3.3.1. Hot Strips

Microalloying with V was applied to two base metallurgies;

Fe-22Mn-0.6C and Fe-17Mn-0.9C. Table 4a lists ten differ-
ent compositions which were tested. Table 4b and c contain
the hot and cold strip parameters and Table 4d the cold strip
Not for use in internet or intranet sites. Not for electronic distribution.

precipitation data. Precipitates such as VN, V(C, N) or VC

formed during casting are re-dissolved during reheating at
1 180 8C. From previous work on vanadium microalloyed
TRIP steels [33], the solubility of V in austenite is such that
there will be little or no V precipitation during hot rolling
and coiling as long as the coiling temperature is kept below
*500 8C. This is confirmed by the low concentration of
precipitated V measured in the hot bands (Table 4b) and
also by TEM investigations on thin foils taken from hot
strips. Thus DYSP can be neglected for all but the most
highly alloyed heats. The presence of V in solid solution
does not appear to greatly modify the hot strip grain size
so DYSH–P is small in all cases (< 26 MPa). No significant
increase in YS due to V solid solution strengthening was ob-
served, in accordance with the weak strengthening coeffi-
cient given by Irvine et al. (17 MPa/wt.% V) [30]. The
increase in YS seen at very high V concentrations
(Vtot > 0.4 wt.%) is due to the fact that the hot band recrys-
tallisation was not 100 % complete; non-recrystallised
grains cause the YS to increase and the ductility drops shar- (b)
ply. In conclusion, it is possible to optimise the process pa-
rameters such that nearly all of the added V can be main- Fig. 6. (a) TEM extraction replica from heat Fe-22Mn-0.6C-0.4V cold
tained in solid solution in the hot strip and hence in full strip annealed at 850 8C. (b) Size distribution histogram of V(C, N) in
hard cold rolled strips. It therefore remains available to pre- Fe-22Mn-0.6C-0.4V cold strips annealed at 850 8C.
cipitate during the continuous annealing step.

3.3.2. Cold strips

www.ijmr.de

In common with Ti and unlike Nb, we observe a strong re-

finement of the cold strip grain size with increasing V addi-
tions for both matrix compositions (Table 4c). The quantity
of V precipitated in annealed cold strips can exceed 60 % of
the nominal value, particularly if a small amount of N is
present. For example, it was found that the total amount of
W 2011 Carl Hanser Verlag, Munich, Germany

precipitated V and the mean precipitate radius for the heat

Fe-22Mn-0.6C-0.4V containing 280 ppm wt. N and an-
nealed at 850 8C/120 s are 0.290 wt.% and 3.4 nm respec-
tively (Fig. 6a and b). The equivalent figures for a similar
composition but without any added N under the same pro-
cessing conditions were 0.267 wt.% and 7.6 nm. The bene-
ficial influence of N is further illustrated in the 3D-TAP
analysis of a small (*3 nm diameter) V precipitate in heat
Fe-22Mn-0.6C-0.21V annealed at 800 8C (Fig. 7). Almost
all of the V precipitates analysed contained significant
amounts of N. TEM data showed that the mean V(C, N)
precipitate radius in this alloy was 3.5 nm. Small angle neu-
tron scattering (SANS) studies carried out on an Fe-18Mn-
0.6C-0.2V alloy which underwent an identical thermome-
chanical treatment indicated a Porod radius of 3.3

0.2 nm [34], providing further confirmation of the TEM
analysis.
With both matrices, a distinct improvement in the cold
strip mechanical properties with V additions was observed.
The YS increases by + 215 MPa for an addition of 0.4 wt.% Fig. 7. 3D-TAP analysis of a small (*3 nm diameter) V precipitate in
V to the Fe-22Mn-0.6C matrix. This increase in strength is heat Fe-22Mn-0.6C-0.21V annealed at 800 8C.

