UNIVERSITY OF OULU P .O. Box 8000 F I -90014 UNIVERSITY OF OULU FINLAND
A C T A U N I V E R S I T A T I S O U L U E N S I S
Professor Esa Hohtola
University Lecturer Santeri Palviainen
Postdoctoral research fellow Sanna Taskila
Professor Olli Vuolteenaho
University Lecturer Veli-Matti Ulvinen
Director Sinikka Eskelinen
Professor Jari Juga
University Lecturer Anu Soikkeli
Professor Olli Vuolteenaho
Publications Editor Kirsti Nurkkala
ISBN 978-952-62-1214-2 (Paperback)ISBN 978-952-62-1215-9 (PDF)ISSN 0355-3213 (Print)ISSN 1796-2226 (Online)
U N I V E R S I TAT I S O U L U E N S I SACTAC
TECHNICA
U N I V E R S I TAT I S O U L U E N S I SACTAC
TECHNICA
OULU 2016
C 569
Anna Kisko
MICROSTRUCTURE AND PROPERTIES OF REVERSION TREATED LOW-Ni HIGH-Mn AUSTENITIC STAINLESS STEELS
UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF TECHNOLOGY
C 569
ACTA
Anna K
isko
C569etukansi.kesken.fm Page 1 Wednesday, April 27, 2016 2:24 PM
A C T A U N I V E R S I T A T I S O U L U E N S I SC Te c h n i c a 5 6 9
ANNA KISKO
MICROSTRUCTURE AND PROPERTIES OF REVERSION TREATED LOW-Ni HIGH-Mn AUSTENITIC STAINLESS STEELS
Academic dissertation to be presented, with the assent ofthe Doctoral Training Committee of Technology andNatural Sciences of the University of Oulu, for publicdefence in the Wetteri auditorium (IT115), Linnanmaa, on10 June 2016, at 12 noon
UNIVERSITY OF OULU, OULU 2016
Copyright © 2016Acta Univ. Oul. C 569, 2016
Supervised byProfessor David Porter
Reviewed byProfessor Paulo FerreiraProfessor Ahmad Kermanpur
ISBN 978-952-62-1214-2 (Paperback)ISBN 978-952-62-1215-9 (PDF)
ISSN 0355-3213 (Printed)ISSN 1796-2226 (Online)
Cover DesignRaimo Ahonen
JUVENES PRINTTAMPERE 2016
OpponentsProfessor Paulo FerreiraProfessor Hannu Hänninen
Kisko, Anna, Microstructure and properties of reversion treated low-Ni high-Mnaustenitic stainless steels. University of Oulu Graduate School; University of Oulu, Faculty of TechnologyActa Univ. Oul. C 569, 2016University of Oulu, P.O. Box 8000, FI-90014 University of Oulu, Finland
Abstract
In this thesis, the influence of reversion and recrystallization annealing on microstructure andmechanical properties was studied in metastable austenitic low-Ni high-Mn stainless steels, somealloyed with up to 0.45 wt.% Nb. Further, the effect of the various microstructures created byreversion and recrystallization on strain-induced martensite transformation in tensile testing wasinvestigated. The aim was to achieve excellent combinations of strength and ductility in the steelsand to improve understanding of the behaviour of ultrafine-grained austenitic stainless steelsduring deformation. All the steels were cold-rolled up to 60% thickness reduction producing up to60% strain-induced α’-martensite in the austenitic structure. Annealing was carried out using aGleeble thermomechanical simulator between 450–1100 °C for durations of 0.1–1000 s. Theresultant microstructures were examined using different research equipment and methods.
Regardless of the amount of Nb alloying, shear- and diffusion-controlled reversion could becompleted by annealing at 700 °C, although at this temperature no recrystallization of theuntransformed cold-rolled austenite occurred. At 800 °C, however, the cold-rolled austeniterecrystallized, producing a non-uniform grain structure comprising ultrafine-grained areas formedvia reversion and coarser ones formed by recrystallization of the retained austenite. At 900 °C, auniform fine austenite grain size of about 2 μm was obtained. At higher annealing temperatures of1000–1100 °C, normal grain growth of fine grains took place during prolonged annealing in steelwith no Nb. However, grain growth was effectively retarded by alloying with 0.28 wt.% Nb.
The non-uniform structures consisting of reverted and retained austenite exhibited excellentcombinations of yield strength and uniform elongation. The results also showed that tensile strain-induced martensite nucleation sites and α’-martensite formation vary in a complex way dependingon grain size.
Keywords: austenitic stainless steel, grain size refinement, niobium alloying, reversionand recrystallization annealing, strength
Kisko, Anna, Reversiokäsiteltyjen matalanikkelisten, korkeamangaanistenausteniittisten ruostumattomien terästen mikrorakenne ja ominaisuudet. Oulun yliopiston tutkijakoulu; Oulun yliopisto, Teknillinen tiedekuntaActa Univ. Oul. C 569, 2016Oulun yliopisto, PL 8000, 90014 Oulun yliopisto
Tiivistelmä
Väitöstyössä tutkittiin reversiohehkutuksen vaikutusta metastabiilin 1% nikkeliä ja 9% mangaa-nia sisältävien austeniittisten ruostumattomien terästen mikrorakenteeseen ja mekaanisiin omi-naisuuksiin sekä austeniitin raekoon ja mikrorakenteen vaikutusta muokkausmartensiitin syn-tyyn vetokokeessa. Koeteräksistä osa oli lisäksi niobiseostettuja. Tavoitteena oli nostaa teräksi-en lujuutta ja ymmärtää ultrahienorakeisen austeniittisten ruostumattomien terästen käyttäyty-mistä muokkauksessa. Teräkset kylmämuokattiin 60% valssausreduktiolla, jolloin austeniitti-seen rakenteeseen muodostui muokkausmartensiittia enimmillään 60%. Reversiohehkutuksettehtiin Gleeble termomekaanisella simulaattorilla lämpötiloissa 450–1100 °C ja 0.1–1000 s pito-ajoilla. Saatuja mikrorakenteita tutkittiin eri tutkimuslaitteistoilla ja -menetelmillä.
700 °C hehkutuksessa leikkautumalla ja diffuusion välityksellä tapahtuva reversio oli nopeamyös niobi-seostetuilla teräksillä, mutta rekristallisaatiota ei tapahtunut. 800 °C hehkutuksessamuokkauksessa teräksiin jäänyt austeniitti rekristallisoitui, mutta raerakenne muodostui epäta-saiseksi koostuen reversion tuottamasta ultrahienoista rakeista ja jäännösausteniitin rekristallis-aation tuottamista karkeammista rakeista. Sitä vastoin hehkutus 900 °C:ssa tuotti tasainen 2 μmausteniitin raekoon. Pitkissä hehkutuksissa korkeammissa lämpötiloissa 1000–1100 °C niobi-seostamattomissa teräksissä tapahtui hienojen rakeiden normaalia rakeenkasvua. Kuitenkin0.28p-% niobi-seostuksen havaittiin oleva riittävä estämään rakeenkasvu.
Reversion ja osittaisen rekristallisaation tuottamilla raerakenteilla saatiin erinomaiset myötö-lujuus-tasavenymäyhdistelmät. Vetokokeissa martensiitin ydintymispaikat ja -nopeus vaihtelivatmonimutkaisella tavalla raekoosta riippuen.
Asiasanat: austenittiinen ruostumaton teräs, lujuus, niobiseostus, raekoonhienontaminen, reversio- ja rekristallisaatiohehkutus
7
Acknowledgements
This thesis comprises work carried out in 2010–2015 in the Materials Engineering
and Production Technology research unit of the Centre for Advanced Steels
Research (CASR) at the University of Oulu. The research was funded by the
University of Oulu Graduate School on Advanced Materials Doctoral Programme
(ADMA-DP) and the Finnish Funding Agency for Technology and Innovation
(Tekes) as a part of the Light and Efficient Solutions program (LIGHT) of the
Finnish Metals and Engineering Competence Cluster (FIMECC Ltd). The
experimental materials were provided by Outokumpu Oyj. This thesis was also
supported by scholarships awarded by the Jenny ja Antti Wihuri Foundation, the
KAUTE foundation, Metallinjalostajien rahasto, Tauno Tönningin säätiö,
Tekniikan Edistämissäätiö (TES) and the Walter Ahlström Foundation.
During this project I have received guidance and support from several experts
and other people whom I would like to acknowledge. I would like to thank my
supervisor, Professor David Porter, for his guidance and support in my research. I
would like to present my sincere gratitude and best thanks to Professor Emeritus
Pentti Karjalainen for his scientific supervision, kindness and suggestions for
improving the quality of my research and papers during this entire work.
I would like to express my gratitude to Dr. Juho Talonen of Outokumpu Oyj
for his extremely valuable comments throughout the research project. I also wish
to thank Professor Anthony DeArdo from the University of Pittsburgh, Professor
Devesh Misra from the University of Louisiana at Lafayette, Assistant Professor
Puspendu Sahu from Jadavpur University, Associate Professor Atef Hamada from
Suez University and Dr. Ludovica Rovatti from Polytechnic University of Milan
for their contributions to the published papers. I would also like to thank Professor
Paulo Ferreira from the University of Texas at Austin and Professor Ahmad
Kermanpur from Isfahan University of Technology for the pre-examination of this
thesis and for the suggestions for improving the manuscript.
I express my thanks to my colleagues in the Materials Engineering Research
Group. I especially thank Dr. Saara Mehtonen for her peer support.
Finally, I would like to thank my friends and family, especially my Mom, Dad
and sister Kaisa, for their endless support. I thank my dearest Pasi and Pauli for
encouragement in every precious moment during these years.
Oulu, April 2016 Anna Kisko
8
9
Symbols and abbreviations
Af Austenite finish temperature [°C]
Ag Uniform elongation [%]
As Austenite start temperature [°C]
d Average grain size [μm or nm]
D Average radius of the grains [μm]
d0 Initial grain size [μm]
f Volume fraction of particles
Fd Driving force for grain growth [MPa]
Fp Pinning force for grain growth [MPa]
ΔGactivation Activation energy for grain growth [J/mol] ∆ → Free energy difference between austenite and martensite [J/mol]
k Constant (page 30)
k Grain growth kinetic parameter (page 29)
k0 Material constant
Md Upper limit for strain-induced transformation
Ms Martensite start temperature
n Grain growth exponent
N Number of measured particles
r Radius of the spherical particles [nm]
R Gas constant [J/(K mol)]
Rp0.2 Yield strength [MPa]
Rm Tensile strength [MPa]
t Annealing time [s]
T Annealing temperature [K] (Equation 2)
T Annealing temperature [°C] (except Equation 2)
T0 Temperature, where austenite and martensite are in
thermodynamic equilibrium
W Width [mm]
W0 Initial width [mm]
α’ Alpha prime (used to denote martensite as opposed to α for
ferrite)
γ Austenite
γ Grain boundary energy [J/m2] (page 29)
ε Epsilon-martensite
σy Yield strength [MPa]
10
σ0 Yield strength without grain boundaries [MPa]
CG Coarse austenite grains (> 18 μm)
EBSD Electron backscatter diffraction
FBM Flexible boundary model
FEG-SEM Field emission gun scanning electron microscope
FETEM Field emission transmission electron microscope
FG Fine austenite grains (∼4 μm)
MG Micron-scale austenite grains (∼1.5 μm)
LOM Light optical microscope
RA Retained austenite
SFE Stacking fault energy
SHR Strain hardening rate
TEM Transmission electron microscope
UFG Ultrafine austenite grains (∼0.5 μm)
XRD X-ray diffraction
11
List of original publications
This thesis is based on the following publications, which are referred to throughout
the text by their Roman numerals:
I Kisko A, Hamada A, Karjalainen LP & Talonen J (2011) Microstructure and mechanical properties of reversion treated high Mn austenitic 204Cu and 201 stainless steels. The 1st International Conference on High Manganese Steels. Seoul, Korea, 15-18 May. Research Institute of Iron and Steel Technology, Yonsei University, B-19.
II Kisko A, Rovatti L, Miettunen I, Karjalainen P & Talonen J (2011) Microstructure and anisotropy of mechanical properties in reversion-treated high-Mn type 204Cu and 201 stainless steels. 7th European Stainless Steel Conference Science and Market. Como, Italy, 21-23 September.
