Paravicinibagliani 2013
Paravicinibagliani 2013
Paravicinibagliani 2013
a r t i c l e i n f o a b s t r a c t
Article history: The effect of different microstructures on the tensile and toughness properties of a low alloy medium
Received 19 March 2012 carbon steel (0.28C–1.4Si–0.67Mn–1.49Cr–0.56Mo wt%) was investigated, comparing the properties
Received in revised form obtained after the application of selected quenching and partitioning (Q&P) and quenching and
18 July 2012
tempering (Q&T) treatments. After Q&T the strength–toughness combination was the lowest, whereas
Accepted 25 August 2012
the best combination was achieved by Q&P, as a result of the carbon depletion of the martensite and the
Available online 1 September 2012
high stabilization of the austenite. Nonetheless, the presence of islands of martensite/austenite (MA)
Keywords: constituents after Q&P treatments prevented the achievement of toughness levels comparable to the
Quenching ones currently obtainable with other steels and heat treatments.
Partitioning
& 2012 Elsevier B.V. All rights reserved.
Tempering
Retained austenite
Bainite
1. Introduction For these applications, when a yield strength (YS) greater than
1000 MPa is required, the current state-of-the-art commercially
The quenching and partitioning (Q&P) process is a recently available consists mainly of medium carbon quenched and tempered
proposed heat treatment to produce high strength steels with good (Q&T) steels, alloyed with elements such as chromium, molybdenum
ductility, improving the mechanical properties that can be obtained and nickel. Grades with a minimum YS of 1140 MPa and good
with other consolidated products based on transformation-induced- toughness at 20 1C are available [5]. Medium-carbon carbide-free
plasticity (TRIP), dual phase or martensitic steels [1]. Q&P consists of bainitic steels also lead to yield strengths above 1000 MPa with 50%
a partial or full austenitization, followed by an interrupted quench at fracture appearance transition temperature (FATT) of 20 1C or less
a quenching temperature (QT) between the martensite-start tem- [6,7], although these steels in some cases require quite expensive
perature (Ms) and the martensite-finish temperature (Mf) to obtain a nickel alloying. In the case of carbide-free bainitic steels, the
predetermined combination of martensite and austenite. A partition- toughness improvement is believed to be a consequence of the
ing stage at the same or at higher temperature (PT) allows carbon to suppression of the cementite precipitation by alloying the steel with
escape from the supersaturated martensitic phase into the untrans- 1.5 wt% silicon, which results in the retention of interlath films of
formed austenite in the absence of carbide precipitation. Thus, retained austenite in a matrix of upper bainite [8].
austenite is stabilized and retained in the final cooling to room Steels resulting from the application of adequate Q&P routes are
temperature [2,3].This cycle aims at producing a microstructure of formed by microstructures containing carbon-depleted martensite
carbon-depleted martensite and retained austenite, whose properties and retained austenite, which have some morphological similarities
offer advantages when high strength and good formability are with the microstructures present in carbide-free bainitic steels.
required, like in thin sheets for the automotive industry, as they However, little attention has been devoted to the combination of
allow improvements in crashworthiness and weight reduction [4]. yield strength and toughness achievable with this new class of
The design of products for structural, engineering, pressure or oil microstructures. Hong et al. [9] showed the possibility of reaching
and gas applications is based on the optimization of the strength and higher yield strength and impact toughness by Q&P routes in
toughness combinations at the minimum operating temperature. comparison with other traditional treatments, by studying the
These are decisive factors to prevent sudden and brittle failure and, behavior of a low carbon steel treated to reach the properties of
in pressure vessels, to promote a ‘‘leak before break’’ behavior. an X100 grade. Nonetheless, most of the research in Q&P steels has
been focused on strength and elongation analyses [1,3,10,11].
The object of the present work is to investigate the combina-
n
Corresponding author. Tel.: þ39 0355602831; fax: þ39 0355602845. tions of strength and toughness reachable in a medium-carbon
E-mail address: epbagliani@tenaris.com (E. Paravicini Bagliani). steel alloyed with 1.5 wt% silicon, after the application of specially
0921-5093/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.msea.2012.08.130
E. Paravicini Bagliani et al. / Materials Science & Engineering A 559 (2013) 486–495 487
Table 2
Critical temperatures (1C), obtained by because these phenomena would overlap and compete with the
dilatometry (average of three tests). process of carbon partitioning. This requires the selection of
appropriate partitioning temperatures at which new phases such
Ac1 Ac3 Ms
as bainite do not form. With this aim, the possible formation of
771 75 896 710 358 74 phases during isothermal treatments after full austenitization
above the Ms-temperature was experimentally investigated.