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 545

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

due to a combination of grain refinement and precipitation which are problematic both for processing and for end user
strengthening. Unlike Ti, these effects do not saturate properties. Thus in practice V is the preferred microalloy-
quickly i. e. in the Fe-22Mn-0.6C matrix the grain size is re- ing element for cold rolled high Mn TWIP steels.
duced to 1.2 lm and DYS attains + 347 MPa for 0.9 wt.% V
additions. Some decrease in the strengthening coefficient 4. Discussion
does occur for the highest V additions but not for the same
reason as the Ti alloyed steels. In fact for V additions 4.1. The influence of microalloying on the strain
> 0.4 wt.% the recrystallisation kinetics are slowed so that hardening properties.
Not for use in internet or intranet sites. Not for electronic distribution.

higher annealing temperatures are required and, once the

recrystallisation isotherm exceeds *825 8C, there is a ten- Previous work has shown that the excellent strain hardening
dency for V precipitates to coarsen and the smaller particles properties of high Mn TWIP steels are essentially deter-
are dissolved. mined by the rate of formation of deformation twins [35].
It is therefore important to establish whether or not precipi-
3.4. A comparison of the effects of Ti, V and Nb additions tates interact with the twinning mechanism and thus modify
on YS the work hardening. The presence of dynamic strain ageing
(DSA) causes discontinuous yielding in high Mn TWIP ten-
A summary of the effect of the various microalloying addi- sile test samples and this is visible as sharp serrations which
tions on the cold strip YS for the alloys tested is shown in are superimposed on the normal stress–strain curves. It has
Fig. 8. Here we have fitted the V data with polynomial been shown that the contribution to strain hardening of
curves. Obviously the grain size and the precipitated frac- DSA is negligible compared with that from the dynamic
tions are not constant in this representation so that Fig. 8 Hall–Petch effect introduced by twinning [36]. However,
should be treated very much as an “engineering” diagram. the presence of DSA serrations makes it very difficult to
Nevertheless, the trends shown reflect accurately the micro- study strain hardening, for example with Mecking–Kochs
structural evolutions and are useful from an alloy design plots, as the data must be filtered or smoothed after differ-
point of view. The strengthening potential of each of the entiation. The alternative approach of fitting the raw
elements is quite clearly displayed in this diagram. It is in- stress–strain data by a polynomial function and then differ-
teresting to note that the best fit curves for the two different entiating the fitted curve can lead to errors, especially at
matrices microalloyed with V are very close, in spite of the small strains. We have therefore taken the simplest ap-
large difference in C concentration. For alloy additions proach which is to compare the true stress/true strain flow
< 0.1 wt.% the strengthening order is Ti > V > Nb. Table 5 curves after subtracting the 0.2 % offset YS and introducing
contains the strengthening coefficients for this range. Thus the required strain offset. The results are plotted in Fig. 9
Ti additions are the most efficient way to increase the YS for the three different base compositions.
by up to *+ 150 MPa. Above this limit, Ti saturates and V The data highlights the difference in strain hardening re-
must be used. However, as stated in Section 3.1.1, even sponse between the three basic alloys. In agreement with
small Ti additions generate large amounts of inclusions Fig. 1, we find that Fe-17Mn-0.9C shows the highest work
hardening. It is clear that the presence of precipitates does
www.ijmr.de

not appear to modify the strain hardening behaviour at low

strains (e < 0.25), irrespective of the base composition or
the microalloying element employed. At strains above
0.25 the situation is less clear. There is a monotonic de-
crease in the total amount of work hardening (rmax – r0)
with increasing amounts of microalloying additions for all
W 2011 Carl Hanser Verlag, Munich, Germany

of the studied systems. Sometimes this is simply due to the

alloy fracturing at lower strains; however in other cases
there does appear to be a decrease in the rate of work hard-
ening. The effect is especially noticeable for the Fe-17Mn-
0.9C heat with the highest V additions. It is impossible to
be precise, due to the masking effects of the plastic instabil-
ity from DSA, but in general it appears that the higher the
initial strain hardening response of the matrix and the larger
Fig. 8. A comparison of the yield strength increase in Fe-(18 – 22)Mn-
the precipitated volume fraction, the more obvious is the
(0.6 – 0.9)C cold strips as a function of microalloying additions. saturation effect. We have tried to indicate with arrows in
Fig. 9 the strain values at which the work hardening re-
sponse starts to diverge from the reference non-microal-
Table 5. The strengthening coefficients of microalloying addi- loyed state.
tions of Ti, V, Nb below 0.1 wt.%. This analysis suggests that there is no interaction be-
tween precipitation and deformation twinning at low
Microalloying Strengthening Coefficient for strains. The results of a recent TEM study by Dumay on
element microalloying additions < 0.1 wt.% Fe-22Mn-0.6C-0.2V after straining by 10 % in uniaxial ten-
sion [37] tend to support this hypothesis – at no point did
Nb 187 MPa/wt.%
V > 530 MPa/wt.%
the author detect any observable interaction between twin
Ti 1 380 MPa/wt.% boundaries and V(C, N) precipitates. It is extremely diffi-
cult to study higher strains using TEM as the overall density

546 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels
Not for use in internet or intranet sites. Not for electronic distribution.