III Kisko A, Rovatti L, Misra D, Sahu P, Talonen J & Karjalainen P (2013) Studies on martensite transformation in a metastable austenitic Cr-Mn stainless steel. Materials Science Forum 762: 424-430.
IV Kisko A, Hamada A.S, Talonen J, Porter D & Karjalainen LP (2016) Effects of reversion and recrystallization on microstructure and mechanical properties of Nb-alloyed low-Ni high-Mn austenitic stainless steels. Materials Science & Engineering A 657: 359-370.
V Kisko A, Talonen J, Porter D & Karjalainen LP (2015) Effect of Nb microalloying on reversion and grain growth in a high-Mn 204Cu austenitic stainless steel. ISIJ International 55(10): 2227-2234.
VI Kisko A, Misra D, Talonen J & Karjalainen LP (2013) The influence of grain size on the strain-induced martensite formation in tensile straining of an austenitic 15Cr-9Mn-Ni-Cu stainless steel. Materials Science & Engineering A 578: 408-416.
Anna Kisko has been the main and corresponding author of all the publications.
She has planned the tests and performed all the experiments, except the Gleeble
annealings for Papers I–VI, the tensile tests for Papers I and II, the TEM
examinations for Papers II, III and VI, the XRD runs for Papers III and IV and the
dislocation density analysis for Paper III. She has also done the data analysis. Juho
Talonen and Pentti Karjalainen have commented on the contents of Papers I–VI
and David Porter on Papers IV and V.
Paper I is an overall, comprehensive study of the microstructures and
properties of reversion-treated 15Cr-9Mn-1Ni-2Cu (coded 0Nb) steel. It compares
the enhanced mechanical properties with conventional values (reported in the
materials standards) and other steels.
In Paper II, the microstructure and formed anisotropy of mechanical properties
of the 0Nb steel in the temper-rolled stage and the annealed stage are discussed.
12
Paper III focuses on α’-martensite transformation during reversion annealing
of the 0Nb steel. The properties of α’-martensite, formed during cold rolling or
annealing, was studied.
In Paper IV, the reversion mechanisms, recrystallization of retained austenite
and mechanical properties of the 0Nb and Nb-alloyed 0.28Nb and 0.45Nb steels
are described in detail. Hall-Petch analysis was carried out and the strengthening
mechanisms of the steels were discussed.
Paper V focuses on grain growth in the 0Nb, 0.05Nb, 0.11Nb, 0.28Nb and
0.45Nb steels during prolonged, high-temperature annealing. Driving and pinning
forces for grain growth are calculated and the grain growth behaviour of the steels
is explained.
Paper VI discusses ε- and α’-martensite nucleation sites and volume fraction
of α’-martensite at the beginning of tensile straining in the reversion-treated 0Nb
steel with various grain structures, i.e., reverted grains and retained austenite,
micron-sized grains and retained austenite, fine-grained austenite and coarse-
grained austenite.
13
Table of contents
Abstract
Tiivistelmä
Acknowledgements 7 Symbols and abbreviations 9 List of original publications 11 Table of contents 13 1 Introduction 15
1.1 Preface ..................................................................................................... 15 1.2 Plastic deformation of metastable austenitic stainless steels ................... 16
1.2.1 Thermodynamics of strain-induced martensite
transformation .............................................................................. 16 1.2.2 Stacking faults, ε-martensite and deformation twinning .............. 18 1.2.3 α’-martensite ................................................................................ 19
1.3 Reversion, recrystallization and grain growth ......................................... 20 1.3.1 Reversion mechanism ................................................................... 20 1.3.2 Factors affecting reversion ........................................................... 20 1.3.3 Recrystallization and grain growth ............................................... 21
1.4 Advantages of reversion treatment .......................................................... 23 1.5 Aims of the study .................................................................................... 24
2 Experimental 25 2.1 Materials ................................................................................................. 25 2.2 Research methods and equipment ........................................................... 25
3 Results 27 3.1 Microstructure and properties of cold-rolled steel .................................. 27 3.2 Microstructural evolution during annealing ............................................ 28
3.2.1 Reversion of martensite ................................................................ 28 3.2.2 Recrystallization of retained austenite .......................................... 32 3.2.3 Austenite grain growth ................................................................. 33
3.3 Nucleation and growth of ε- and α’-martensite in tensile
straining ................................................................................................... 36 3.4 Mechanical properties ............................................................................. 38
3.4.1 Hardness evolution ....................................................................... 38 3.4.2 Tensile properties .......................................................................... 40
4 Discussion 43 4.1 Anisotropy of mechanical properties ...................................................... 43
14
4.2 ε- and α’-martensite transformation ........................................................ 43 4.3 The impact of reversion and recrystallization on microstructure,
grain size and grain growth ..................................................................... 45 4.4 Nucleation and growth of ε- and α’-martensite during plastic
deformation ............................................................................................. 47 4.5 Effect of microstructural evolution on mechanical properties ................ 48 4.6 Recommendations for further research ................................................... 52
5 Summary and conclusions 55 6 Novel features 57 List of references 59 Original publications 67
15
1 Introduction
1.1 Preface
For more than 60 years manganese has been considered a replacement for nickel in
austenitic stainless steels, particularly at times of high nickel prices. Instead of
adding nickel, these so-called “200-series” austenitic stainless steels are often
alloyed with nitrogen and copper to stabilize the austenitic phase. As a result, a
good combination of strength and ductility is provided. This enables the use of
“200-series” austenitic stainless steels in a wide range of applications, with much
lower alloying costs. [1,2]
Due to their good mechanical properties, formability and architectural image,
typical applications for high-Mn steels are lightweight applications, such as indoor
or outdoor architecture and bus bodies, as well as kitchenware, e.g., cutlery. In fact,
in these applications high-Mn grades are designed to replace commonly used Cr-
Ni austenitic grades (“300-series”). However, these steels are less resistant to
corrosion than the “300-series”, which limits their applications. [1,2]
Austenitic stainless steels typically have good ductility and formability, but
their yield strength is quite low. There are several methods for improving strength,
e.g., solid solution strengthening, work hardening and grain size refinement. [3-5]
Many “200-series” steels are metastable at room temperature, so that during cold
working the formation of strain-induced martensite occurs readily, increasing
strength. The disadvantage of strengthening by cold rolling is the formation of
undesirable anisotropy in mechanical properties, with strength being different in
different directions relative to the rolling direction. Therefore, other strengthening
methods are desirable. [6]
Proper annealing treatment of cold-strengthened metastable austenitic stainless
steel will revert strain-induced martensite back to austenite. Reverted austenite can
have a fine grain size of a few hundred nanometers. As a result of this efficient
grain size refinement, an excellent combination of yield strength and elongations
is achievable, as shown in numerous studies during the last 25 years. [4,7-12]
However, the annealing cycle is a critical processing stage, as during prolonged
or high-temperature annealing rapid grain growth tends to occur readily [3,4,9,13-
15]. Therefore, prevention or retardation of grain growth becomes important. Fine
precipitation is a method for used for this [16].
16
Alloying elements can be classified as either austenitic stabilizers, such as
nickel, nitrogen, carbon, manganese, copper and cobalt, or ferrite stabilizers, such
as chromium, silicon and molybdenum. Nickel addition is an effective way to
ensure the austenitic structure of stainless steel. However, by adding other
austenitic stabilizers, the austenitic structure can be achieved even without nickel,
or with a low nickel content such as 1%. Nitrogen provides high-strength properties
and copper improves drawability. When copper is added, nitrogen content can be
reduced to obtain a softer material with better formability. Lowering nickel content
without adding other austenite stabilizers, however, reduces the possible amount of
chromium in the alloy, weakening corrosion resistance and narrowing the range of
applications for which the steel is suited. [1,2,17]
Niobium is a ferrite stabilizer, but it is also a strong carbide- and nitride-
forming element. Niobium carbides are much less soluble in austenite than, for
instance, more unstable chromium carbide, and they form at relatively high
temperatures [18]. In this way, niobium alloying helps refine grain size and retard
recrystallization [19].
1.2 Plastic deformation of metastable austenitic stainless steels
Austenitic stainless steels have an fcc crystal structure due to the alloying. However,
this structure is not thermodynamically stable at ambient temperatures. Therefore,
applied stress or plastic deformation, e.g., cold rolling or tensile straining, may
induce a martensitic phase transformation where the metastable austenite phase
transforms into the thermodynamically more stable martensite phase, i.e.,
γ→(ε)→α’. Two different martensite phases may exist in austenitic stainless steels,
hexagonal close-packed (hcp) ε-martensite and body-centred cubic (bcc) α’-
martensite. Due to the relatively low interstitial content, the crystal structure of α’-
martensite is normally referred to as bcc instead of bct (body-centred tetragonal).
1.2.1 Thermodynamics of strain-induced martensite transformation
The chemical free energy changes of the austenite and martensite phases as a
function of temperature are illustrated schematically in Fig. 1. At temperature T0
austenite and martensite are thermodynamically in equilibrium and at Ms
martensitic transformation starts upon cooling. Martensitic transformation can
occur only if the difference between the chemical free energies of the austenite and
martensite phases reaches the critical value of ∆ → . This takes place at
17
temperature Ms. However, the transformation can also occur at temperatures
between Ms and T0 if a mechanical driving force U’ is applied. If T0 is low enough,
the microstructure remains austenitic. [20]
Fig. 1. Chemical free energy changes of the austenite and martensite phases according
to Cahn and Haasen [20].
Fig. 2 shows that the critical applied stress to induce martensite formation increases
linearly with an increase in stressing temperature in the range between Ms and .
Below temperature yielding occurs by martensitic transformation, whereas at a
higher temperature, by means of slip in the austenite phase. [20-21] According to
Olson and Cohen [21], below temperature martensitic transformation is called
stress-assisted and above it, strain-induced nucleation. The former occurs below
the yield strength of the austenite phase with the aid of applied stress, and the latter
occurs after plastic deformation of the austenite phase. Temperature Md is the upper
limit for strain-induced transformation, as the chemical driving force of the phase
transformation becomes so small that nucleation of martensite cannot be
mechanically induced. [20,21] The characteristics of strain-induced martensite
transformation can be modified by several factors, including chemical composition,
temperature, strain rate, strain state and austenitising conditions [22].
18
Fig. 2. Critical stress required to initiate the formation of martensite as a function of
temperature according to Olson and Cohen 1972 [21].
1.2.2 Stacking faults, ε-martensite and deformation twinning
Stacking faults in austenite are formed between Shockley partial dislocations that
form from the dissociation of a perfect dislocation. A stacking fault is a
discontinuity in the stacking order of the {111} planes. Stacking fault energy (SFE)
controls the width between two partial dislocations, since it balances the repulsive
force between the partial dislocations. Austenitic stainless steels have low SFE, and
therefore stacking faults are relatively wide, which makes cross-slip relatively
difficult. SFE controls the formation of dislocation cells, twins and ε-martensite
during plastic deformation. The lower the SFE of the steel, the lower the stability
of austenitic stainless steel, and the higher the tendency for ε-martensite to form
instead of twins [3,23-25].
A change in the sequence of the {111} plane produces a thin layer of a
hexagonal close-packed phase. Hence, even a single stacking fault can be
considered a nucleus of hcp ε-martensite [26]. The ε-martensite grows by an
overlapping of stacking faults on every second {111} plane [27-29]. It is difficult
to distinguish a single stacking fault, bundles of overlapping stacking faults, faulted
and perfect ε-martensite, and therefore, the term “shear band” is often used to
describe such microstructural features. [30]
Plastic deformation of austenitic stainless steel can induce deformation
twinning. A deformation twin is formed as the result of gliding of Shockley partial
dislocations of the same sign on successive adjacent {111} planes. Then the twin
19
grows in thickness. The driving force for twinning is the applied stress, while the
driving force for martensite formation is the free energy difference between the
austenite and martensitic phases, which may, however, be assisted by the applied
stress. [3,31,32].
