Materials Science and Engineering A346 (2003) 189 /195
www.elsevier.com/locate/msea
Formation of ultra-fine ferrite microstructure in warm rolled and
annealed C Mn steel /
D.B. Santos a, R.K. Bruzszek a, P.C.M. Rodrigues b, E.V. Pereloma c,*
a
Department of Metallurgical and Materials Engineering, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil
b
Department of Mechanical Engineering, FUNREI, Sao João Del’Rei, MG, Brazil
c
School of Physics and Materials Engineering, P.O. Box 69M, Monash University, Vic., 3800, Australia
Received 28 January 2002; received in revised form 10 July 2002; accepted 11 July 2002
Abstract
Laboratory simulations of warm rolling followed by intercritical annealing of a low carbon 0.15%C /1.39%Mn steel have been
performed. The effects of austenitising temperature, amount of deformation and annealing time on microstructure and mechanical
properties have been investigated. The results have shown that the homogeneous ultra-fine ferrite grain microstructure (/1.1 /1.2
mm average grain size) has been achieved after austenitising at 900 8C, warm rolling with 3 passes of 20% reduction each at 700 8C
and annealing at 800 8C for 60 min. This correlates to the 20% improvement in mechanical properties compared to traditional
industrial hot rolled steel.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Ultra-fine ferrite; Low carbon steel; Warm rolling; Intercritical annealing
1. Introduction (carried out in the temperature region between Ac1 and
Ac3 temperatures) [3,4]. For steels with martensitic
There has been significant interest in the formation of microstructure cold rolling is generally performed before
fine ferrite grain structure (B/2 mm) in low carbon steels. intercritical annealing [6], while in this work warm
This ultra grain refinement offers significantly higher rolling is applied.
yield strengths and lower ductile-brittle transition tem- The objective of this work was to study the micro-
peratures with minimal alloying [1,2]. This method is structural evolution during thermomechanical proces-
important as a hardening mechanism in these steels sing to obtain ultra-fine ferrite grain size in a carbon-
when they are used in welding applications, and is less manganese steel using quenching, laboratory warm
expensive than utilising high alloy steels [3 /8]. rolling and intercritical annealing. It was also an aim
There are several processing routes to obtain the of this work to evaluate the mechanical performance of
ultra-fine ferrite grain microstructure. It has been the steel with ultra-fine grains.
demonstrated in a variety of steel compositions that
ultra-fine ferrite can be produced using the so called
strain-induced transformation rolling process [9,10].
2. Experimental procedure
However, another way to achieve ferrite grain refine-
ment is to utilise the dynamic recrystallisation of ferrite The chemical composition of the steel investigated is
during warm working [11,12]. Fine microstructure can given in Table 1. The warm rolling process was carried
also be produced in steels subjected to heat treatment out on a 500 kN laboratory mill with rolling speed of 25
which combines warm rolling and intercritical annealing m min 1. The processing schedules are shown in Fig. 1.
The selection of the processing parameters was justified
* Corresponding author. Tel.: /61-399054916; fax: /61-399054940
elsewhere [4]. The specimens of 16.2-mm thick slab
E-mail address: elena.pereloma@spme.monash.edu.au (E.V. commercial steel were reheated at 900 and 1200 8C for
Pereloma). 30 min and then quenched in ice brine (cooling rate of
0921-5093/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 5 0 9 3 ( 0 2 ) 0 0 5 1 9 - 1
190 D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195
Table 1 10 /55 mm3) have been carried out at /20 8C. Data
Steel composition (weight%) from tensile and impact tests represent the mean of three
C Mn Si P S Al N2
specimens.
0.16 1.39 0.39 0.016 0.009 0.039 0.0042
3. Results and discussion
3.1. Ice brine quenching of non-deformed samples
The microstructures formed after quenching from 900
and 1200 8C are shown in Fig. 2. The average prior
austenite grain size was measured to be 120 mm in
samples quenched from 1200 8C (Fig. 2a) and 19 mm
(Fig. 2b) in samples quenched from 900 8C. A fully
Fig. 1. Processing schedule of the steel investigated. AC1 /720 8C, martensitic microstructure was formed in the specimens
AC3 /834 8C.