Fig. 9. The effect of Ti, Nb and V additions on the flow curves of Fe-17Mn-0.9C, Fe-22Mn-0.6C and Fe-18Mn-0.6C cold strips. The arrows indi-
cate the strain values at which the work hardening response starts to diverge from the reference non-microalloyed state.

of defects quickly becomes too great. Other techniques such

as SEM EBSD do not have the spatial resolution required.
We therefore carried out synchrotron transmission X-ray
diffraction experiments on strained samples in order to
further investigate the interaction between precipitates and
work hardening. The results are discussed in the following
section.

4.2. The effect of microalloying on defect densities.

www.ijmr.de

Plastic deformation of metals introduces point (vacancies,

self-interstitials), line (dislocations), surface (stacking
faults, dislocation cell walls) and volume (shear bands,
twins, voids) defects into the microstructure. As the ap-
plied strain increases, so does the density of these defects.
The relative proportion of each defect depends upon the
W 2011 Carl Hanser Verlag, Munich, Germany

straining conditions and the hardening/damage mecha-

nisms which are intrinsic to the material. Defects disrupt
the perfection of the crystal lattice and thus, in X-ray dif-
fraction experiments, the coherence of each diffracted
beam is reduced. In general the amplitude and symmetry
of the (hkl) peaks in an X-ray diffraction experiment will
be decreased, their half-width broadened and the peak Fig. 10. The effect of plastic strain (uniaxial tension) on the {220}
centre may be shifted. An example of the variations seen X-ray diffraction peak of the Fe-22Mn-0.6C matrix. Data measured in
in the (220) reflection from a Fe-22Mn-0.6C sample as a transmission mode at the ESRF [37]. Note that the intensity scale is
function of tensile strain is shown in Fig. 10. These data logarithmic.
were recorded in transmission mode on pre-strained ten-
sile test samples on the BM02 and ID31 lines at the ESRF
synchrotron facility in Grenoble. The density of different the method derived by Wilkens [39] modified to take into
types of defects can, in principle, be determined by a care- account the effects of planar slip in these alloys. It is
ful study of the position and form of a suitable set of dif- possible, by Fourier analysis of the peak profiles, to deter-
fraction peaks. The problem is that separating out the dif- mine the quantity (1.5
+ b) which is related to the den-
ferent contributions is often difficult or even impossible. sity of stacking faults and twin boundaries. The parameter
In our case we are interested in defects which contribute
is the probability of finding a stacking fault (intrinsic or
directly to strain hardening i. e. dislocations and twins. In extrinsic) on a {111} plane and b is the probability of
his Ph.D. thesis J.-L. Collet [38] showed that it is possible finding a twin boundary on a {111} plane. Unfortunately
to measure dislocation densities in Fe-22Mn-0.6C using it is not possible to unambiguously separate the two com-

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 547

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

ponents, however to a first approximation we can write to suggest that that twinning kinetics in the microalloyed
[38]: grade should begin to show saturation from this point, in
good qualitative agreement with the flow curves in Fig. 9.
ð1:5
þ bÞ The actual physical mechanism controlling the interaction
b ð3Þ
2:5 between precipitation and twinning kinetics at high strains
is not known. Further work is necessary to investigate this
The mean distance d (in nm) between twin planes in the
phenomenon.
<111> direction can then be written:
Not for use in internet or intranet sites. Not for electronic distribution.