The incidence of ε-martensite, mechanical twins or dislocation glide are
closely related to the SFE of the steel. ε-martensite formation is preferred when the
chemical free energy difference between metastable austenite and ε-martensite
phases is negative, as the hcp structure has lower energy than a twin with fcc
stacking. An early work by Rémy and Pineau [24] on an Fe-Mn-Cr-C alloy
suggested that ε-martensite transformation occurs when SFE is below ∼13 mJ/m2,
twinning when SFE is between ∼8 and ∼40 mJ/m2 and dislocation cells when SFE
is above ∼18 mJ/m2. Later Allain et al. [33] studied an Fe-Mn-C system and Lee et
al., [34] examined Fe-Cr-Mn-N-C alloys and concluded that strain-induced
martensite transformation occurs if SFE is below 15 or 18 mJ/m2 and formation of
twins in steels if SFE is above 20 mJ/m2, or between 12 and 35 mJ/m2.
1.2.3 α’-martensite
α’-martensite has been found to nucleate on deformation-induced defects, such as
shear bands [35,36] and their intersections [21,26,27,29,37-43], or at the
intersection of deformation twins [29,43-47]. However, α’-martensite can also
nucleate straight from austenite [48-52] and at grain boundaries [53-56]. As a shear
band may consist of faulted and perfect ε-martensite, it is considered to be an
intermediate phase in the formation of α’-martensite. [30,31,49,57]. Generally, α’-
martensite nucleation involves a process by which an array of Shockley partial
dislocations can penetrate through another shear band.
α’-martensite grows by repeated nucleation of new α’-martensite embryos and
coalescence. At small strains, α’-martensite has been found to grow along slip
planes, and with higher strains, outside of the active slip plane. Hence, the
morphology is irregular and blocky. [43,58]
The γ → α’ phase transformation affects the mechanical properties of
metastable austenitic stainless steel. Firstly, it increases the strain hardening rate
(SHR) and thereby increases tensile strength. For a metastable steel, a curve
showing strain hardening as a function of increasing strain first exhibits a minimum
and then a subsequent maximum [31,39,59-62]. Secondly, the strain hardening
capability governs ductility, measured by uniform elongation. It has been shown
that the rate and the point on the stress-strain curve at which α’-martensite
20
transformation takes place are important, rather than the total amount of formed α’-
martensite. The deformation temperature has a major impact on the microstructure
formed, as austenite stability depends on temperature. With increasing temperature,
α’-martensite transformation is prevented. With decreasing temperature, α’-
martensite transformation and strain hardening occur readily at low strains, causing
premature fracture. [60,63, 64]
1.3 Reversion, recrystallization and grain growth
1.3.1 Reversion mechanism
α’-martensite transformation into austenite (α’→ γ) in ferrous alloys between the
austenite start (As) and austenite finish (Af) temperatures is known as reversion [65].
Two different reversion mechanisms are proposed in the literature: diffusion-type
and shear-type mechanisms [7,8].
Diffusion-type reversion consists of random nucleation of austenite in α’-
martensite laths or at the lath boundaries. Austenite orientation in the
crystallography is random. Then the nucleated grains grow into equiaxed grains.
[8,10]
In the case of shear-type reversion, the α’-martensite phase reverts back to the
austenite phase via diffusionless transformation. The austenite formed contains a
high dislocation density and its morphology is associated with the parent α’-
martensite. During prolonged annealing, defects recover to form sub-grains that
then recrystallize into larger austenite grains in the range of a few μm [9]. Shear-
type reversion occurs within a narrow temperature range of ∼50 °C, whereas
diffusion-type reversion occurs over a larger temperature range of ∼200 °C
[8,66,67]. Tomimura et al., [8] among others, studied 16Cr-10Ni and 18Cr-9Ni
steels and found that at 10 s annealing reversion occurs through shear in a narrow
temperature range of around 577 °C for the 16Cr-10Ni steel and through diffusion
in a wider temperature range of 527–727 °C for the 18Cr-9Ni steel.
1.3.2 Factors affecting reversion
The reversion mechanism that is found to take place depends on the heating rate
and chemical composition. A change in the reversion mechanism from diffusion-
type to shear-type with an increasing heating rate has been reported in the literature
21
by, e.g., Apple and Krauss, who studied Fe-Ni-C alloys with heating rates of 3 and
1500 °C/s [68], Montanari et al., who studied AISI 304 stainless steel with heating
rates of 0.017–1.67 °C/s [69] and Leem et al., who found that the reversion
mechanism in Fe-13Cr-7Ni-3Si martensitic stainless changed at 10 °C/s [70].
Moreover, the reversion mechanism can change from the shear-type, which occurs
during heating, to the diffusion-type during subsequent isothermal holding [66].
Alloy composition affects the temperature dependence of the free energies of
the α’ and γ phases. An increase in the Ni/Cr ratio causes an increase in the Gibbs
free energy difference between α’ and γ, leading to a change in the reversion
mechanism from shear-type to diffusion-type. With decreasing T0, the kinetics of
diffusion required to transform γ to α’ by diffusion-type reversion decreases. The
critical driving force for shear-type reversion is estimated to be about -500 J/mol.
[8]
In an early work by Kaufman et al. [71], the free energy difference of the
ternary Fe-Cr-Ni system was determined. Later, Somani et al. [9] further extended
the equation to include the effects of other alloying elements, such as Si, Mn, Mo,
C and N, which have a strong influence on the microstructure constitution, by
adopting nickel and chromium equivalents instead of pure Ni and Cr. [72].
1.3.3 Recrystallization and grain growth
Austenite, which is deformed but not transformed during cold rolling, termed here
as retained austenite (RA), can recrystallize during annealing. The driving force for
recrystallization is the stored energy from the cold working. The new grains
nucleate particularly at defects, such as shear bands and grain boundaries, and their
growth consumes the whole RA structure. Recrystallization in austenitic stainless
steels depends on several factors, such as cold rolling reduction, i.e., the amount of
stored energy, initial grain size, alloy composition and annealing conditions, i.e.,
time, temperature and heating rate. [3,13,73]
In most studies concerning reversion annealing, e.g., [8,9,12], a mainly
martensitic structure was used as a starting material for the annealing. In a partially
martensitic structure, i.e., consisting of α’-martensite and RA, the evolution of the
microstructure is more complicated due to the slower kinetics of recrystallization
compared with that of reversion [9,74-76]. However, only preliminary observations
have been made concerning the competition between reversion and retained
austenite.
22
After recrystallization, grain growth may occur with prolonged annealing. The
driving force for grain growth is reduction of total grain boundary energy. At the
atomic scale, this occurs by a net transfer of atoms from the concave side of curved
grain boundaries to the convex side due to the lower free energy of the atoms on
the convex side. Other things being equal, the greater the curvature of the boundary,
the faster is its movement. Grain growth can be influenced e.g., by annealing
temperature, solutes and particles. [3,13]
The kinetics of grain growth can be analysed using the generalized grain
growth law, which can be written as:
− = (1)
where d is average grain size, d0 is initial grain size, n is the grain growth exponent,
k is the kinetic parameter of grain growth and t is annealing time. In theory, n is
equal to 2 for an ideal material, but for real materials n values are greater because
grains are not spherical and there are chemical segregations. [3,77]. Therefore,
higher values of n, i.e., 2.4–10, have been reported in the literature for austenitic
stainless steels and nanocrystalline materials [78,79].
The kinetic parameter k can be used to estimate the activation energy for the
grain growth process, which can be written as follows:
= (2)
where k0 is a material constant, ΔGactivation is the activation energy for grain growth,
R is a gas constant and T is annealing temperature. [3] Activation energies of grain
growth between 280–455 kJ/mol have been reported in the literature for austenitic
stainless steels [79-81].
Grain size is controlled by competition between the driving force for grain
growth and the pinning force, so that if the pinning force exceeds the driving force,
the boundary will be pinned, i.e., immobile. There are models for describing the
driving force for grain growth, and Zener described it as follows:
= (3)
where Fd is the driving force predicted by Zener, γ is grain boundary energy and D
is the average radius of the sphere. [82]
Second-phase particles exert a pinning force on grain boundaries and they can
inhibit grain growth. The pinning pressure depends on the size, volume fraction and
distribution of the particles. At high annealing temperatures, second-phase particles
can form, coarsen or dissolve. [3,13]
23
There are various models for explaining how precipitates can suppress
austenite grain growth by pinning grain boundaries. The flexible boundary model
(FBM) assumes that an infinitely flexible boundary is capable of interacting with
every particle in a three-dimensional array until it is fully pinned. It can be
expressed as follows:
= / (4)
where Fp is the pinning force according to the flexible boundary model, f is the
volume fraction of the particles and r is the radius of the spherical particles.
[3,13,83]
1.4 Advantages of reversion treatment
Generally, improved mechanical properties can be obtained in metastable austenitic
stainless steels subjected to reversion annealing, since the reversion treatment
refines grain size, e.g., [4,8,9,12,72]. During cold forming, ε- and α’-martensite
form from the metastable austenitic structure, and α’-martensite grows by repeated
nucleation of the embryos and their coalescence. The fraction of α’-martensite
increases with increasing straining. In diffusional reversion, austenite nucleates at
martensite laths or grain boundaries of α’-martensite, and generally there are a large
number of nucleation sites for austenite inside α’-martensite. Hence, the final grain
size will be much finer than before the treatment. In the instance of shear reversion,
the reverted austenite inherits the martensitic grain structure with a high dislocation
density. The reversion is followed by recrystallization, leading to a refined grain
size. [8,43,58,67,84]
A highly refined grain size of 0.54 μm has been reported for commercial
301LN austenitic stainless steel, giving a superior combination of yield strength
and uniform elongation: 700 MPa and 35%, respectively [4]. However,
precipitation of (Fe,Cr,Mo)23C6 carbides and CrN and Cr2N nitrides may occur at
reversion annealing temperatures, which can have a detrimental effect on corrosion
properties [4,84,85].
Grain size strengthening can be described with the Hall-Petch equation, which
can be written as follows:
= + / (5)
where σy is yield strength, σ0 is yield strength without the contribution of grain
boundaries, k is a constant, and d is average grain size. Yield stress increases with
24
decreasing grain size, as pile-ups in fine-grained materials contain fewer
dislocations and the stress at the tip of the pile-up decreases. Therefore, a larger
applied stress is required to generate dislocations in adjacent grains. Overall, the
strengthening effect in austenitic stainless steels is due to various strengthening
mechanisms, such as solid solution strengthening, precipitate strengthening, strain
hardening and grain boundary strengthening, all of which are dependent on
annealing temperature. [4,13]
1.5 Aims of the study
The objective of this thesis work was to study the relationships between processing
and achieved properties of high-manganese low-nickel austenitic stainless steel.
The steel was chosen due to its metastable characteristics. Austenitic stainless steels
have potential for further optimization of mechanical properties through adjustment
of the deformation mechanisms in cold working and control of subsequent
treatments, such as reversion and recrystallization annealing. The objectives of this
work can be divided into the following tasks:
– To study the details of the transformation of martensite and its reversion.
– To characterize various microstructures, i.e., reverted, retained and recovered
austenite and their combinations, achieved via reversion annealing.
– To study the potential of reversion treatment to improve mechanical properties
and to understand the influence of the various microstructures formed during
reversion annealing on the mechanical properties of low-nickel austenitic
stainless steels.
– To investigate if reversion annealing can eliminate anisotropy at a high strength
level.
– To reveal how the grain size distribution of reverted ultrafine-grained austenite
and coarse cold-worked austenite affects strain-induced martensite formation,
strength and ductility during tensile testing.
– To study the influence of microalloying on microstructures and mechanical
properties, and to clarify the ways in which microalloying affects reversion and
recrystallization kinetics, grain size refinement and the mechanical properties
achieved.
– To establish the potential of microalloying to widen the processing window for
reversion heat treatment of the steel concerned.
25
2 Experimental
2.1 Materials
The steel used as experimental material was “200-series” austenitic stainless steel,
15Cr-9Mn-1Ni-2Cu (204Cu steel; coded as 0Nb). It was a cold-rolled and annealed
final product in 2B delivery condition from Outokumpu Stainless Oy (Tornio,
Finland) production and it was cold-rolled to 5, 12 or 20% thickness reduction in a
laboratory rolling mill in order to study of the influence of light cold rolling
reductions on the anisotropy of the mechanical properties of the steel. Reduction of
60% was taken as the highest practical cold rolling reduction on a production line.