quenched from both austenitising temperatures. The
blocks of martensite, distinctly evident due to the
:/300 8C s 1). After that, all specimens were reheated different etching contrast, consist of fine martensite
at 740 8C for 30 min and submitted to warm rolling at laths parallel to each other. The prior austenite grains
700 8C, with three equal reductions of 22.4% true strain contain several martensitic blocks with different orien-
each pass. The final thickness of the specimens was 8.3 tations of laths. The width of the martensite blocks is
mm. After the first and second rolling passes, the not uniform. In samples quenched from 1200 8C, some
samples were returned to a furnace at 740 8C and after blocks are 100 mm thick, although others are wider than
the last rolling pass the specimens were air cooled. All 200 mm. In samples quenched from 900 8C, the
rolled specimens were annealed at 800 8C for 1, 60, 120 martensite structure is more refined due to the finer
and 180 min. After annealing, some of the specimens prior austenite grain size.
were air cooled, while the others were quenched.
Transverse sections of the heat treated and rolled 3.2. Warmed rolled and annealed microstructures
specimens were examined by optical, scanning and
transmission electron microscopy (TEM). Samples The effects of the deformation on the final micro-
were prepared following the standard procedures [13]. structure of the processed samples can be seen in Fig.
For TEM studies, thin foils of 3 mm diameter were 3(a /c). With increasing amount of deformation the final
punched out from /0.2 mm thick slices, mechanically microstructure becomes finer. The microstructures after
thinned to 0.12 mm and electropolished in a solution of single or double passes, and even in some localized parts
methanol with 5% perchloric acid at /30 8C in a of the samples after three pass deformation, are very
Struers Tenupol double jet unit, set at 60 V. Thin foils heterogeneous. This may be explained by the coarse
were examined using transmission electron microscope prior austenite grain structure developed during auste-
Philips CM 20 operated at 200 kV. To determine the nitising at 1200 8C and by incomplete recrystallisation
prior austenite grain size after austenitising at 900 and of ferrite due to the preferential recrystallisation at some
1200 8C, the quenched specimens were polished and regions, leading to a different grain size.
etched with picric acid [4]. Grain size was measured The evolution of the microstructure with annealing
using the linear average intercept method [13]. The final time after three pass deformation could be seen for
microstructure of the specimens was revealed by 2% specimens austenitised at 900 8C in Fig. 4 and for
Nital and Le Pera etching [13]. Ferrite grain size and specimens austenitised at 1200 8C in Fig. 5. At short
volume fractions of phases present in the final micro- annealing time (1 min), Fig. 4(a, b) and Fig. 5(a) the
structure were determined using the IMAGE PRO- ferrite grain size is very small (close to 1 mm) but the
PLUSTM image analyser software and following the microstructure is heterogeneous. Simultaneously with
ASTM standards [14,15]. All results were determinated fine ferrite grains the regions of twice as coarse ferrite
according to mean, standard deviation and relative error grains (/7 mm) were also observed. The annealing time-
values for 95% confidence range. Vickers microhardness temperature and total deformation dependence for
was measured using a 4.903 N load and result represents ferrite recrystallisation have been already established
the mean of 10 impressions from the analysed section. in some works for cold and hot rolling [16 /18]. In the
Tensile tests were performed on an Instron 4400 testing present work, the total deformation and annealing
machine using a crosshead speed of 0.033 mm s 1 and temperature were fixed. According to [16 /18], the time
samples with 4 mm diameter and 25 mm gauge length. for complete recrystallisation is longer than 1 s. For
Impact Charpy tests of notched sub-size specimens (5 / instance, for recrystallisation of C /Mn steel cold rolled
D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195 191
Fig. 2. (a) Optical micrographs of samples quenched from 1200 8C and (b) 900 8C and etched with 2% Nital.
Fig. 4. Optical (a,c,e) and SEM (b,d,f) micrographs of the samples
quenched from 900 8C, then warm rolled at 700 8C and annealed at
Fig. 3. Effect of the deformation on the final microstructure of the 800 8C for (a, b) 1min, (c,d) 60 min and (e,f) 180 min. LePera etching
annealed at 800 8C for 1 min samples. a, single pass; b, double pass; c, in (a,c,e).
three passes. Reheating temperature is 900 8C.