1:5 5. Conclusions
d ð4Þ
ð1:5
þ bÞ
We have studied the effects of strengthening by transition
The Fourier analysis technique was carried out for two an- metal carbides of Ti, V and Nb on different Fe–Mn–C and
nealed cold strips, Fe-22Mn-0.6C and Fe-22Mn-0.6C- Fe–Mn–C–Al hot and cold strips. The first important result
0.2V. The alloys were prepared using the same thermo- is that, at low strains (e < 0.25), the precipitation of microal-
mechanical treatment and had a similar final grain size. loying elements has no significant effect on work hardening
The main results are shown in Fig. 11 and Fig. 12. From and that the strengthening effect can be well described by a
Fig. 11 there does not appear to be any significant differ- simple increase in the r0 term in the flow stress equations.
ence in the evolution of the dislocation density with strain The amount of strengthening can be calculated using the
between the two heats. Thus, any difference in strain hard- Ashby–Orowan approach. However, at higher strains
ening must be caused by another mechanism. Figure 12 in- (e > 0.3) there appears to be an interaction which results in
dicates that there is a difference in the rate of formation of reduced strain hardening. This effect is more pronounced
planar faults (twins + stacking faults) at high strains i. e. at for alloys with higher initial work hardening rates and high-
e = 0.4 using Eq. (4) the inferred mean distance between er precipitate volume fractions. Synchrotron X-ray diffrac-
twin planes is 60 nm for Fe-22Mn-0.6C and 84 nm for tion analysis suggests that this is due to a reduction in the
Fe-22Mn-0.2C-0.2V. In fact the two curves start to diverge kinetics of twin formation at high strains. A comparison of
at true strains between 0.2 and 0.3. It is therefore reasonable the strengthening effects of the different microalloying ele-
ments revealed that, under industrially relevant processing
conditions, the element which provides the highest harden-
ing effect in cold strips is Ti. For Ti additions £ 0.1 wt.%,
the strengthening coefficient is +1 380 MPa/wt.% Ti. Un-
fortunately for Ti additions > 0.1 wt.% the strengthening ef-
fect saturates (at *+150 MPa). The effect of Nb additions
is much more modest and only +187 MPa/wt.% could be
achieved, mainly because the mean particle size was too
large to be effective for hardening. Finally, the strengthen-
www.ijmr.de

ing effect due to V is > 530 MPa wt.% for V additions

£ 0.4 wt.%.

The authors would like to thank colleagues at ArcelorMittal Research

for their precious aid, especially B. Lambottin for the selective
chemical dissolution measurements and P. Cugy, Dr. O. Bouaziz and
Fig. 11. The variation of dislocation density q with true strain for heats Dr. S. Allain for stimulating discussions.
W 2011 Carl Hanser Verlag, Munich, Germany

Fe-22Mn-0.6C and Fe-22Mn-0.6C-0.2V calculated by Fourier analysis

of X-ray data from pre-strained samples measured in transmission References
mode at the ESRF.
[1] G. Frommeyer, E.J. Drewes, B. Engl: La Revue de Metallurgie-
CIT 10 (2000) 1245.
[2] C. Scott, S. Allain, M. Faral, N. Guelton: La Revue de Metallur-
gie-CIT 103 (2006) 293. DOI:10.1051/metal:2006142
[3] L. Chen: ISIJ International 47 (2009) 1804.
[4] M.C. Chaturvedi, R.W.K. Honeycombe, D.H. Warrington: J. Iron
and Steel Inst. Dec. (1968) 1236.
[5] Y.G. Matveev, V.S. Bukhtin, D.I. Tarasko: Metallovedenie I Ter-
micheskaya Obrabotka Metallov 3 (1975) 51.
[6] S.J. Harris, N.R. Nag: J. Mat. Sci. 11 (1976) 1320.
DOI:10.1007/BF00545153
[7] K.H. Miska: Mat. Eng. 8 – 77 (1977) 42.
[8] H. Masumoto, H. Yoshimura, T. Akasawa, S. Ohba, T. Harada,
K. Suemune, T. Inoue, T. Ogawa, Y. Hida: Nippon Steel Techni-
cal Report 22 (1983) 47.
[9] O.A. Atasoy: Z. Metallkd. 75 (1984) 463.
[10] Y.G. Kim, Y.S. Park, J.K. Han: Met. Trans. A 16 (1985) 1689.
[11] O.A. Atasoy, K. Ozbaysal, O.T. Inal: J. Mat. Sci. 24 (1989) 1393.
DOI:10.1007/BF02397078
Fig. 12. A comparison of the variation of stacking fault and twin den- [12] K. Hee Han: Mat. Sci. Eng. A 279 (2000) 1.
sities with true strain for heats Fe-22Mn-0.6C and Fe-22Mn-0.6C- DOI:10.1016/S0921-5093(99)00690-5
0.2V calculated by Fourier analysis of X-ray data from pre-strained [13] Z. Guo, T. Furuhara, T. Maki: Scripta Mat. 45 (2001) 525.
samples measured in transmission mode at the ESRF. DOI:10.1016/S1359-6462(01)01053-3