To investigate the influence of alloying, four 60-kg charges of pure raw materials,
alloyed with various contents of Nb, were melted in a laboratory vacuum induction
furnace. Cast ingots were machined, subsequently heat-treated at 1265 °C for 60
min, hot-rolled with 9 passes to a final thickness of 4 mm, annealed at 1100 °C for
5 min, water quenched and pickled and cold-rolled to a final thickness of 1.5 mm
(corresponding to ~60% thickness reduction) at the Tornio Research Center. The
materials were coded, based on their Nb content, as 0.05Nb, 0.11Nb, 0.28Nb and
0.45Nb. The final thicknesses were 0.4 mm for the 0Nb and 1.5 mm for the Nb-
alloyed sheets. The chemical compositions of the steels are listed in Table 1.
Table 1. Chemical composition (in wt.%) of the steels investigated.
Steel C Si Mn Cr Ni Cu N Nb Fe
0Nb 0.079 0.40 9.00 15.2 1.1 1.68 0.115 - bal.
0.05Nb 0.070 0.30 9.13 15.2 1.1 1.70 0.165 0.05 bal.
0.11Nb 0.072 0.28 9.19 15.2 1.1 1.74 0.130 0.11 bal.
0.28Nb 0.083 0.28 9.19 15.1 1.1 1.74 0.160 0.28 bal.
0.45Nb 0.100 0.30 8.90 15.2 1.1 1.70 0.160 0.45 bal.
2.2 Research methods and equipment
The cold-rolled samples were annealed at the University of Oulu using a Gleeble
1500 or Gleeble 3800 thermo-mechanical simulator. The heating rate was 5 °C/s
for annealing tests performed at 450, 500, 550, 585 and 620 °C for 0.1, 1, 10, 100,
300 and 600 s, 50 °C/s, 150 °C/s or 200 °C/s for annealing tests performed at 700,
800, 900, 1000 and 1100 °C for 0.1, 1, 10, 100, 200, 500 and 1000 s. In every test
the cooling rate was 200 °C/s down to 400 °C, followed by free cooling.
26
A Feritscope ® (Helmut Fisher FMP 30) instrument was used to quantify the
ferromagnetic α’-martensite in the steel. The readings of the instrument were
multiplied by a correction factor of 1.7 [86]. X-ray diffraction (XRD) (Siemens-
D500) was used to determine the volume fraction of ε-martensite at the cold-rolled
stage and to estimate the dislocation density of α’-martensite.
Microstructural examinations were performed using a light optical microscope
(LOM) (Nikon Eclipse MA100), field emission scanning electron microscopes
together with electron backscatter diffractions units (EBSD) (Zeiss Ultra Plus FEG-
SEM with a HKL EBSD detector and Zeiss Sigma FEG-SEM with an EDAX
EBSD detector) and transmission electron microscopes (TEM) (LEO 912 FETEM
and Hitachi H7600 TEM).
Mechanical properties were determined at room temperature using tensile
testing (Zwick / Z100) at strain rates of 7.5 x 10-4 s-1 up to 2% strain (0.2% proof
stress) after which the strain rate was 5 x 10-3 s-1 for cold-rolled specimens and 5 x
10-4 s-1 for reversion-treated specimens. Because of the limitations of the size of the
uniformly heated zone in the Gleeble specimens, the tensile test specimens had a
non-standard specimen geometry with a short 15 mm long gage length [87]. The
hardness of the specimens was determined with micro-hardness measurements
(CSM Instrument, micro-indentation tester) with a 10 mg load and a Berkovich
indenter.
Table 2 presents the test materials and the equipment and methods for
microstructure characterization employed in Papers I–VI.
Table 2. Test steels and microstructure characterization methods employed in papers.
Paper Steel Microstructure characterization equipment and method
I 0Nb Feritscope, LOM, Zeiss Ultra Plus FEG-SEM + HKL EBSD
detector, tensile testing, micro-indentation tester
II 0Nb Feritscope, LOM, Zeiss Ultra Plus FEG-SEM + HKL EBSD
detector, LEO 912 FETEM, tensile testing
III 0Nb Feritscope, XRD, Hitachi H7600 TEM, micro-indentation tester
IV 0Nb, 11Nb,
0.45Nb
Feritscope, XRD, LOM, Zeiss Sigma FEG-SEM + EDAX EBSD
detector, tensile testing, micro-indentation tester
V 0Nb, 0.05Nb,
0.11Nb, 0.28Nb, 0.45Nb
Feritscope, LOM, Zeiss Sigma FEG-SEM + EDAX EBSD
detector, LEO 912 FETEM
VI 0Nb Feritscope, LOM, Zeiss Ultra Plus FEG-SEM + HKL EBSD
detector, Hitachi H7600 TEM, tensile testing
27
3 Results
3.1 Microstructure and properties of cold-rolled steel
Initially, the 0Nb steel was fully austenitic without any α’-martensite, (Papers I–
VI). After 5, 12 or 20% cold rolling reduction, the α’-martensite fractions were 2,
10 and 24% (Paper II). After 60% cold rolling reduction, the α’-martensite
fractions of the 0Nb steel were 33% (Papers I and II) or 60% (Papers III–VI). This
indicates a high sensitivity towards α’-martensite formation during cold rolling.
The α’-martensite fractions of the 0.05Nb, 0.11Nb, 0.28Nb and 0.45Nb steels were
53, 63, 47 and 49%, respectively (Papers IV and V). No ε-martensite was detected
in the steels after 60% cold rolling (Paper IV).
The microstructures of the 0Nb steel after various cold rolling reductions are
shown in Fig. 3. At low reductions, slip bands containing ε/α’-martensite and wide
stacking faults become visible (Fig. 3a,b). With increasing rolling reduction, the
number of slip bands increased and finally α’-martensite covered several grains
(Fig. 3c–e).
Fig. 3. LOM (a,c–e) and TEM (b) images of the 0Nb steel after a,b) 5%, c) 12%, d) 20%
and e) 60% cold rolling reduction (coded as CR). [Paper II, Published by permission of
AIM Editorial Office]
The anisotropy of the tensile properties of the 0Nb steel caused by cold rolling was
analysed in Paper II. Yield strengths (Rp0.2 = 0.2% proof stress) in the longitudinal
28
direction and the transverse direction after 5, 12, 20 and 60% cold rolling are shown
in Table 3. It is seen that yield strength increased with rolling reduction and the
values were higher in the transverse direction than in the longitudinal direction.
Anisotropy appeared already at 5% reduction, and anisotropy seemed to be the
smallest with 60 % reduction.
Table 3. Yield strengths (Rp0.2) of the 0Nb steel after 5, 12, 20 and 60% cold rolling
reductions in the longitudinal (LD) and transverse directions (TD). [Paper II, Published
by permission of AIM Editorial Office]
Cold rolling reduction [%]
5 12 20 60
Rp0.2 LD 583 692 815 1522
Rp0.2 TD 632 740 881 1544
3.2 Microstructural evolution during annealing
In reversion annealing, the steel sheet is heated at a certain rate to the final
annealing temperature. Depending on the heating rate and soaking temperature,
during either this heating stage or soaking, α’-martensite will revert back to
austenite. Reversion occurs by either shearing or diffusion. At a higher temperature
than required for reversion, cold-deformed retained austenite recrystallizes and
reverted grains grow. During high-temperature annealing, i.e., 1000–1100 °C, grain
growth occurs. The microstructural evolution during reversion and recrystallization
annealing is discussed in detail in the following sections.
3.2.1 Reversion of martensite
α’-martensite transformation to austenite in the 0Nb steel was studied in Paper III.
Fig. 4 shows the relative change in the specimen’s width (Fig. 4a) and its first
derivative (Fig. 4b) as a function of heating temperature. In the course of
continuous heating, in addition to the approximately linear thermal expansion of
the austenite-martensite structure, the width increased more rapidly in the
temperature range of 550–700 °C, indicating formation of athermal martensite.
Then the width decreased, indicating phase transformation, and finally expanded
approximately linearly as the structure became austenitic. Fig. 4b shows that the
peaks in the curves shifted to higher temperatures with the increasing heating rate.
29
Fig. 4. a) Relative change in width (ΔW/W0) and b) its first derivate (d(ΔW/W0)dT) as a
function of heating temperature at heating rates of 5–150 °C/s for the 0Nb steel. [Paper
III, Published by permission of Trans Tech Publications Inc.]
Fig. 5a shows the formation of new martensite, i.e., athermal martensite, during
heating to 620 °C for 0.1 s. Compared with the value prior to heating, there is about
12% units more α’-martensite. In view of the above results, it was considered
interesting to study whether α’-martensite would continue to form during
isothermal holding after heating. Isothermal martensite was found to form after
heating, e.g., at 550 °C for 100 s. No α’-martensite was detected after annealing at
1100 °C for 1000 s and cooling, suggesting that no α’-martensite is formed during
cooling to room temperature.
Fig. 5b displays a TEM image of α’-martensite present after annealing at
550 °C for 100 s, revealing that the α’-martensite presented before heating has
already tempered. To find out the changes in the dislocation density of α’-
martensite, XRD spectra were subjected to Rietveld analyses using the MAUD
program [88]. The dislocation density of α’-martensite after 60% cold rolling was
found to be 5.04 x 1015 m-2. During heating to 620 °C and quenching, i.e., 0.1 s
annealing, the dislocation density was decreased to 1.55 x 1015 m-2, even though
reversion did not yet occur. After 10, 100 and 300 s soaking at 550 °C, the
dislocation densities were 2.52, 1.41 and 1.03 x 1015 m-2. Hence, the overall
dislocation density of the samples (containing old and newly formed martensite)
decreased over the course of soaking even at this low temperature, i.e., tempering
of α’-martensite took place.
30
Fig. 5. Development of fraction and structure of α’-martensite in the 0Nb steel. a) the
volume fraction of α’-martensite after annealing at various temperatures for various
times and b) TEM micrograph of the specimen after annealing at 550 °C for 100 s. [Paper
III, Published by permission of Trans Tech Publications Inc.]
Fig. 4b shows that the reversion starting and finishing temperatures during
continuous heating depend on the heating rate. The reversion kinetics of α’-
martensite to austenite during isothermal annealing was investigated in Papers I–
VI. The α’-martensite fractions measured after heating to 700 °C at 200 °C/s and
holding for various annealing times are plotted in Fig. 6, which also includes data
for the Nb-alloyed steels (Papers IV and V). According to the figure, reversion of
the 0Nb steel is fast, but it is apparent that Nb alloying retards reversion. At 800 °C
all the α’-martensite reverted back to austenite within 1 s (Papers I, II and VI).
Fig. 6. α’-martensite fractions of the 0–0.45Nb steels after heating at 200 °C/s to 700 °C
and holding for various durations. [Paper V, published by permission of The Iron and
Steel Institute of Japan]
Fig. 7 shows the microstructural evolution during annealing at 700 °C for various
times. The microstructure of the 0Nb steel after 10 s was non-uniform, consisting
of ultrafine reverted austenite grains and large RA grains (Fig. 7a). Within 100 s of
31
annealing, all the α’-martensite was reverted back to austenite (Fig. 7b), but it took
a long time before RA started to recrystallize, with recrystallization being
incomplete after 1000 s (Fig. 7c).
Evidence of both shear-type and diffusion-type reversion mechanisms was
observed, as shown in Figs. 6–7. The reversion was fast, being completed within
10 s in the 0Nb steel and within 100 s in the Nb-alloyed steels (Fig. 6). The EBSD
image quality was low in some areas (examples marked as Shear rev. in Fig. 7a,d)
and new fine grains could hardly be seen after 1000 s, as typical features of shear-
type reversion. However, large areas consisting of new, fine, equiaxed grains were
also visible (examples marked as Diff. rev. in Fig. 7a,d), as typical features of
diffusion-type reversion. Recrystallization was slower in the Nb-alloyed steels, so
that even after 1000 s of soaking, no traces of recrystallized RA grains were seen
yet (Fig. 7e,f).
Fig. 7. Inverse pole figure (a,d) and LOM (b,c,e,f) images after cold rolling and annealing
at 700 °C for the 0Nb (a–c), 0.11Nb (d,e) and 0.45Nb (f) steels for a,d) 10 s, b) 100 s and
c,e,f) 1000 s. [Paper IV, published by permission of Elsevier Ltd.]