A more homogeneous and refined ferrite microstruc-
ture was produced in samples after austenitising at
with 50% reduction 30 s hold at intercritical annealing 900 8C comparing to the ones austenitised at 1200 8C.
temperature of 760 8C is required. However, the With increasing annealing time, the average ferrite grain
formation of austenite during this intercritical annealing size increases by /50% in the specimens austenitised at
continues until the completion of ferrite recrystallisation 1200 8C and by only /10% in the samples after
[16]. Therefore, 1 min at 800 8C was not enough to austenitising at 900 8C (Fig. 6). The corresponding
recrystallise the whole sample. The microstructures ferrite grain size in samples after austenitising at 900 8C
formed after longer annealing times were more uniform is also much finer for each annealing temperature than
Fig. 4(c /f) and Fig. 5(b, c). for the specimens austenitised at 1200 8C. These
192 D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195
Fig. 7. Effect of the annealing time on the volume fraction of phases
present in the microstructure. Samples were austenitised at 1200 8C,
quenched, warm rolled with three passes, annealed at 800 8C and air
cooled.
1200 8C. Fig. 4(a, c and e) show the changes in volume
fraction (from 1 to 8%) of MA constituent (white islands
in microstructure) with annealing time in the samples
austenitised at 900 8C. This may be due to the increase
in the volume fraction of granular bainite formed during
the cooling after intercritical annealing.
In Fig. 4(b, d) and Fig. 5(a, b), it is also possible to
identify the carbides as isolated islands precipitated at
proeutectoid ferrite grain boundaries. No pearlite for-
Fig. 5. SEM micrographs of the samples quenched from 1200 8C,
mation has been observed using SEM after 1 min
then warm rolled at 700 8C and annealed at 800 8C for (a) 1min, (b) annealing (Fig. 4b and Fig. 5a), while pearlite could be
90 min and (c) 180 min. seen in the microstructures after 180 min annealing (Fig.
4f and Fig. 5c).
During the cooling after intercritical annealing, aus-
tenite transforms to a mixture of ferrite and carbide,
Fe3C, or pearlite. As the time of annealing increases
austenite became more homogeneous and transforms on
cooling into more defined pearlite (Fig. 4f).
The evolution of final microstructure with annealing
time is clearly seen from TEM micrographs (Fig. 8).
Typically, two types of ferrite grains were observed:
polygonal ferrite grains which were formed on cooling
from the annealing temperature and ferrite grains
formed as a result of recrystallisation of deformed a
grains. Some ferrite grains with cell dislocation structure
Fig. 6. Effect of the annealing time on the ferrite grain size. in which recovery took place are also present. In samples
after 1 and 60 min annealing some elongated grains with
features are due to the finer average prior austenite grain high dislocation density were observed (Fig. 8a and b).
size at 900 8C than at 1200 8C leading to the finer It means that these annealing times were not sufficient
martensitic structure. Deformation during warm rolling for them to recover or recrystallise. Islands and blocks
introduces in-grain shear bands and by this also of retained austenite and martensite were observed in
increases the number of ferrite nucleation sites [16]. the final microstructure (Fig. 8b and c). The minor
The changes in the volume fractions of phases present phases present were granular bainite and pearlite.
are given in Fig. 7. With increasing annealing time, in air The austenite conditioning parameter Sv is generally
cooled after annealing samples ferrite volume fraction used to quantify ferrite nucleation sites. It represents the
slightly decreases, while the pearlite and bainite volume effective austenite interfacial area available for nuclea-
fractions increase. The volume fraction of martensite/ tion of ferrite grains. Increase in Sv causes ferrite grain
austenite (MA) constituent does not change significantly refinement [19 /22]. During warm rolling there is a
with annealing time for the specimens austenitised at mixture of a and g grains. As a result, a crystals deforms
D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195 193
Fig. 8. Thin foil TEM micrographs of the samples austenitised at 900 8C and annealed at 800 8C for (a) 1 min, (b,c) 60 min, (d) 120 min and (e,f)
180 min.
and subgrains form [3] while microshear bands are The microstructure after 180 min annealing is homo-
introduced into austenite [7]. These shear bands are geneous but with much coarser ferrite grain size. The
additional nucleation sites for ferrite formation during optimum combination of ferrite grain size and homo-
the cooling from the temperature of warm rolling. geneity of microstructure has been achieved after
During annealing, static recrystallisation of ferrite annealing for 60 min.
may take place leading to the formation of new ferrite To summarise, intercritical annealing leads to further
grains, otherwise ferrite transforms to austenite. Then Sv refinement of ferrite grains due to the (i) formation of
is significantly reduced as austenite now is not de- new recrystallised fine ferrite grains during annealing
formed. As a consequence, the nucleation of ferrite (extent of their growth is controlled by the annealing
during air cooling after annealing is restricted to the time) and (ii) phase transformation of austenite present
grain boundary area of austenite. at annealing temperature to fine ferrite crystals on
The lowering of the austenitising temperature from cooling.