The TTT diagram, determined starting from a fully austenitic
Theoretical calculations were also performed to estimate the structure was carried out to identify possible ranges of tempera-
partitioning time that ensures full carbon partitioning from the ture where the austenite does not transform into bainite or ferrite
martensite to the austenite at different partitioning temperatures. that could be exploited to avoid phase transformations during
The model presented by Santofimia et al. [17] was applied to a partitioning. Results of such investigation can be used as a guide;
Fe-0.28 1C system quenched at a temperature at which the however it is important to note that these results may slightly
volume fraction of martensite formed is 0.77. A stationary differ from the case of isothermal transformations at the same
martensite–austenite interface and a total width of martensite temperatures after partial quench below Ms, since the presence of
lath plus austenite film of 0.26 mm were assumed. This value was martensite may affect nucleation, grain size and internal stresses
based on the hypothesis of a typical width of the martensite laths of the new forming phase [17–19].
of 0.20 mm, constant during the martensitic transformation, and The isothermal treatments were carried out by the dilatometer
of a width of the austenite films of 0.06 mm, dependent on the at temperatures between 370 1C and 500 1C and the results are
austenite fraction present at QT [17]. Three partitioning tempera- shown in Fig. 3a. The ‘‘finish’’ curve indicates the end of the phase
tures, later selected in the experimental study, were considered in transformation according to dilatometry, but this does not imply
the calculations—420 1C, 470 1C and 500 1C. The minimum parti- the complete transformation of the austenite [20]. Above 470 1C,
tioning time needed for full carbon partitioning from martensite and at least up to 500 1C, the steel is characterized by a range of
to austenite in such conditions is 3 s, 0.4 s and 0.2 s, respectively. temperatures, close to the Bs temperature, where the austenite is
stable for very long times, whereas an increasing amount of
3.2. Experimental study of the steel bainite is formed when the isothermal temperature is decreased
towards Ms. An example of microstructure, obtained after iso-
On the basis of the theoretical analyses, a preliminary dilato- thermal treatment at 400 1C, is shown in Fig. 3b. It is mainly
metry study of the steel was carried out to identify the conditions composed of fine plates of bainite with films of retained austenite,
that stabilize the retained austenite in Q&P treatments. although coarse blocks of high-carbon martensite plus retained
The critical temperatures Ac1 and Ac3, listed in Table 2, were austenite (MA-constituent) can also be observed. This constituent
determined by heating a cylindrical specimen of 10 mm length was also present after isothermal treatments at lower tempera-
and 4 mm diameter, according to ASTM A 1033, at 28 1C/h above tures and it is related to an incomplete transformation of bainite
650 1C. On the basis of this result, an austenitizing condition controlled by the T0 0 line, at which the bainite and ferrite have the
above Ac3 of 940 1C for 10 min was chosen for the design of Q&P same free energy [20].
and Q&T treatments.
Continuous-cooling tests were carried out after full austenitization 3.3. Design of Q&P treatments
to experimentally determine the Ms-temperature and the critical
cooling rate that gives 100% martensite. As shown in Fig. 2, the The Q&P routes designed in this work followed the sequence
dilatometric curve at 2 1C/s did not reveal traces of phase transforma- displayed in Fig. 4. The considered criteria for the design of heat
tions above Ms, whereas the curve at 1 1C/s displayed the formation treatments are listed below:
of bainite. The rate of 2 1C/s was identified as the minimum rate
to have a fully martensitic structure, as was previously predicted (a) Complete austenitization above the Ac3 temperature;
from the calculations with JMatPro. The experimental value of the (b) cooling rate fast enough to avoid any phase transformations
Ms-temperature is given in Table 2. before martensite formation;
After the interrupted quench at QT the untransformed austenite (c) interrupted quench at the temperature QT between Ms and Mf;
must undergo the partitioning stage. During this step, the decom- (d) short holding time at QT but long enough to ensure the thermal
position of the residual austenite into new phases is undesired, homogenization of the specimen;
E. Paravicini Bagliani et al. / Materials Science & Engineering A 559 (2013) 486–495 489
Table 3
Details of the Q&P cycles carried out by the dilatometer.