An example of TEM examinations of the 0Nb steel showed nucleation of new,
dislocation-free and equiaxed austenite grains at the lath boundaries of the α’-
32
martensite phase when annealed at 700 °C for 100 s, Fig. 8a. The size of reverted
austenite was 300–400 nm, as presented in Fig. 8b.
Fig. 8. Microstructure of the 0Nb steel after annealing at 700 °C for 100 s, showing a)
nucleation of austenite from α’-martensite and b) a reverted equiaxed austenite
structure. [Paper II, Published by permission of AIM Editorial Office]
3.2.2 Recrystallization of retained austenite
Recrystallization of RA required higher temperatures than reversion, as seen in Fig.
9, Papers I, II, IV and V. Nucleation of recrystallization in RA grains tended to
occur at 800 °C, starting at shear bands crossing RA grains, their intersections and
grain boundaries (see white circles in Fig. 9).
Fig. 9. Inverse pole figure images after annealing at 800 °C for 10 s for the a) 0Nb, b)
0.11Nb and c) 0.45Nb steels. [Paper IV, published by permission of Elsevier Ltd.]
Reversion was completed within 1 s at 800 °C, but RA recrystallized later during
1000 s of annealing. The occurrence of two processes at this temperature, i.e.,
33
reversion and recrystallization, resulted in a non-uniform grain size distribution, as
seen in Fig. 10 (Paper IV). In Fig. 10a, fine, reverted areas are shown in red and
coarser recrystallized RA areas in blue. The bimodal grain size distribution was due
to the fact that ultrafine grains are formed from the reverted α’-martensite and
larger grains from RA (Fig. 10b). After annealing at 800 °C for 1000 s, the average
grain sizes were 1.7, 1.1 and 1.0 μm for the 0Nb, 0.11Nb and 0.45Nb steels,
respectively.
Fig. 10. a) Image quality map and b) grain size distribution (taken from the coloured
areas) of the 0Nb steel after annealing at 800 °C for 1000 s. [Paper IV, published by
permission of Elsevier Ltd.]
3.2.3 Austenite grain growth
To obtain a uniform austenite grain size, an annealing temperature of 900 °C or
higher is preferred, as demonstrated in Papers I, II, IV and V. Fig. 11a–c shows an
example, determined using EBSD, of the uniform grain structure of the 0Nb,
0.11Nb and 0.45Nb steels, with average grain sizes of 1.9, 1.4 and 1.4 μm,
respectively. These grain sizes are fine, but significant grain coarsening occurred
during annealing at 1100 °C for 1000 s, as evident in Fig. 11d–f.
34
Fig. 11. Inverse pole figure images of the 0Nb (a,d), 0.11Nb (b,e) and 0.45Nb (c,f) steels
after annealing at 900 °C for 10 s (a–c) and 1100 °C for 1000 s (d–f). [Paper IV, published
by permission of Elsevier Ltd. and Paper V, published by permission of The Iron and
Steel Institute of Japan]
The efficiency of Nb alloying in retarding grain growth was investigated in paper
V. Various grain sizes are achievable via reversion treatments, for instance 2–9 μm
after annealing at 900–1100 °C for 1 s, as shown in Fig. 12, and in Papers I and V.
Grain size naturally increased with increasing temperature or time, but decreased
with increasing Nb content due to retardation of grain coarsening by Nb. According
to Fig. 12, the concentration of 0.11 wt.% Nb was high enough to inhibit grain
growth at 900 and 1000 °C (Fig. 12a,b), but a higher Nb content of 0.28 wt.% was
needed to have an effect at 1100 °C (Fig. 12c). Despite some grain growth,
reversion can be seen to be an effective way to refine grain size, as the initial grain
size of the 0Nb steel before cold rolling was 18 μm.
The kinetics of grain growth can be estimated using the measured grain sizes,
annealing durations and the grain growth equation (Equation 1). Further, the
calculated kinetic parameter can be applied to determine the activation energy for
grain growth (Equation 2). Based on the experimental data, the activation energies
for grain growth in the 0Nb, 0.11Nb and 0.45Nb steels were calculated to be 363,
373 and 458 kJ/mol, respectively.
35
Fig. 12. Average grain size as a function of annealing time for the 0–0.45Nb steels
annealed at a) 900 °C, b) 1000 °C and c) 1100 °C. Note: different scales for grain sizes.
[Paper V, published by permission of The Iron and Steel Institute of Japan]
Grain size is controlled by competition between the driving force (Fd) (Equation 3)
and pinning force (Fp) (Equation 4) for grain growth. If the pinning force exceeds
the driving force, the boundary will be pinned and no grain growth occurs. The
volume fraction of precipitates was estimated from TEM replicas, Paper V. As an
example, Fig. 13 presents typical TEM images of the precipitation structure of the
0.45Nb steel after annealing at 1100 °C for 1 s (Fig. 13a) and 1000 s (Fig. 13b).
The results of the calculations using Equations 3 and 4 after annealing at 1100 °C
for 1 and 1000 s are shown in Fig. 14. It seems that 0.28 wt.% Nb was efficient in
retarding grain coarsening after annealing at 1100 °C, as also observed in grain size
measurements.
36
Fig. 13. TEM extraction replicas of the 0.45Nb steel after annealing at 1000 °C for a) 1 s
and b) 1000 s. [Paper V, published by permission of The Iron and Steel Institute of Japan]
Fig. 14. Estimated driving (Fd) and pinning (Fp) forces of the 0.05, 0.11Nb, 0.28Nb and
0.45Nb steels after annealing at 1100 °C for a) 1 s and b) 1000 s. [Paper V, published by
permission of The Iron and Steel Institute of Japan]
3.3 Nucleation and growth of ε- and α’-martensite in tensile
straining
To clarify the influence of refined grain size on strain-induced martensite
transformation in tensile straining, the formation of ε- and α’-martensite during
straining in various structures, i.e., ultrafine grains (UFG, average grain size 0.5
μm) and retained austenite (RA), micron-scale grains (MG, average grain size 1.5
μm) and RA, fine grains (FG, average grain size 4 μm) and coarse grains (CG,
average grain size 18 μm), was investigated using EBSD in Paper VI. The
structures were obtained by various reversion treatments and α’-martensite
37
fractions were determined as a function of straining by interrupting the tensile test
for the measurement, Fig. 15.
Fig. 15. α’-martensite contents in different structures as a function of tensile strain for
the 0Nb steel. [Paper VI, published by permission of Elsevier Ltd.]
In UFG, no ε-martensite was detected at any stage of deformation, but in MG, FG
and CG, traces of ε-martensite were found after 10% (Fig. 16a) and 20% (Fig. 16b)
strain. In RA grains, ε-martensite appeared already after 2% strain and was found
even after 20% strain (Fig. 16c). α’-martensite formed along shear bands even after
2% of straining in the UFG-RA structure and later in the MG-RA structure.
In UFG, TEM examinations revealed numerous stacking faults (Fig. 16d) and
thin mechanical twins (Fig. 16e) already after 2% strain. After 20% strain, austenite
grains seemed heavily deformed, with shear bands covering the grains (Fig. 16f).
In MG, extended dislocations in one or two slip systems and mechanical twins
were seen after 2% strain (Fig. 16g). At 10% strain, ε- and α’-martensite were
formed in the coarser austenite grains, while shear bands crossed some finer grains
entirely (Fig. 16h).
After 2% strain, extended dislocations and stacking faults were seen in the FG
and CG microstructures (Fig. 16i). After 10% strain in FG, shear bands appeared
in one slip system, and in two slip system in CG (Fig. 16a). Both ε- and α’-
martensite were detected at shear bands in the both structures, but only α’-
martensite at austenite grain boundaries. The α’-martensite nucleated and grew at
the intersection of two ε-martensite laths, shear bands and along austenite grain
boundaries, and sometimes via ε-martensite. Hence, the types of nucleation sites
changed with grain size refinement.
38
Fig. 16. EBSD (a–c) and TEM (d–i) images of various structures with average grain size
(GS) after 2, 10 or 20% straining for the 0Nb steel. [Paper VI, published by permission
of Elsevier Ltd.]
3.4 Mechanical properties
3.4.1 Hardness evolution
To improve our understanding of the occurrence of the reversion and
recrystallization processes required for grain refinement, the softening of the α’-
martensite, reverted grains and RA grains during the course of annealing was
investigated using microhardness (HV0.01) measurements, Fig. 17, Papers I and
IV. After 60% cold rolling, the microhardness of α’-martensite was 665, 746 and
39
777 HV and that of RA grains 626, 622 and 629 HV for the 0Nb, 0.11Nb and
0.45Nb steels, respectively. The hardness of α’-martensite decreased gradually in
the beginning of the annealing process, obviously due to tempering, Fig. 17a. As
shown earlier in Section 3.2.1, pronounced tempering of α’-martensite occurred
already in the heating stage, seen similarly here corresponding to 0.1 s time. At
700 °C, the hardness of reverted grains dropped with continuing annealing time,
especially in the 0Nb steel, where there were no precipitates to hinder the decrease
in dislocation density of shear-reverted grains or the grain coarsening of diffusion-
reverted ultrafine grains. At 700 °C, the hardness of RA grains decreased slowly in
the beginning of the annealing process, Fig. 17b. However, after 100 s, the hardness
of the 0Nb steel dropped, revealing the start of static recrystallization. A similar
drop did not happen in the Nb-alloyed steels, indicating the retarding effect of Nb
alloying on recrystallization kinetics.
At 800 °C, the hardness of reverted grains decreased slowly due to grain
growth, Fig. 17c. In RA grains, the hardness in the 0Nb steel dropped after 10 s of
annealing, and later in the 0.11Nb and 0.45Nb steels, after 40 and 60 s, Fig. 17d.
The data clearly show the retarding effect of Nb alloying on the start of static
recrystallization. Based on the hardness behaviour, it seems that recrystallization
became completed within 20 s in the 0Nb steel, whereas only within 200 s in the
0.45Nb steel. Hence, recrystallization took place distinctly slower than reversion,
especially in the Nb-alloyed steels.
40
Fig. 17. Evolution of microhardness in α’-martensite, reverted austenite (Rev.), RA and
recrystallized austenite (Rex.) in the 0Nb, 0.11Nb and 0.45Nb steels after annealing at
a,b) 700 °C and c,d) 800 °C for various times. [Paper IV, published by permission of
Elsevier Ltd.]
3.4.2 Tensile properties
To understand the dependence of mechanical properties on microstructure, the
mechanical properties of the reversion-treated steels were determined using tensile
tests. The strength and ductility properties were investigated in Papers I, II, IV and
VI. Fig. 18 shows that anisotropy in mechanical properties present after cold rolling
practically disappeared as the grain structure became reverted and recrystallized.
41
Fig. 18. Anisotropy of yield strength in the reversion-treated 0Nb steel containing α’-
martensite, reverted austenite (Rev.), RA and recrystallized austenite (Rex.). LO =
longitudinal direction, TR = transverse direction. [Paper II, Published by permission of
AIM Editorial Office]
Fig. 19 illustrates the true stress-true strain and strain hardening behaviour of the
steels annealed at 700, 800 and 900 °C for various durations. It is seen that after
annealing at 700 °C for 10 s (Fig. 19a), i.e., the microstructure consisting of α’-
martensite, reverted and RA, yield strength was in the range of 1060–1140 MPa,
increasing with Nb content, while uniform elongations (Ag) was 38–42%.
Annealing at 700 °C for a longer time, 1000 s (Fig. 19b), which resulted in reverted
fine-grained austenite and recovered and partially recrystallized RA, caused yield
strength to decrease significantly. Annealing at 800 °C for 10 s (Fig. 19c), i.e., the
similar reverted and recovered, and partially recrystallized retained austenitic
microstructure as above, provided lower yield strength similarly as the uniform
austenitic grain structure obtained at 900 °C for 10 s (Fig. 19d). Table 4 compares
the mechanical properties achieved and average grain size (measured using EBSD),
excluding RA of the studied steels.
For all the steels, strain hardening exhibited two-stage hardening behaviour,
with small strains the strain hardening rate (SHR) dropped steeply, reaching a
minimum at small strains, and then increased to a peak at higher strains before
finally declining.
42
Fig. 19. Strain hardening rate and true stress as a function of true strain for the 0Nb,
0.11Nb and 0.45Nb steels after annealing at a) 700 °C for 10 s, b) 700 °C for 1000 s, c)
800 °C for 10 s and d) 900 °C for 10 s. [Paper IV, published by permission of Elsevier
Ltd.]