1200 to 900 8C led to the decrease in prior austenite
grain size and subsequent increase in Sv. The refinement 3.3. Mechanical properties
of the microstructure might be further improved after
warm rolling by varying the cooling rate and time- 3.3.1. Microhardness
temperature parameters of annealing. As can be seen As was anticipated, increase in the number of rolling
from Fig. 4 and Fig. 5, short annealing times promote passes from 1 to 3 for the specimens austenitised at
the formation of the finest ferrite grain structure, albeit 1200 8C led to higher work hardening and higher
heterogeneous (significant variation in ferrite grain size). microhardness values, around 200 VHN. As the anneal-
194 D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195
ing time increased the values of microhardness de-
creased for samples austenitised at 1200 8C, although
for samples austenitised at 900 8C hardness is almost
the same for different times (Fig. 9). This is probably
due to the formation of a considerable amount of MA
constituent, around 8% for longer annealing times. The
hardness slightly oscillates between the various anneal-
ing times (Fig. 9), but with a tendency to decrease. This
is the result of competition between three processes
taking place: (i) the recovery and static recrystallisation
of ferrite, (ii) the grain growth of recrystallised ferrite
and (iii) the austenite formation and grain growth.
Fig. 10. Effect of the annealing time on the mechanical properties.
However, the total drop in microhardness from as-
rolled state to after 180 min of intercritically annealed
was not very large (/10 VHN). Longer annealing times steel, typically characterised by a yield strength of 356
result in coarser ferrite grain structure and subsequently MPa, a tensile strength of 538 MPa and 24% elongation.
lower microhardness. The microhardness of all samples This improvement in mechanical properties is due to the
austenitised at 900 8C, subjected to three pass warm reduction of ferrite grain size from 10 mm in industrial
rolling and intercritically annealed was significantly hot rolled steel to 1/1.5 mm grain size achieved in this
higher than the ones for the corresponding samples warm rolled and intercritically annealed steel. It is also
austenitised at 1200 8C. Comparing the homogenety of suggested that the variation in volume fractions of
the microstructure, ferrite grain size and microhardness, secondary phases (pearlite, bainite) does not affect the
it could be concluded that the best combination of these mechanical properties to the same degree as the varia-
features has been achieved in samples after 900 8C tion in ferrite grain size. Thus, ferrite grain size is the
austenitising, 60% total deformation and 60 min anneal- main factor controlling the mechanical properties in this
ing time at 800 8C. steel.
3.3.2. Stress-strain data 3.3.3. Charpy V-notch test
The tensile tests results are summarised in Fig. 10. All Charpy V-notch impact test results are shown in Fig.
samples subjected to austenitising at 900 8C have higher 11. For samples austenitised at 1200 8C the absorbed
values of yield and ultimate tensile strength for all energy is the highest after 60 min annealing time and
annealing conditions than samples austenitised at intermediate ferrite grain size of /1.15 mm. With
1200 8C. This could be explained from the point of increasing ferrite grain size at longer annealing times
view of ferrite grain size as has already been discussed. the absorbed energy decreases. This is due to the
There is no linear dependence of the elongation on formation of pearlite instead of a mixture of ferrite
annealing time, even though the higher elongation has and carbides after longer annealing times. On the other
been achieved in samples after 900 8C austenitising and hand, the samples austenitised at 900 8C have a higher
60 and 180 min annealing times, than in corresponding volume fraction of MA constituent, which impair the
samples austenitised at 1200 8C. These results are in absorbed energy [25]. Both the impact transition tem-
agreement with the literature [23,24]. In this work, the perature and absorbed energy are lower for ultra-fine
increase in strength and elongation values by /20% has
been achieved in comparison with industrially hot rolled
Fig. 11. Charpy test results for all warm rolled (three passes) and
Fig. 9. Effect of the annealing time on microhardness. annealed specimens.
D.B. Santos et al. / Materials Science and Engineering A346 (2003) 189 /195 195
ferrite steel due to the formation of separation during [5] T. Hayashi, M. Saito, K. Tsuzaki, K. Nagai, Proceedings of the
the impact test [19,23]. International Conference Recrystallization’99, JIM, Tsukuba,
Japan, July 1999, p 333.
