Materials Science and Engineering A 527 (2010) 7600–7604
Full Pearlite Obtained by Slow Cooling in
Medium Carbon Steel
H. L. Yi a,b
a Graduate
Institute of Ferrous Technology, Pohang University of Science and
Technology, Pohang 790-784, Republic of Korea
b Technical
center, Laigang Steel, Laiwu 271104, Shandong, P.R. China
Abstract
Fully pearlitic steels are applied widely in engineering structures in the form of
strong cables. They conventionally contain 0.8 wt% of carbon and therefore can be
poor in ductility. One solution is medium carbon steel but which is fully pearlitic.
This can be achieved only by rapid cooling which is not convenient for comercial manufacture. In the present work an almost fully pearlitic microstructure was
desinged in a 0.4 wt% carbon containing steel, in the as-cast condition, by slow
cooling. The mechanism involves the suppression of allotriomorphic ferrite formation from austenite during casting because of the big austenite grain size. Fine
divorced-cementite lamellae formed in pearlite due to the huge driving force.
Key words: Pearlite, Medium carbon, Para-equilibrium, Lamellae, Cementie
1
Introduction
A pearlite colony is a bicrystal of cementite (θ)and ferrite (α), with these two
phases growing at a common front with the austenite (γ)[1, 2]. Fully pearlitic
microstructure can be obtained with a eutectoid composition in convention[3].
The fully pearlitic steel is strong and therefore is widely applied as engineering
alloys[4]. It is however not ductile due to the brittle cementie lamellaes. Reduce the size and fraction of cementite is a way to enhance the ductility[3, 5].
Austenite is possible to transform into full pearlite at the temperature lower
than the equilibrium eutectoid even in low carbon or hyper-eutectoid steel,
Email address: hityihl@postech.ac.kr, Tel.: +86 634 6822127; fax: +86
634 6822127; corresponding author (H. L. Yi).
Preprint submitted to Elsevier
7 October 2010
which is known as the Hultgren extrapolation region, identified by extending the (γ + θ/γ)or (α + γ/γ)phase boundaries to temperatures below the
eutectoid[3, 6–9]. This idea is one potential solution to acheive high strength
combined with ductiltiy by fully pearlitic microstructure in medium carbon
alloys. The ducitliy is resulted from the refined size and reduced fraction of
cemenite lamellaes. Very fast cooling is neccessary to avoid the hypo-ferrite
formation. The less the carbon is contained, the lower temperature of eutectoid
transformaiton is required and therefore higher cooling rate is needed[6, 9], for
example, a Fe-0.4C-0.62Mn-0.30Si-0.024Al wt% alloy can acheive full pearlite
between 670-560◦ C cooling from austenitizing temperature 900◦ C at very fast
rate between 70 ◦ C s−1 and 140 ◦ C s−1 [9]. Such high cooling rate may not be
easy in practice when dealing with large dimensions of steels.
Para-equilibrium solidification was found in a high aluminium alloy during
solidification, called δ–TRIP[10–13]. The pearlite transformation is however
a local equilibrium process. The transition of state from para-equilibrium to
equilibrium may give huge driving force for cementite precipitating and it is
therefore possible to acheive fine cementite. A 0.4C wt% high aluminium–
bearing steel was designed to form full pearlite in this research. The science
behind this concept is discussed in the point view of the huge driving force
for pearlitic transformation due to transition from the non-equilibrium state
to equilibrium state.
2
Experimental Method
Two alloys were designed with different aluminium concentrations (Table 1).
The alloys were manufactured as 34 kg ingots of 100×170×230 mm dimensions
using a vacuum furnace. The ingot was reheated to 1200◦ C for rough rolling to
make 25–30 mm slabs followed by air cooling. These slabs were then reheated
to 1200◦ C and hot–rolled to 3 mm in thickness. Optical microscopy samples
were prepared using standard methods and etched in 2% nital. Then the phase
volume fraction was estimated from the optical microscopy by image analysis. Higher resolution observations were done using a field-emission scanning
electron microscope operating at 10 kV accelerating voltage. Microhardness of
pearlite in each alloy was measured on a Vickers hardness (FM–700) tester by
using 300 gf loading and 15 s holding time. Room temperature tensile testing
for the as-cast alloys was carried out using a computer controlled Zwick/Roell
testing machine (Z-100) with an extensometer. The test specimens were prepared as per ASTM Standard (ASTM: Vol. 01.02: E8M–00). Uniaxial tensile
tests of samples of 10 mm diameter and 50 mm gauge length was carried out
at a strain rate of 3.3 × 10−3 s−1 .