Table 4
Austenitization, quenching and partitioning temperatures (TAust, QT and PT,
respectively) and austenitization and partitioning times (tAust, Pt) applied with
a salt bath setup. Pt ¼tempering time for QT500 and QT550.
e)
3.3.2. Salt baths
g)
c) QT The Q&P routes for salt bath experiments were chosen on the
d)
basis of the results obtained with the dilatometer. The heat
treatments were carried out with slight differences between real
and target temperatures. The indicative values of austenitization,
quenching and partitioning temperatures recorded with a ther-
time
mocouple and the corresponding times are displayed in Table 4.
Fig. 4. Scheme of the Q&P cycles. The quenching temperatures were between 240 1C and 315 1C in
order to obtain different fractions of austenite and martensite.
The partitioning stage was carried out reheating to 500 1C
(e) reheating of the specimen to the partitioning temperature PT, (above Bs), cooling in air down to approximately 470 1C and
at a reasonably fast rate to minimize the concurrence of quenching in water. In three cases the piece was immediately
processes, but not too far from industrially feasible rates; quenched in water after reheating in the second bath. One Q&P
(f) adequate partitioning temperatures and times (Pt), in which treatment was applied with quenching to approximately 290 1C
the formation of bainite is avoided and the homogenization of and partitioning at a temperature of about 420 1C corresponding
carbon within the austenite is complete; and to bainite formation after full austenitization. This specimen
(g) final quenching to room temperature. was identified as QP-B. In addition, two Q&T treatments with
tempering at 500 1C and 550 1C were applied.
In the following, the selected Q&P treatments for dilatometry
and salt bath experiments are described. 4. Results and discussion
between 10 s and 600 s. The theoretical values of Fig. 1 are also The effect of partitioning time at 470 1C is shown by the tests
reported. The effect of QT can be observed in the tests carried out with holding times of 10 s, 60 s and 600 s after quench to 300 1C.
with partitioning at 470 1C for 60 s. Considerable amounts of After 10 s, the process of carbon partitioning was not completed
retained austenite were obtained (0.12–0.18) stopping the as the maximum content of austenite was not reached. Note that
quench between 260 1C and 325 1C. The maximum of 0.18 was the remarkable stability of the process was up to 600 s. The
reached between 285 1C and 300 1C. The maximum retained predicted partitioning times at 470 1C (Section 3.1) were less than
austenite and the optimum temperature are quite in agreement 1 s, in the assumption of martensite lath plus austenite film of
with the predictions, whereas at higher QT the experimental 0.26 mm. This would indicate that longer diffusion paths of carbon
retained austenite values are higher than the predicted ones. must be taken into account to achieve a homogeneous stabiliza-
Differences between experiments and calculations can be caused tion of the austenite.
by the fact that the applied model does not consider the kinetics After partitioning at 500 1C for 30 s, the volume fractions of
of carbon partitioning [17]. retained austenite were slightly lower than those obtained at
During partitioning, the dilatometric curves do not show 470 1C for 60 s with the same quenching temperatures; moreover,
evidence of phase transformations, which would have been a considerable decrease of the retained austenite content was
revealed by an expansion of the specimen. This is in agreement observed for longer holding times. A possible explanation for
with the information provided by the TTT (Fig. 3a) regarding the this observation is the starting of carbide precipitation, although
stability of the austenite in this range of temperatures. further analyses with TEM would be necessary to confirm this
point.
The final microstructure of the specimen quenched to 260 1C,
which is about 30 1C below the optimum quench temperature,
and partitioned at 470 1C for 60 s, is shown in Fig. 6a. It is possible
to observe the carbon-depleted martensite laths, with some
carbides inside. Their presence was likely due to a slight auto-
tempering of the martensite before the partitioning, as they
were also observed in tests with a partitioning time of 1 s. No
quantitative measurements of carbides density were carried out,
hence a further precipitation during partitioning cannot be
excluded. The unetched areas can be either retained austenite
(RA) or untempered high carbon martensite (UM). The fraction of
these areas is greater than the fraction of retained austenite
measured by the magnetic technique, therefore some high-
carbon martensite formed during the last quench appears to be
present.