Table 4. Mechanical properties (Rp0.2, Rm and Ag) and grain size (GS) of the 0Nb, 0.11Nb
and 0.45Nb steels. The α’-martensite fraction formed during cold rolling is marked as α’
and reversion annealing treatment (annealing temperature – time) as Ann. Some steels
also include RA, marked as *. [Papers IV and VI, Published by permission of Elsevier
Ltd.]
Steel α’
[%]
Ann.
[°C – s]
Rp0.2
[MPa]
Rm
[MPa]
Ag
[%]
GS
[μm]
0Nb 60 700 – 10 1185 1308 43 0.5*
30 700 – 100 899 1345 38 0.5*
60 700 – 1000 605 1180 36 0.8*
60 800 – 10 542 1067 63 0.6*
60 900 – 10 441 1049 59 1.9
0.11Nb 63 800 – 10 813 1117 51 0.6*
900 – 10 564 1005 59 1.4
0.45Nb 49 800 – 10 778 1105 52 0.6*
900 – 10 575 1031 63 1.4
43
4 Discussion
4.1 Anisotropy of mechanical properties
The disadvantage of strengthening by cold rolling is the formation of harmful
anisotropy in mechanical properties, which limits applications. Reversion
annealing and grain size refinement is an alternative strengthening method that
causes less anisotropy in mechanical properties.
In temper rolling of the 0Nb steel, directional anisotropy in yield strength was
already present at such a low thickness reduction as 5%, similarly as also previously
observed for austenitic 301LN stainless steel [6,87]. At this stage, only 2% α’-
martensite is formed in the steel. Hence, α’-martensite cannot be a major reason
for the anisotropy. It has been suggested that anisotropy in mechanical properties
at light reductions results from a dislocation structure formed in the austenite, i.e.,
numerous stacking faults, slip activity in two systems (Fig. 3), twinning (Fig. 16)
and texture, i.e., a mixture of copper, brass and Goss components [89-92].
After reversion treatment, the anisotropy in mechanical properties disappeared
in the structure containing annealed soft grains with low dislocation density, as also
observed by Miettunen [87]. Slight anisotropy was still left in partially softened
structures (Fig. 18), indicating that the anisotropy is related to the dislocation
structure of deformed austenite. It has been observed in 301LN steel [9] that the
texture of reverted fine-grained austenite is strong compared with the typical
texture of commercially cold-rolled and annealed steel, dominated by strong Brass
and Goss orientations and minor Copper and S orientations. Hence, we can notice
that this texture does not seem to be a factor affecting anisotropy. Because almost
full reversion and recrystallization was required to completely eliminate the
anisotropy, maximal strength cannot be utilized; however, grain size after complete
reversion treatment is more than one order of magnitude finer than that of the initial
structure, providing a relatively high yield strength, about 600 MPa.
4.2 ε- and α’-martensite transformation
The formation ε- and α’-martensite during tensile testing and cold rolling was
followed in order to understand the microstructures formed in reversion annealing.
In the 0Nb steel, after 2–20% tensile straining some ε-martensite formed in
structures consisting of grains larger than 4 μm (Paper VI), but after 60% cold
44
rolling no ε-martensite was detected in XRD measurements (Paper IV). This clearly
indicates that ε-martensite is an intermediate phase in α’-martensite nucleation, as
also shown previously, e.g., [27,34,41,43,57]. The dislocation density of the cold-
rolled 0Nb steel was high, 5.04 x 15 m-2, which is consistent with the values
reported in the literature for austenitic stainless steels, such as 304 and 301LN
[62,93]. However, the stability of the studied steels was rather high. Therefore, the
60% cold rolling did not result in complete transformation to α’-martensite. In the
partially transformed structure, α’-martensite reversion can only refine part of the
structure and recrystallization is required for grain size refinement of the cold-
rolled RA. During heating of 17Mn-0.06C TRIP/TWIP steel, ε-martensite has been found
to revert to austenite in the very beginning of the annealing process, over a
temperature range of 100–250 °C [94]. However, upon heating the 0Nb steel
without ε-martensite at the heating rate of 5–150 °C/s, the specimen’s width
increased around the temperature of 550–700 °C (Fig. 4, Paper III), depending on
the heating rate. The formation of athermal bcc-martensite from fcc-austenite
resulted in this volume expansion, as confirmed by α’-martensite measurements,
Fig. 5. However, dislocation density measurements only showed a decrease in the
dislocation density compared with the cold-rolled state, due to the tempering of old
α’-martensite. As the volume fraction of initial α’-martensite was higher than the
fraction of athermally formed new martensite, the overall dislocation density of the
specimen decreased.
During heating at a rate of 5–150 °C/s, reversion occurred over the
temperature range 600–800 °C and was seen as a decrease in the width of the
specimen. In addition to athermal martensite formation during continuous heating,
new martensite formed isothermally during annealing at a constant temperature,
e.g., at 550 °C for 300 s, Fig. 5. The overall dislocation density of the specimen
annealed at 550 °C for 300 s decreased due to the tempering of α’-martensite and
relaxation of microstresses at the γ/α’ interfaces, enabling the growth of new
martensite (Paper III) [73,95]. Another proposed mechanism for the formation of
new martensite in annealing is the formation of precipitates that deplete the matrix
from austenite stabilizer elements, allowing the Ms temperature to increase locally
[9,73,76,95]. However, no evidence of precipitates was seen in the microstructure
(Fig. 5b).
45
4.3 The impact of reversion and recrystallization on
microstructure, grain size and grain growth
Reversion and recrystallization of RA during annealing were investigated in Papers
I–V. Upon annealing at a temperature of 700 °C, α’-martensite reverted back to
austenite. Reversion was so rapid in all the studied steels that within 1 s most of the
α’-martensite disappeared (Fig. 6). The fast reversion suggests that it is of the shear
type (Papers I and IV) [7,8]. Typical shear-reverted, highly dislocated austenite
containing low-angle grain boundaries and having a lower image quality was
visible with the aid of EBSD (Fig. 7). However, also fine, equiaxed austenite grains
with random orientations were observed using EBSD (Fig. 7), suggesting that
diffusion-aided nucleation of new grains and their growth had occurred. Therefore,
it seems that both types of martensitic reversion tended to take place in annealing
at 700 °C. Misra et al., [67] found that the presence of secondary precipitates did
not influence the reversion mechanism in Cr-Ni-N steels. The same reversion
mechanisms also occurred in the present Nb steels, as confirmed in Fig. 7. Nb
alloying distinctly retarded reversion kinetics (Fig. 6). Sadeghpour et al. [96] also
found a retarding effect of Ti alloying on reversion in 201L austenitic stainless steel
and Baghbadorani et al. [12] found Nb-rich carbonitrides in 201 austenitic stainless
steel. The slower reversion kinetics might be a result of fine precipitates whose size
and density are suitable to obstruct the nucleation of new austenite grains in α’-
martensite, even though the driving force for the phase transformation is much
higher than the pinning force created by the precipitates (Papers IV and V).
However, all the α’-martensite reverted back to austenite during heating at 200 °C/s
up to 800 °C (Paper I).
Recrystallization of highly dislocated shear-reverted and retained austenite
grains has been found to take place more slowly than reversion, i.e., at higher
temperatures and over longer times, e.g., [9,74-76,97]. According to hardness
measurements (Fig. 17), Nb alloying retarded the start and progress of
recrystallization of RA, as annealing of 20, 100 and 200 s at 800 °C was required
to recrystallize RA in the 0Nb, 0.11Nb and 0.45Nb steels, respectively (Paper IV).
The retardation in the Nb steels can result from Nb atoms in the solid solution or
Nb(C,N) precipitates, which are effective at pinning grain boundaries [98] and can
even stop recrystallization in microalloyed carbon steels [99].
Recrystallization of RA grains tended to initiate both in deformation bands
traversing large grains and on grain boundaries (Fig. 9), as also reported in the
literature [74]. After reversion of martensitic areas, e.g., after annealing at 700 °C
46
for 100 s, grain size was 300–400 nm (Fig. 8), but during longer annealing or
annealing at higher temperature, where RA recrystallizes, ultrafine grains tended to
grow. As a consequence of these two processes having different kinetics, grain size
distribution became non-uniform in annealing at 800 °C for 1000 s (Fig. 10).
Annealing at 900 °C resulted in quite uniform grain size distribution with a grain
size of 1.4–1.9 μm (Fig. 11a–c), even though grain size was not as fine as when
formed at a lower temperature. Anyhow, annealing for reversion and
recrystallization of RA is an effective way to refine grain size, as uniform grain size
decreased in the 0Nb steel from 18 to 1.9 μm as a result of the treatment.
Grain growth occurred during prolonged annealing at 900 °C or higher, as
shown in Figs. 11d–f and 12, and Paper V. Nb alloying seems to be an effective
means to obstruct grain coarsening, as 0.11 wt.% Nb was enough to retard grain
growth at 1000 °C, and 0.28 wt.% at 1100 °C (Fig. 12). After 1000 s of annealing
at 1100 °C, the grain size of the 0Nb steel was 59 μm but only 10 μm in the 0.28Nb
steel. Therefore, a significantly finer grain size is achievable by alloying with Nb
in the instance of annealing at high temperature for such a long time.
The lowest activation energy for grain growth in the studied steels was
calculated for the 0Nb steel, 363 kJ/mol. The activation energy increased with Nb
alloying up to 458 kJ/mol. This is what we would expect, as precipitates create a
pinning stress on the grain boundaries, hindering their motion and thereby
increasing the activation energy for grain growth (Paper V) [77,100,101].
Activation energy values ranging from 280 to 420 kJ/mol are reported in the
literature for 321 and 347 austenitic stainless steels containing TiC and NbC
carbides, respectively [80,81].
Grain size is controlled by competition between driving (Fd) and pinning (Fp)
forces. By comparing the results predicted by FBM for Fp and Zener’s equation for
Fd, Fp was higher until Nb alloying was at least 0.11 wt.%, suggesting grain growth
in the 0Nb and 0.05Nb steels at 1100 °C (Fig. 14). For the 0.28Nb and 0.45Nb
steels, Fp was higher than Fd and particles retarded grain coarsening. The Fd of the
0.11Nb steel was slightly higher than the Fp and some grain coarsening occurred at
1100 °C, as also observed in Fig. 12. Therefore, 0.28 wt.% Nb seems to be
appropriate for reaching the maximum efficiency of alloying. Similarly, Sawada et
al. [16] found that addition of 0.1 wt.% Nb is sufficient to retard grain coarsening
at 900 °C in 301L austenitic stainless steel.
47
4.4 Nucleation and growth of ε- and α’-martensite during plastic
deformation
Microstructure and austenite grain size affect the nucleation and growth of ε/α’-
martensite during plastic deformation. It was noticed that the amount of α’-
martensite formed in tensile straining was the lowest in the FG (4 μm) and MG-RA
(1.5 μm) microstructures, Fig. 15. This diminishing trend is well reported in the
literature, e.g., [102-108]. It has been argued that the stability is based on the energy
associated with austenite to martensitic transformation [103,108]. Jin et al. [104]
suggested that fine austenite grains have higher strength, which makes the shear
transformation more difficult in them and therefore delays triggering of strain-
induced martensitic transformation. The effective SFE also increases with
decreasing grain size [105,106], which tends to increase the stability of the
austenite phase.
In agreement with the present results, Maréchal [102] showed that at a given
strain, the lowest amount of α’-martensite was formed in the instance of an
austenite grain size of 0.9 μm in 301LN steel. With a larger or finer austenite grain
size, the stability of austenite decreased, promoting martensite formation. In the
tensile tests reported in Paper VI, during 20% straining only 4% α’-martensite was
formed in the micron-scale MG-RA structure (Fig. 15), and in fact, even this α’-
martensite was formed mainly in coarse RA grains (Fig. 16c). On the other hand,
12% α’-martensite was formed in the CG structure. However, in a microstructure
comprising UFG-RA, the α’-martensite fractions were even higher than in the CG
structure, so that 20% α’-martensite was formed in 20% straining. In the UFG-RA
structure, RA grains were still heavily deformed with a high dislocation density, as
evident from the high recorded hardness of 520 HV (Fig. 17), which promoted the
transformation, and therefore α’-martensite was also formed in RA grains (Fig. 16).