548 Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5

 C. Scott et al.: Precipitation strengthening in high manganese austenitic TWIP steels

[14] S. Kajiwara, D. Liu, T. Kikuchi, N. Shinya: Scripta Mater. 44 [35] O. Bouaziz, S. Allain, C. Scott: Scripta Mater. 58 (2008) 484.
(2001) 2809. DOI:10.1016/S1359-6462(01)00978-2 DOI:10.1016/j.scriptamat.2007.10.050
[15] Z. Guo, N. Kimura, S. Tagashira, T. Furuhara, T. Maki: ISIJ Int. [36] S. Allain, P. Cugy, C. Scott, J.-P. Château, A. Rusinek, A. Deschamps:
42 (2002) 1033. Int. J. Mat. Res. 99 (2008) 734.
[16] T. Furuhara, T. Shinyoshi, G. Miyamoto, J. Yamaguchi, N. Sugita, [37] A. Dumay: Ph.D. thesis, Institut National Polytechnique de Lor-
N. Kimura, N. Takemura, T. Maki: ISIJ 43 (2003) 2028. raine (2008).
[17] H. Kubo, K. Nakamura, S. Farjami, T. Maruyama: Mat. Sci. Eng. [38] J.-L. Collet: Ph.D. thesis, Institut Polytechnique de Grenoble
378 (2004) 343. (2009).
[18] Y. Yazawa, T. Furuhara, T. Maki: Acta Mater. 52 (2004) 3727. [39] M. Wilkens: Phys. Stat. Sol. B 171 (1970).
Not for use in internet or intranet sites. Not for electronic distribution.

DOI:10.1016/j.actamat.2004.04.027
[19] H. El-Faramawy: Steel Grips 3 (2005) 360. (Received July 2, 2010; accepted March 2, 2011)
[20] V.V. Sagaradze, I.I. Kositsyna, M.L. Mukhin, Y.V. Belozerov, Y.R.
Zaynutdinov: Mat. Sci. Eng. A 481 (2008) 747.
DOI:10.1016/j.msea.2007.02.158
[21] A.K. Srivastava, K. Das: Mat. Sci. Eng. A 516 (2009) 1. Bibliography
DOI:10.1016/j.msea.2009.04.041
[22] T. Gladman: The Physical Metallurgy of Microalloyed Steels, DOI 10.3139/146.110508
The Institute of Materials, London (1997). Int. J. Mat. Res. (formerly Z. Metallkd.)
[23] W. Huang: Met.Trans. A 22 (1991) 1911. 102 (2011) 5; page 538 – 549
[24] K. Han: Scripta Mater. 28 (1992) 699. # Carl Hanser Verlag GmbH & Co. KG
[25] S. Koyama, T. Ishii, K. Narita: Nippon Kink. Gakk 35 (1971) 1089. ISSN 1862-5282
[26] M.G. Abken, I. Weiss, J.J. Jonas: Acta. Mater. 29 (1981) 111.
DOI:10.1016/0001-6160(81)90092-4
[27] G. DaCosta, F. Vurpillot, A. Bostel, M. Bouet, B. Deconihout: Correspondence address
Review of Scientific Instruments 76 (2005) 013304.1 – 01334.8.
[28] P. Cugy, N. Guelton, C. Scott, F. Stouvenot, M.-C. Theyssier: Pa- Dr. Colin Scott
tent WO 2006/056670. Areva NP
[29] O. Bouaziz, H. Zurob, B. Chebab, J.D. Embury, S. Allain, M. Huang: 10 rue Juliette Récamier, 69006 Lyon, France
Mat. Sci. Tech. (2010) article in press. Tel.: +33 4 72 74 75 52
DOI 10.1179/026708309X12535382371852 Fax: +33 4 72 74 81 45
[30] K.J. Irvine, T. Gladman, F.B. Pickering: J. Iron and Steel Inst., E-mail: [email protected]
July (1969) 379.
[31] T. Gladman: Mat. Sci. Tech. 15 (1999) 30.
DOI:10.1179/026708399773002782
[32] S. Allain: Ph.D. thesis, Institut National Polytechnique de Lor- You will find the article and additional material by enter-
raine (2002). ing the document number MK110508 on our website at
[33] F. Perrard, C. Scott: ISIJ Int. 47 (2007) 1168. www.ijmr.de
[34] B. Malard: in preparation.
www.ijmr.de
W 2011 Carl Hanser Verlag, Munich, Germany

Int. J. Mat. Res. (formerly Z. Metallkd.) 102 (2011) 5 549

View publication stats