The insufficient carbon enrichment of the austenite that comes
Fig. 5. Retained austenite contents for various QT, after partitioning at 470 1C and
from quenching above the optimum temperature can be observed
500 1C. The arrows indicate the trends with increasing partitioning times. The dot- in Fig. 6b, which shows the presence of coarse islands of untem-
dashed line reproduces the calculated values of Fig. 1. pered martensite (UM). When the QT of 325 1C was reached, the
Fig. 6. Microstructures partitioned at 470 1C for 60 s after quenching to (a) 260 1C (volume fraction of retained austenite ¼0.14), (b) 325 1C (volume fraction of retained
austenite¼ 0.12) and (c) microstructure partitioned at 500 1C for 600 s after quenching to 300 1C (retained austenite fraction 0.11). RA: retained austenite, TM: tempered
martensite and UM: untempered martensite.
E. Paravicini Bagliani et al. / Materials Science & Engineering A 559 (2013) 486–495 491
austenite stabilization. The retained austenite content of QP-B was corresponding to the maximum in the volume fraction of retained
in line with the Q&P samples and its average carbon concentration austenite was shifted to lower quenching temperatures in the
was a bit higher. After the Q&T heat treatments no retained case of the specimens treated with salt bath (250 1C instead of
austenite was detected. 280 1C). These differences can be due to a less accurate control
The retained austenite fractions measured in salt bath speci- of the thermal parameters during the application of the heat
mens were generally lower and less sensitive to QT variations treatments in salt bath, with respect to the better-controlled
than those measured in the dilatometric specimens. The peak temperature profile in the dilatometer. In order to clarify this
aspect, microstructures of some selected specimens treated in the
salt baths are displayed in Fig. 10. Two types of constituents can
be observed after Q&P treatments in the salt bath:
Fig. 10. Scanning electron micrographs of the following specimens: (a,b) QP252; (c) QP243; (d,e) QP263 and (f) QT550 (550 1C). RA: retained austenite, MA: high carbon
martensite-retained austenite constituent, TM: tempered martensite, UM: untempered martensite, LB: lower bainite.
E. Paravicini Bagliani et al. / Materials Science & Engineering A 559 (2013) 486–495 493
of the microstructure
Table 5
4.3.1. Tensile properties Tensile properties and 50% fracture appearance transition temperature (FATT) of
Fig. 11a shows the yield strength, ultimate tensile strength (UTS) QT500, QT550, QP243 and QP253 specimens.
and total elongation (TE) versus the quenching temperature of the
ID YS (MPa) UTS (MPa) TE (%) 50% FATT (1C)
specimens treated following Q&P schedules with the salt bath. The
steels exhibited almost constant UTS between 1450 and 1500 MPa, QT500 1194 7 12 1415 7 7 14.0 7 0.5 4100
except for the case with the highest quenching temperature, with QT550 1141 7 0 1300 7 0 13.8 7 0 68
UTS¼1600 MPa, where the presence of untempered martensite QP243 1170 7 0 1505 7 5 13.7 7 0 13
was significant. The YS values were between 800 and 1175 MPa and QP253 1175 7 15 1457 7 5 14.4 7 0.3 10
90
Ref.5
80
70 Present
study
50
Ref.6
40 QP243, QP253, YS = 1170 MPa
Ref. 5 Q&T, YS = 1165 MPa
30 Ref. 7 Bainitic steel YS = 965 MPa
Ref. 7 Bainitic steel YS = 1230 MPa
Ref. 6 Bainitic steel YS = 1150 MPa
20
-120 -100 -80 -60 -40 -20 0 20 40 60 80 100
T C
Fig. 13. Charpy V notch impact energies after Q&P compared with data from the
literature of bainitic steels and Q&T steels.
effects: the higher stabilization of the austenite, the carbon [3] J.G. Speer, D.V. Edmonds, F.C. Rizzo, D.K. Matlock, Curr. Opin. Solid State
depletion of martensite and the reduction of untempered Mater. Sci. 8 (2004) 219–237.
[4] M. Ma, H. Yi, in: J. Weng, H. Dong, Y. Gan (Eds.), Advanced Steels: The Recent
martensite fraction. The films and islands of retained austenite Scenario in Steel Science and Technology, Metallurgical Industry Press,
between the blocks and packets of the tempered martensite can Beijing and Springer-Verlag, Berlin Heidelberg, 2011, pp. 187–198.
also have a role in the refinement of the Q&P microstructure. [5] M. Jellison, R.B. Chandler, Asia Pacific oil and gas conference and exhibition,
in: Conference Proceedings, SPE145404, 2011.