Also, in another study on 304L and 316L austenitic stainless steels, it has been
proposed that the initial high dislocation density accelerates the formation of strain-
induced martensite [57].
Somani et al. [9] also observed decreased stability of austenite in a refined grain
structure in 301LN steel. This lower stability was observed and explained by
Maréchal [102]. He showed that in grains smaller than 0.9 μm, martensite nucleates
on grain boundaries and twins, while in coarser grains it nucleates at shear bands
inside the grains. As a result of this change in the mechanism of martensite
nucleation, it can become faster in CG structures as well as ultrafine structures.
48
It seems that in the UFG and MG structures, α’-martensite nucleates straight
from austenite without the intermediate step of ε-martensite formation, but in large
RA grains and FG and CG structures the route can be either straight from austenite
or sometimes via ε-martensite (Fig. 16a–c). As commonly reported in the literature,
the amount of ε-martensite increases with strain up to a maximum, and then
decreases with increasing plastic strain. For instance, Lee et al. [34] reported that
the maximum amount of ε-martensite was obtained at 20% strain, after which the
α’-martensite started to form at the expense of ε-martensite in an Fe-18Cr-10Mn-
N-C alloy.
For the FG, CG and RA structures, various nucleation sites of the α’-martensite
were seen, such as grain boundaries and slip or shear bands, intersection of
deformation bands, including the bundles of stacking faults, dislocation pile-ups,
deformation twins and plates of ε-martensite, some shown in Fig. 16, and also
reported for Cr-Ni alloys [67,62,109].
The incidence of active deformation mechanisms, such as martensite formation,
mechanical twinning and dislocation glide, are dependent on SFE. Curtze et al. [110]
calculated the SFE of the 0Nb steel (204Cu) is 16.8 mJ/m2. Saeed-Akbari et al.
[111] estimated the decrease in effective SFE to be 4 mJ/m2 for an increase in grain
size from 5 to 20 μm. Therefore, in addition to α’-martensite, both twinning and ε-
martensite formation should be possible in the 0Nb steel. TEM examinations
revealed twins and many extended dislocations close to or connected with grain
boundaries in UFG and MG structures (Fig. 16d–f). Misra et al. [10] found that
twinning facilitated nucleation of α’-martensite in 301LN austenitic stainless steel
in 100–500 nm grains by the emission of extended dislocations from grain
boundaries.
4.5 Effect of microstructural evolution on mechanical properties
Grain refinement can be performed by cold rolling followed by subsequent
reversion annealing. Obviously, the finest grain size is obtained by annealing a fully
martensitic structure, e.g., [9-11,16,79,112-115], even though such a situation was
not tested in this work. To keep this refined ultrafine grain size during practical
industrial heat treatments, coarsening of the reverted grains must be prevented or
at least retarded. As discussed above, this can be done by using nano-sized
precipitates.
For the 0Nb steel, the highest yield strength of 1185 MPa and tensile strength
of 1310 MPa with uniform elongation of 43% was achieved after annealing at
49
700 °C for 10 s (Fig. 19, Papers I, II and IV). The high yield strength is due to the
complex microstructure consisting of unreverted tempered α’-martensite, shear-
reverted austenite with a high dislocation density and sub-boundaries, diffusion-
reverted austenite with an ultrafine grain size (hardness of reverted grains 525 HV)
and slightly recovered RA (hardness 520 HV), as discussed in Paper IV. However,
to obtain better ductility, a longer annealing time is preferred.
For the 0Nb steel, after longer annealing at 700 °C, i.e., 100 s, α’-martensite
was reverted and the structure consisted of ultrafine new austenite grains with an
average grain size (measured using LOM) of 0.5 μm and a hardness of 490 HV, and
recovered, partially recrystallized RA grains with a hardness of 515 HV. Since the
hardness of both constituents, UFG and RA, is almost the same, further straining
produces α’-martensite in both constituents. The stability of such a structure is low,
as shown in Fig. 15, promoting α’-martensite formation. Therefore, yield strength
and tensile strengths are high, 900 and 1345 MPa, respectively, and uniform
elongation is 38%.
For the 0Nb steel, features similar to those observed after annealing at 700 °C
for 100 s were obtained after annealing at 800 °C for 10 s, except more recovery
and recrystallization occurred in both reverted (440 HV) and RA (445 HV) grains.
The grain size of the reverted grains increased from 0.5 to 1.5 μm (average grain
size measured using LOM). The stability of such a structure is high due to the
optimal grain size. As discussed above, with smaller or larger grain sizes the
stability of the studied 0Nb steel decreased. Further straining promoted α’-
martensite formation mainly in RA (Fig. 15, Paper VI). Yield strength decreased to
540 MPa and tensile strength to 1065 MPa, but uniform elongation increased to
62%.
A uniform grain structure in the studied steels is achievable with 10 s of
annealing at 900 °C. However, the grain size of reverted grains increases, although
they are still micron-sized (Fig. 12). The stability of these fine grains (3 μm) is as
high as in the above MG-RA structure (Fig. 15) and further straining promotes α’-
martensite formation in the largest grains. For the 0Nb steel, due to the presence of
recrystallized grains, yield strength, tensile strength and uniform elongation are
lower, i.e., 440 MPa, 1050 MPa, and 59% respectively, than in the non-uniform
grain structure (540 MPa, 1065 MPa and 62%, respectively). Yield strength, tensile
strength and uniform elongation are still higher than those of conventional annealed
austenitic stainless steels, i.e., according to the Metals Handbook [116], typical
values are 380 and 760 MPa, and 52% for 201, and 275 and 515 MPa, and 60% for
301.
50
According to the results, for the 0Nb steel the best combination of uniform
elongation and yield strength is achieved with a composite structure (annealed at
800 °C for 10 s) consisting of fine grains and partially recovered RA, i.e., 62% and
540 MPa.
It seems that the highest austenite stability in the 0Nb steel is obtained with a
grain size of 1.5–4 μm (measured using LOM). However, a structure with a uniform
grain size of 1.5 μm (measured using LOM) was not achieved in the 0Nb steel due
to the lack of a fully martensitic structure after cold rolling. However, by changing
the alloy composition, a 100% martensitic structure is achievable, e.g., [9-
11,79,112-115], but the highest austenite stability as a function of grain size may
change.
In Fig. 20, data from the literature has been collected to compare the
mechanical properties achieved in this study with some values reported for some
Cr-Ni- and Cr-Mn-type austenitic stainless steels [4,9-11,79,113,114,117,118]. The
strain rates vary between the steels, which obviously affects the results [62]. It is
seen that Cr-Mn steels generally have a higher yield strength than Cr-Ni steels, but
their ductility varies. The figure also shows that a wide range of mechanical
properties is achievable via reversion treatment by controlling the heat treatment
and thereby the microstructure. Nb alloying itself increases uniform elongation, but
the main difference comes from variation of the microstructure and grain sizes, as
discussed above. LOM images inserted in Fig. 20 display typical microstructures
achieved via the reversion treatment of the studied steel with the corresponding
mechanical properties.
Fig. 20 also shows data from the temper-rolled 0Nb steel [87,119]. It is seen
that reversion is a more effective way to increase the combination of yield strength
and uniform elongation than temper rolling.
51
Fig. 20. Comparison of the mechanical properties of the 0Nb, 0.11Nb and 0.45Nb steels
with the values reported in the literature for Cr-Ni and Cr-Mn steels, strain rates (SR),
and the corresponding LOM images after annealing at a) 700 °C for 100 s for the 0.45Nb
steel (α’+Rev.+RA structure), b) 800 °C for 1 s for the 0.11Nb steel (Rev.+RA structure)
and c) 1000 °C for 10 s for the 0.45Nb steel (Rex. structure). [Paper II, Published by
permission of AIM Editorial Office and Paper IV, published by permission of Elsevier
Ltd.]
The degree of grain size strengthening can be estimated from the Hall-Petch
relationship between yield strength and grain size (Equation 5, Paper IV). Fig. 21
shows the experimental data and the fitted line for the 0Nb steel. The figure also
shows the Hall-Petch plot of the yield strength of fine- and ultrafine-grained Cr-Ni
and Cr-Mn austenitic stainless steels, the data were collected from the literature.
According to Fig. 21, the slope k seems to be 256 MPa μm-0.5 for the Cr-Ni
steels [4,9-11,16,79,117]. This value is slightly higher than the 221 MPa μm-0.5
obtained by Di Schino et al. [120] for 301LN, but much lower than the 495 MPa
μm-0.5 reported by Rajasekahra et al. [4] for 304. However, based on this collected
data, a distinctly higher slope of 367 MPa μm-0.5 is obtained for the Cr-Mn type
steels [97,112-115,118] than for the Cr-Ni steels. The k of 377 MPa μm-0.5 obtained
52
here for the 0Nb steel is in good agreement with this value. This analysis implies
that a high Mn content increases grain size dependence. This is also supported by
the fact that a high value of k, i.e., 350 MPa μm-0.5, has been reported for 22% Mn
TWIP steel [121]. Therefore, we can conclude that grain size strengthening is an
important strengthening mechanism for austenitic stainless steels, even more so in
the case of Cr-Mn steels than Cr-Ni steels.
It is seen that without RA, for the finest grain size of 1.4 μm (Table 4), the
amount of grain size strengthening is 324 MPa, and for a grain size of 1.9 μm the
value is 277 MPa. These high values emphasize the great importance of grain size
refinement by reversion treatment in enhancing yield strength. As seen in Table 4,
after annealing at 900 °C for 10 s, the experimentally measured yield strengths of
564 and 575 MPa for the Nb-alloyed 0.11Nb and 0.45Nb steels were 120–130 MPa
higher than that of the 0Nb steel due to the contribution of precipitates.
Fig. 21. Comparison of Hall-Petch analyses of the 0Nb steel, austenitic Cr-Ni [4,9-
11,16,79,117] and Cr-Mn [97,112-115,118] stainless steels reported in the literature.
[Paper IV, published by permission of Elsevier Ltd.]
4.6 Recommendations for further research
In the present study, the relationship between annealing temperature and time in
microstructure and mechanical properties were studied in 15Cr-9Mn-1Ni-2Cu
austenitic stainless steels, some alloyed with Nb, in order to achieve better
combinations of strength and elongation. However, the current study leaves room
for further investigations as follows:
The reverted and partly recrystallized structure obtained here after annealing
at 700 °C seems to provide an excellent strength-ductility balance. However, to
limit the work involved, only a few experiments were carried out here. Therefore,
53
a more systematic study creating various microstructures consisting of tempered
α’-martensite, reverted ultrafine austenite and recovered and partly recrystallized
RA by annealing at a temperature of around 700–750 °C for various durations, and
the dependence of strength-ductility combination on the fraction and condition of
each microstructural component, should be performed. This could be done using
Nb-free and Nb-alloyed steels.
In Paper II, the anisotropy of mechanical properties after 5, 12, 20 or 60 % cold
rolling of the Nb-free steel was studied and discussed, as also seen in Table 3.
However, the data series did not contain any Nb-alloyed steels. It is assumed that
the behaviour of Nb-alloyed steel is similar to that of the Nb-free steel, but to
confirm this, cold rolling and tensile tests should be performed.
In Paper III, the transformation from α’-martensite to austenite and the
reversion temperature of the Nb-free steel were studied and discussed, as also seen
in Figs. 4 and 5. However, the data series did not contain any Nb-alloyed steels. It
is assumed that the behaviour of Nb-alloyed steel is similar to that of the Nb-free
steel, but to confirm this, dilatometric studies together with XRD analysis should
be performed.
Paper VI discussed the influence of austenite grain size and microstructure on
strain-induced martensite transformation for the 0Nb steel, as also seen in Figs. 14
and 15. To thoroughly understand the differences in the kinetics of ε- and α’-
martensite transformation between the 0Nb and Nb-alloyed steels, i.e., the
nucleation sites of both types of martensite and their growth phenomena,
interrupted tensile tests together with microstructural analysis, including XRD,
EBSD and TEM work, are required. The results would also bring out new
knowledge concerning the different behaviour of the steels during strain hardening
and the effect of Nb alloying.
The corrosion properties of the “200-series” steels limit their applications. It is
possible that sensitization due to precipitation tends to impair the intergranular
corrosion resistance of ultrafine-grained steels. This was not tested in this study.