2
3
Metallography and properties
The microscopic features of two alloys designed on the basis of equilibrium to
contain substantial amounts of δ–ferrite, have been examined and found to display zero or much reduced fractions of this phase in the solidified condition[10–
13]. It is concluded that this is because the austenite that forms during cooling by solid–state transformation, does so without the required partitioning of
substitutional solutes. This is responsible for the diminished quantities of δ–
ferrite found in the cast microstructures[11]. The alloy1 contains only 9±1wt%
of allotriomorphic ferrite and the remanent pearlite, where the dendritic solidification is revealed through coring effects that shows the original dendritic
thickness is 89±9µm(Fig. 1a,b). That indicates the alloy1 was fully austenized
during the cooling process then followed by γ → α + θ transformation. The
allotriomorphic ferrite was transformed from austenite in the para equilibrium
condition and was confirmed by the microanalysis and the thermodynamics
data. The partition coefficient of alloying element is defined as the ration between concentration in ferrite and in austenite before pearlite formation. The
composition in austenite can be measured from the pearlite region because it
in the final microstructure inherits the chemical composition of parent austenite. The partition coefficients of both manganese and aluminium are close to
1 which is far from the equilibrium value 0.32 and 3.00 listed in Table 2. It
means the alloy elements don’t partition during the allotriomorphic ferrite.
The ferritic transformation before the cementite beginning was suppressed
owing to the big original austenite grain size. This can be confirmed by the
fact that approximately 50 ± 3wt% allotriomorphic ferrite was found in the
hot rolled steel due to the grain refinement for the same alloy (Fig. 1c), where
the original austenite is as small as only 3.1±0.2µm estimated by using the
mean lineal intercept method[14]. The low cementite fraction due to the low
carbon concentration causes it to become discontinuous in the pearlite.
It is emphasized again, that the para equilibrium calculation is a limiting
estimate since the real process can be anywhere between local equilibrium and
local para equilibrium. The interpretation presented here explains subsequent
solid–state transformation under conditions where partitioning does not follow
expectations from local equilibrium at the advancing interfaces. In conclusion,
the allotriomorphic ferrite was suppressed in the cast condition but grow fast
during the hot-rolling due to kinetics reason resulted from large austenite grain
size during solidification.
In Alloy 2, 33 ± 2 vol% of dendritic δ–ferrite was retained to the ambient
temperature and the remanent phase is pearlite (Fig. 2). The pearlite in Alloy
2 contains both the perfect pearlite and discontinuous fine cementite dispersed
in the pearlitic ferrite matrix.
3
Supposing ferrite contains 0.02 wt% carbon in the as–casted microstructure,
the overall carbon concentration in pearlite can be calculated based on the
carbon mass balance, illustrated in Table 3. The microhardness of Alloy1 containing 0.39 wt% carbon in the pearlite region was measured to be 266±20 HV
compared with 253 ± 16 HV in pearlite region of Alloy2 which even contains
more carbon, 0.54 wt%, there. Alloy1 performed mechanically better than Alloy2 did in cast condition(Fig. 3). Alloy1 has similar yield strength and work
hardening rate but slightly better ductility and resultant better ultimate tensile strength compared with alloy2. The ductility in alloy1 was enhanced by
the fine and dispersed cementite in pearlite.
4
Kenitics
The big original austenite grain size limits the growth of allotriomorphic ferrite is approved by the simulation of DICTRA combined with thermodynamics database TCFE (version 1.21) and mobility database MOB2, capable of
dealing with diffusional growth in multicomponent systems given the availability of thermodynamic and atomic–mobility data[15–18]. The scale of the
microstructure selected for simulation is half of the approximate distance between δ–ferrite dendrites, i.e., 50 µm because the ferrite grows from both
direction of austenite. Since the cooling rate of the 34 kg ingots used is not
known, we have assumed a slow rate of −20Ks−1 , which also is representative of many continuous casting operations. During hot rolling, the alloy was
heated to 1473K and followed by rolling process what was finished at 1173
K. Both in cast process and hot rolling process, the alloy can be supposed
to be full austenite at 1200K followed by allotriomorphic ferrite formation.
The composition of austenite used was therefore just the alloy composition
and the computation was initiated with a thin (0.001 µm) layer of ferrite of
the same composition as the austenite. The simulation was based on paraequilibrium condition and fulfilled between 1200 K and 1000 K. The results
(Fig. 4) are fascinating in that they prove that it is quite feasible for discrapency of the amount of allotriomorphic ferrite between as–cast alloy and
hot rolled one, where 9.24 and 65.17 wt% of allotriomorphic ferrite can form
at 1000 K cooled from 1200 K at the rate of −20Ks−1 are well consistent with
the experimental value 9 ± 1 and 50 ± 3 wt%. It is emphasized again, that
the paraequilibrium calculation is a limiting estimate since the real process
can be anywhere between local equilibrium and local paraequilibrium. The
interpretation presented here explains subsequent solid–state transformation
under conditions where partitioning does not follow expectations from local
equilibrium at the advancing interfaces. In conclusion, the allotriomorphic ferrite was suppressed in the cast alloy but not in the hot rolling process due
to the big austenite grain size in cast condition and relevant kinetic reason
4
proven by microanalysis and simulation by DICTRA.