(c) The study of toughness properties with different heat treat- [6] F.G. Caballero, H.K.D. Bhadeshia, Int. J. ISSI 1 (2004) 15–23.
ments conditions highlighted the importance to avoid islands [7] G. Gomez, T. Perez, H.K.D.H. Bhadeshia, Mater. Sci. Technol. 25 (2009)
of untempered martensite, if the purpose is to combine 1501–1512.
[8] F.G. Caballero, J. Chao, J. Cornide, C. Garcı́a-Mateo, M.J. Santofimia, C. Capdevila,
high YS and low FATT. The autotempering of martensite was Mater. Sci. Eng. A 525 (2009) 87–95.
regarded as the main competing process to the stabilization of [9] S.C. Hong, J.C. Ahn, S.Y. Nam, S.J. Kim, H.C. Yang, J.G. Speer, D.K. Matlock, Met.
the austenite, leading to the presence of small islands of Mater. Int. 13 (6) (2007) 439–445.
[10] E. De Moor, J.G. Speer, D.K. Matlock, J. Kwak, S. Lee, ISIJ Int. 51 (2011)
austenite that are not sufficiently enriched in carbon to be
137–144.
stabilised to room temperature. [11] E. De Moor, C. Föjer, A.J. Clarke, J. Penning, J.G. Speer, Proceedings of the
(d) The Q&T treatment was not appropriate for the selected steel. International Conference on New Developments on Metallurgy and Applica-
The absence of carbon partitioning led to a very poor tough- tions of High Strength Steels, vol. 2, 2008 pp. 721–730.
[12] S. Bechet, L. Beaujard, La Revue de Métallurgie 52 (1955) 830–836.
ness at similar YS levels, due to the presence of coarse [13] L. Zhao, N.H. van Dijk, E. Brück, J. Sietsma, S. van der Zwaag, Mater. Sci. Eng. A
carbides in the microstructure. 313 (2001) 145–152.
[14] C.F. Jatczak, J.A. Larson, S.W. Shin, Retained Austenite and its Measurements
by X-Ray diffraction, Society of Automotive Engineers, SAE no. SP-453,
Warrendale, PA, 1980.
Acknowledgments [15] N.H. van Dijk, A.M. Butt, L. Zhao, J. Sietsma, S.E. Offerman, J.P. Wright, S. van
der Zwaag, Acta Mater. 53 (2005) 5439–5447.
Dr. Mario Rossi, director of Tenaris Dalmine Research & [16] G. Krauss, Mater. Sci. Eng. A 273–275 (1999) 40–57.
[17] M. Santofimia, J.G. Speer, A.J. Clarke, L. Zhao, J. Sietsma, Acta Mater. 57 (2009)
Development Center, is kindly acknowledged for permission to 4548–4557.
publish this work. [18] D.H. Kim, J.G. Speer, H.S. Kim, B.C. DeCooman, Metall. Mater. Trans. A 40A
(2009) 2048–2060.
[19] J.G. Speer, R.E. Hackenberg, B.C. DeCooman, D.K. Matlock, Philos. Mag. Lett. 87
References (2007) 379–382.
[20] H.K.D.H. Bhadeshia, R. Honeycombe, Steels Microstructures and Properties,
3rd edition, Butterworth–Heinemann, Oxford, 2006, pp. 135–142.
[1] J.G. Speer, E. De Moor, K.O. Findley, D.K. Matlock, B.C. De Cooman, [21] D.V. Edmonds, K. He, F.C. Rizzo, J.G. Speer, D.K. Matlock, Proceedings of the
D.V. Edmonds, Metall. Mater. Trans. A 42A (2011) 3591–3601. International Conference on New Developments on Metallurgy and Applica-
[2] J.G. Speer, D.K. Matlock, B.C. De Cooman, J.G. Schroth, Acta Mater. 51 (2003) tion of High Strength Steels, Buenos Aires 2008.
2611–2622. [22] E. Kozeschnik, H.K.D.H. Bhadeshia, Mater. Sci. Technol. 24 (3) (2008) 343–347.