[6] R. Priestner, A.K. Ibraheem, Proceedings of 39th Annual
Conference of Metallurgists of CIM, CIM, Ottawa, Canada,
4. Conclusions August 20 /23, 2000 p. 351.
[7] S. Takaki, K. Kawasaki, Y. Kimura, Journal of Materials
The low carbon steel has been subjected to warm Proceeding Technology 117(3) (2001).
rolling followed by intercritical annealing. The final [8] Y. Adachi, S. Hinotani, Journal of Materials Proceeding Tech-
nology 117(3) (2001).
microstructure contained ultra-fine ferrite, granular [9] P.D. Hodgson, M.R. Hickson, R.K. Gibbs, in: J.M. Rodriguez, I.
bainite with martensite/retained austenite constituent Gutierrez, B. Lopez (Eds.), Materials Science Forum, vol. 284 /
and small amount of pearlite. The results have shown 286, Transaction and Technology Publication Ltd, Spain, 1998, p.
that processing schedules utilised in this work led to 63.
significant refinement of ferrite grain structure and [10] P.D. Hodgson, M.R. Hickson, R.K. Gibbs, Scripta. Mater. 40
corresponding improvement in mechanical properties. (1999) 1179.
[11] A. Najafi-Zaden, J.J. Jonas, S. Yue, Mat. Sci. Forum 113 /115
The optimum combination of microhardness, strength
(1998) 441.
and ductility has been achieved in samples subjected to [12] T. Hayashi, CAMP-ISIJ 11 (1998) 1031.
900 8C austenitising, total reduction of 66.8% true [13] P.C.M. Rodrigues, E.V. Pereloma, D.B. Santos, Mater. Sci. Eng.
strain during warm rolling and 60 min annealing at A283 (2000) 136.
800 8C. The ferrite grain refinement led to an increase [14] ASTM E 112-88, Annual Book of ASTM Standards, Philadel-
by 20% in yield and tensile strength compared to the phia, PA, 1988, pp. 284 /309.
[15] ASTM E 562-83, Annual Book of ASTM Standards, Philadel-
industrially hot rolled steel. This was accompanied by a phia, PA, 1983, pp. 612 /617.
similar improvement in ductility. [16] D.Z. Yang, E.L. Brown, D.K. Matlok, G. Krauss, Metall Trans.
A16 (1985) 1385.
[17] A.C. Perry, S.W. Thompson, J.G. Speer, Iron Steel Maker 6
Acknowledgements (September 2000) 47.
[18] Y. Matsumura, H. Yada, Trans. ISIJ 27 (1987) 492.
[19] T. Tanaka, Int. Metals Rev. 4 (1981) 185.
The authors acknowledge to CNPq (Process number [20] I. Kozasu, in: I. Kozasu, C. Ouchi, T. Sampei, T. Okita, M.
463777/00-0-CTPETRO) for financial support to this Korchynsky (Eds.), Microalloying’75: Proceedings of an Interna-
work. tional Symposium on High-Strength, Low-Alloy steels, Union
Carbide Corporation, New York, USA, 1976, p. 100.
[21] C.M. Sellars, Proceeding of the International Conference on Hot
Working and Forming Processes, London Metals Society 1980,
References
p.3.
[22] H.J. McQueen, J.J. Jonas, Treatise Mater. Sci. Technol. 6 (1975)
[1] T. Ishikawa, H. Yoshikawa, T. Inoue, Y. Nomiyama, S. Imai,
393.
Nippon Steel Tech. Rep. No.75 (November 1997) 31.
[23] M. Niikura, Y. Hagiwara, K. Nagai, K. Tsuzaki, S. Takaki,
[2] A.K. Ibraheem, R. Priestner, J.R. Bowen, P.B. Prangnell, F.J.
Proceedings of the International Symposium on Ultrafine
Humphreys, Proceedings of the International Conference on
Grained Steels, The ISIJ, Fukuoka, Japan, September, 2001, p.
Thermomechanical Processing of Steels, vol. 2, IOM Commu-
nications Ltd, London, UK, 24 /26 May, 2000, p. 446. 26.
[3] D.N. Hawkins, A.A. Shuttleworth, J. Mech. Working Technol. 2 [24] W. Morrison, Trans. ASM 59 (1966) 825.
(1979) 333. [25] P.C.M. Rodrigues, E.V. Pereloma, D.B. Santos,
[4] D.B. Santos, R.A.N.M. Barbosa, Metalurgia */ABM 43 (354) Journal of Materials Proceeding Technology 117(3)
(1987) 279. (2001).