5
Transformation driving force
The solid transformation is of para equilibrium approved in the previous
work[11]. It is however not possible to include cementite in para equilibrium
calculation using MTDATA with TCFE1.21 database due to a lack of thermodynamic data. Based on the metallography, the pearlitic transformation in
alloy 1 in cast condition is considered to start from full austenite in the case
ignoring the small amount of allotriomorphic ferrite, where the mother phases
are in non-equilibrium state and the products are in equilibrium state, where
the allotriomorphic ferrite was suppressed due to the large original austenite
grain size and the kinetic reason. The equilibrium phase diagrams of both
alloys including or excluding cementite phase are shown in Fig. 5.
The driving force for pearlitic transformation in alloy1 at 1000 K can therefore
be calculated as ∆Gγ→α+θ . In alloy 2, δ–ferrite persisted during the whole cast
process, the pearlite (α + θ) at 1000K therefore transformed from a mixture of
α and γ, where both mother phases and products are in equilibrium state. The
Gibbs free energy is obtained from results of MTDATA[19]. For 100 kg alloys,
the total driving force for pearlitic transformation of Alloy 1 is almost 3 times
as that of Alloy 2, illustrated in Table 4. The huge driving force for pearlitic
transformation is coming from the strong trend of system state transition from
non-equilibrium state to equilibrium state during pearlite transformation in
Alloy1. As a conclusion, the ultra fine cementite lamellae or particle can form
in Alloy 1 due to very high nucleation rate resulted from the huge chemical
driving force due to state transition from non–equilibrium to equilibrium resulting from the suppressed allotriomorphic ferrite formation, which may play
a significant role for mechanical properties.
6
Conclusions
Almost fully pearlitic microstructure was obtained in a cast 0.4 wt% carbon
containing steel with the approximate cooling rate −20Ks−1 , where only 9 ± 1
wt% of allotriomorphic ferrite formed compared with 50 ± 3 wt% of that was
formed from during hot rolling process. The allotriomorphic ferrite transformation in para equilibrium condition during casting was suppressed by the
kinetics reason because of the big austenite grain size proven by microanalysis. The suppressed ferritic transformation leads to a huge driving force for
the pearlite formation confirmed by thermodynamics calculation, which encourages high nucleation rate and fulfils the fine divorced cementite lamellae.
5
Acknowledgments
The author is grateful to Professor H.K.D.H. Bhadeshia for useful discussions,
to Professor Hae–Geon Lee for laboratory facilities at GIFT, and to POSCO
for the Steel Innovation Programme. Support from the World Class University Programme of the National Research Foundation of Korea, Ministry of
Education, Science and Technology, project number R32–2008–000–10147–0
is gratefully acknowledged.
References
[1] N. Ridley, A Review of the Data on the Interlamellar Spacing of Pearlite,
Metallurgical and Materials Transactions A 15 (6) (1984) 1019–1036.
[2] P. Howell, The Pearlite Reaction in Steels Mechanisms and Crystallography* 1:: Part I. From HC Sorby to RF Mehl, Materials Characterization
40 (4-5) (1998) 227–260.
[3] H. K. D. H. Bhadeshia, R. W. K. Honeycombe, Steels: Microstructure
and Properties, 3rd edition, Butterworth–Heinemann, London, 2006.
[4] W. Nam, C. Bae, S. Oh, S. Kwon, Effect of interlamellar spacing on
cementite dissolution during wire drawing of pearlitic steel wires, Scripta
materialia 42 (5) (2000) 457–463.
[5] H. K. D. H. Bhadeshia, Alternatives to the ferrite-pearlite microstructures, Materials Science Forum 284–286 (1998) 39–50.
[6] A. Marder, B. Bramfitt, The effect of morphology on the strength of
pearlite, Metallurgical and Materials Transactions A 7 (2) (1976) 365–
372.
[7] M. Gensamer, E. Pearsall, W. Pellini, J. Low Jr, The tensile properties of
pearlite, bainite, and spheroidite, TRANSACTIONS, American Society
for Metals 30 (1942) 983.
[8] A. Hultgren, et al., Isothermal transformation of austenite, Trans. ASM
39 (1947) 915–1005.
[9] J. P. Houin, A. Simon, G. Beck, Relationship between structure and mechanical properties of pearlite between 0.2% and 0.8%C, Trans. ISIJ 21
(1981) 726–731.
[10] S. Chatterjee, M. Murugananth, H. K. D. H. Bhadeshia, δ–TRIP steel,
Materials Science and Technology 23 (2007) 819–827.
[11] H. L. Yi, S. K. Ghosh, W. J. Liu, K. Y. Lee, H. K. D. H. Bhadeshia,
Nonequilibrium solidification and ferrite in δ–TRIP steel, Materials Science and Technology 26 (2010) 817–823.
[12] H. L. Yi, K. Y. Lee, H. K. D. H. Bhadeshia, Stabilisation of ferrite in
hot-rolled in δ–TRIP steel, Materials Science and Technology 26 (2010)
in press.
6
[13] H. L. Yi, K. Y. Lee, H. K. D. H. Bhadeshia, Extraordinary ductility in
al-bearing δ–TRIP steel, Proceedings of the Royal Society of London A
466 (2010) in press.
[14] R. T. DeHoff, F. N. Rhines, Quantitative Microscopy, McGraw Hill, New
York, 1968.
[15] J. Ågren, Local equilibrium and prediction of diffusional transformations,
Scandinavian Journal of Metallurgy 20 (1991) 86–92.
[16] J.Ågren, Computer simulations of diffusional reactions in complex steels,
ISIJ International 32 (1992) 291–296.
[17] A. Engström, L. Höglund, J. Ågren, Computer simulation of diffusion in
multiphase systems, Metallurgical and Materials Transactions A 25 (6)
(1994) 1127–1134.
[18] A. Borgenstam, A. Engströn, L. Höglund, J. Ågren, DICTRA, a tool
for simulation of diffusional transformations in alloys, Journal of Phase
Equilibria 21 (2000) 269–280.
[19] NPL, MTDATA, Software, National Physical Laboratory, Teddington,
U.K. (2006).
7
Table 1
The design compositions and those actually achieved during manufacture, wt%
Alloy1
Alloy2
Design
Actual
Design
Actual
C
0.4
0.36
0.4
0.37
Si
0.25
0.26
0.25
0.23
Mn
2.0
2.02
2.0
1.99
Al
2.1
2.13
2.5
2.49
Cu
0.5
0.49
0.5
0.49
P
0.02
0.02
0.02
0.02
S
0.0036
0.0036
N
0.0048
0.0048
Table 2
Partition of manganese and aluminium by equilibrium calculation at 1000 K and
microanalysis results
Ferrite
’Austenite’
Partition coefficient
Mn
Al
Mn
Al
Mn
Al
Equilibrium at 1000 K
1.18
2.73
3.74
0.91
0.32
3.00
Measured
1.85±0.21
2.62±0.11
1.82±0.18
2.59±0.11
1.01
1.02
8
Table 3
Microhardness and carbon content in pearlite and pearlite fraction in each alloy,
C
where the VP earlite means volume percentage of pearlite, CAlloy
and CPCearlite present
carbon concentration in alloy and pearlite respectively
VP earlite / vol%
C
CAlloy
/ wt %
CPCearlite / wt %
Hardness in pearlite / HV
Alloy1
90.7
0.36
0.39
266±20
Alloy2
67.1
0.37
0.54
253±16
Table 4
Driving force for pearlitic transformation for the 100 kg alloy at 1000 K
GM other / kJ
GP roduct / kJ
Alloy1
-89244.8 (γ)
-89709.8 (α + θ)
-4.65(γ → α + θ)
Alloy2
-91019.3 (α + γ)
-91192.8 (α + θ)
-1.74( α + γ → α + θ
9
GM other
toP roduct
(a)
(b)
(c)
Fig. 1. Microstructure of the cast Alloy1; the dark regions are fine pearlite and the
white, ferrite. (a) General microstructure; (b) higher magnification image of the
dark areas showing pearlite. (c) hot rolled indicating fine original austenite grain
size.
10
(a)
(b)
Fig. 2. Microstructure of the cast Alloy2; the dark regions are fine pearlite and the
white, ferrite. (a) General microstructure; (b) higher magnification image showing
pearlite and ferrite.
11
(a)
(b)
(c)
Fig. 3. Tensile properties of as-cast alloys. (a) Flow stress of Alloy 1; (b) Flow stress
of Alloy2; (c) Comparison of mechanical properties between Alloy1 and Alloy2.
12
(a)
(b)
Fig. 4. The solid-phase transformation in (a) alloy 1 and (b) alloy2 simulated by
DICTRA from 1200K to 1000K at cooling rate of 20K s−1 .
13
(a)
(b)
(c)
Fig. 5. Calculated phase amount as a function of temperatures and alloy based on
equilibrium. (a) Alloy 1 including cementite, (b) Alloy 2 excluding cementite, (c)
Alloy2 including cementite.
14