Therefore, it would be necessary to perform corrosion tests for reversion-treated
steels with a complex microstructure.
In Paper V the retardation of grain growth using Nb alloying was studied. In
order to be able to use these steels in production, optimization of the process
parameter window size of the studied steels using sensitivity analysis could be
useful.
54
55
5 Summary and conclusions
The behaviour of metastable low-Ni high-Mn austenitic stainless steel 15Cr-9Mn-
1Ni-2Cu, with and without Nb, was investigated in reversion annealing to
determine the effect of the thermal history on the microstructure, grain size and
mechanical properties. The aim of this study was to understand the relationship
between the annealing cycle, the resultant microstructure and mechanical
properties in order to achieve enhanced combinations of strength and ductility. The
rate and mechanism of strain-induced martensite formation in different
microstructures with different grain sizes under plastic straining were also studied.
Reversion annealing after cold rolling was conducted on a Gleeble
thermomechanical simulator and the microstructures were analysed using magnetic
measurements, XRD, LOM, SEM-EBSD and TEM. Mechanical properties were
tested using microhardness measurements and tensile tests. The main results and
conclusions can be summarized as follows:
– In temper rolling, directional anisotropy in yield strength is formed in the Nb-
free steel already at 5% thickness reduction. Strength is greater in the
transverse direction than in the longitudinal direction. It was suggested that this
is due to the dislocation structure, i.e., numerous stacking faults and activity in
two slip systems. In reversion treatment, the anisotropy disappeared as the
grain structure became fully softened. Anisotropy was also much reduced in a
structure consisting of reverted grains and recovered RA without any α’-
martensite.
– Upon fast heating, about 12% athermal martensite forms in cold-rolled steel at
a temperature of around 550–700 °C and some isothermal martensite appears
during holding. The dislocation density of α’-martensite decreases
significantly due to tempering before reversion in a temperature range of 600–
800 °C.
– Reversion from α’-martensite to austenite occurs rapidly at 700 °C in all the
studied steels, even though Nb alloying retards the reversion. Features of both
reversion types are present; large austenite grains with low EBSD image
quality and low-angle grain boundaries of the shear type, and fine equiaxed
grains of the diffusion type.
– Recrystallization of shear-reverted areas and retained austenite occurs at a
higher temperature or during longer annealing than is the case for reversion.
56
– At 800 °C the annealed microstructure becomes non-uniform due to the
reversion forming ultrafine grains and recrystallization leading to coarser
grains. Nb alloying retards the recrystallization kinetics of cold-rolled retained
austenite.
– A uniform austenite grain size of 2 μm was obtained on annealing at 900 °C
for 10 s, being slightly finer in the Nb-alloyed steels than in the Nb-free one.
– 0.11 wt.% Nb alloying is sufficient to retard grain coarsening at least up to
1000 s at 1000 °C, but annealing at a higher temperature of 1100 °C requires
0.28 wt.%. This grain growth behaviour can be explained by the driving force
predicted by Zener’s model and the pinning force predicted by the flexible
boundary model.
– A structure consisting of reverted grains and partially recrystallized austenite
possesses significantly enhanced strength with high uniform elongation. Nb
alloying further increases yield strength without reducing uniform elongation.
– The transformation of strain-induced martensite in tensile straining depends on
the original grain size. The stability of austenite is maximal in the instance of
a fine grain size of about 1–4 μm, and stability decreases with an increase or
decrease in grain size.
– The nucleation sites of α’-martensite vary with grain size. In ultrafine grains,
it nucleates on grain boundaries and mechanical twins. In coarser grains, α’-
martensite nucleates straight from austenite or via ε-martensite at grain
boundaries, shear bands and their intersections. In retained austenite, some
deformation remains from cold rolling, so that further straining promotes
martensite formation.
57
6 Novel features
To the best knowledge of the author, the following findings are original to this work:
– New α’-martensite was formed in austenitic stainless steel upon fast,
continuous heating before reversion.
– Both types of reversion mechanisms, shear and diffusion, occurred
simultaneously during annealing at 700 and 800 °C. Nb alloying did not change
the reversion mechanism, but retarded its rate.
– At 800 °C, the annealed microstructure became non-uniform due to the
reversion forming ultrafine grains and recrystallization forming coarser grains,
the effect being more pronounced without Nb alloying.
– The concentration of Nb required to prevent grain growth up to temperatures
as high as 1100 °C was established as 0.28 wt.%. This was also confirmed
using models to predict the driving and pinning forces for grain growth.
– α’-martensite transformation in tensile tests depends in a complex way on
grain size, but is also affected by cold-deformed, retained austenite grains. The
increase in martensite volume fraction with increasing strain is greatest in a
microstructure consisting of both ultrafine and retained austenite grains, being
even faster in coarse-grained steel.
58
59
List of references
1. Charles J, Mithieux JD, Krautschick J, Suutala N, Antonio Simón J, Van Hecke B & Pauly T (2008) A new European 200 series standard to substitute 304 austenitics? In: 6th European stainless steel conference Science and Market. Helsinki, Finland: 427-436.
2. (2005) “New 200-series” steels: An opportunity or a threat to the image of stainless steels. ISSF technical guide. URI: http://www.centroinox.it/sites/default/files/monografiche/200_series_stainless_steels.pdf. Cited 2015/09/16.
3. Reed-Hill RE (1964) Physical metallurgy principles. New Jersey, D. Van Nostrand Company, INC.
4. Rajasekhara S, Ferreira PJ, Karjalinen LP & Kyröläinen A (2007) Hall-Petch in ultra-fine grained AISI 301LN stainless steel. Metall Mater Trans A 38 (6): 1202-1210.
5. Karjalainen LP, Taulavuori T, Sellman M & Kyröläinen A (2008) Some strengthening methods for austenitic stainless steels. Steel Res Int 79 (6): 404-412.
6. Taulavuori T, Aspegren P, Säynäjäkangas J, Salmén J & Karjalainen P (2004) Anisotropic behaviour of the nitrogen alloyed stainless steel grade 1.4318. In: Proceedings of the 7th international conference on high nitrogen steels, Ostend, Belgium: 405-409.
7. Tomimura K, Takaki S, Tanimoto S & Tokunaga Y (1991) Optimal chemical composition in Fe-Cr-Ni alloys for ultra grain refining by reversion from deformation induced martensite. ISIJ Int 31 (7): 721-727.
8. Tomimura K, Takaki S & Tokunaga Y (1991) Reversion mechanism from deformation induced martensite to austenite in metastable austenitic stainless steels. ISIJ Int 31 (12): 1431-1437.
9. Somani MC, Juntunen P, Karjalainen LP, Misra RDK & Kyröläinen A (2009) Enhanced mechanical properties through reversion in metastable austenitic stainless steels. Metall Mater Trans A 40 (3): 729-744.
10. Misra RDK, Nayak S, Mali SA, Shah JS, Somani MC & Karjalainen LP (2009) Microstructure and deformation behavior of phase-reversion-induced nanograined/ultrafine-grained austenitic stainless steel. Metall Mater Trans A 40 (10): 2498-2509.
11. Forouzan F, Najafizadeh A, Kermanpur A, Hedayati A & Surkialiabad R (2010) Production of nano/submicron grained AISI 304L stainless steel through the martensite reversion process. Marter Sci Eng A 527 (27-28): 7334-7339.
12. Baghbadorani HS, Kermanpur A, Najafizadeh A, Bejahti P, Moallemi M & Rezaee A (2015) Influence on Nb-microalloying on the formation of nano/ultrafine-grained microstructure and mechanical properties during martensite reversion process in a 201-type austenitic stainless steel. Metall Mater Trans A 46 (8): 3406-3413.
13. Humphreys FJ & Hatherly M (1995) Recrystallization and related annealing phenomena. Oxford, Pergamon.
60
14. Shirdel M, Mirzadeh H & Parsa MH (2014) Microstructural evolution during normal/abnormal grain growth austenitic stainless steel. Metall Mater Trans A 45 (11): 5185-5193.
15. Kisko A, Talonen J, Porter DA & Karjalainen LP (2015) Effect of Nb microalloying on reversion and grain growth in a high-Mn 204Cu austenitic stainless steel. ISIJ Int 55 (10): 2227-2234.
16. Sawada M, Adachi K & Maeda T (2011) Effect of V, Nb and Ti addition and annealing temperature on microstructure and tensile properties of AISI 301L stainless steel. ISIJ Int 51 (6): 991-998.
17. Schaeffler AL (1949) Constitution diagrams for stainless steel weld metal. Metal Prog 56 (5): 680–680B.
18. Bhadeshia HKDH & Honeycombe RWK (2006) Steels microstructure and properties. Oxford, Butterworth-Heinemann.
19. Meyer L (2001) History of niobium as a microalloying element. In: Niobium Science & Technology Proceedings of the international symposium niobium, Orlando, Florida: 359-378.
20. Cahn RW & Haasen P (1983) Physical metallurgy, part II. 3rd ed. Amsterdam, Elsevier Science Publisher.
21. Olson GB & Cohen M (1972) A mechanism for the strain-induced nucleation of martensitic transformations. J Less-comm met 28(1):107-118.
22. Post J, Datta K & Beyer J (2008) A macroscopic constitutive model for a metastable austenitic stainless steel. Mater Sci Eng A 485 (1-2): 290-298.
23. Feiedel J (1964) Dislocations. Oxford, Pergamon Press. 24. Rémy L & Pineau A (1977) Twinning and strain-induced f.c.c. →h.c.p. transformation
on the mechanical properties of Co-Ni-Cr-Mo alloys. Mater Sci Eng 28 (1): 99-107. 25. Hull D & Bacon DJ (2011) Introduction to dislocations. Oxford, Butterworth-
Heinemann. 26. Brooks JW, Loretto MH & Smallman RE (1979) Direct observations of martensite
nuclei in stainless steel. Acta Metall 27 (12): 1839-1847. 27. Venables JA (1962) The martensite transformation in stainless steel. Phil Mag 7 (73):
35-44. 28. Fujita H & Ueda S (1972) Stacking faults and f.c.c. (γ) →h.c.p. (ε) transformation in
18/8-type stainless steel. Acta Metall 20 (5): 759-767. 29. Brooks JW, Loretto MH & Smallman RE (1979) In situ observations of the formation
of martensite in stainless steel. Acta Metall 27 (12): 1829-1838. 30. Olson GB & Cohen M (1975) Kinetics of strain-induced martensitic nucleation. Metall
Trans A 6 (4): 791-795. 31. Lecroisey F & Pineau A (1972) Martensitic transformations induced by plastic
deformation in the Fe-Ni- Cr-C system. Metall Trans 3 (2): 387-396. 32. Lee T-H, Oh, C-S, Kim S-J & Takaki S (2007) Deformation twinning in high-nitrogen
austenitic stainless steel. Acta Mater 55 (11): 3649-3662.
61
33. Allain S, Chateau J-P, Bouaziz O, Migot S & Guelton N (2004) Correlations between the calculated stacking fault energy and the plasticity mechanisms in Fe-Mn-C alloys. Mater Sci Eng A 387-389: 158-162.
34. Lee T-H, Shin E, Oh C-S, Ha H-Y & Kim S-J (2010) Correlation between stacking fault energy and deformation microstructure in high-interstitial-alloyed austenitic steels. Acta Mater 58 (8): 3173-3186.
35. Lee W-S & Lin C-F (2000) The morphologies and characteristics of impact-induced martensite in 304L stainless steel. Scr Mater 43 (8): 777-782.
36. Gey N, Petit B & Humbert M (2005) Electron Backscatter diffraction study of α/ε’ martensitic variants induced by plastic deformation in 304 stainless steel. Metall Mater Trans A 36 (12): 3291-3299.
37. Bogers AJ & Burgers WG (1964) Partial dislocations on the {110} planes in the B.C.C. lattice and the transition of the F.C.C. into the B.C.C. lattice. Acta Metall 12 (2): 255-261.
38. Lagneborg R (1964) The martensite transformation in 18% Cr-8% Ni steels. Acta Metall 12 (7): 823-843.
39. Reed RP & Guntner CJ (1964) Stress-induced martensitic transformations in 18Cr-8Ni steel. Trans Met Soc AIME 230 (7): 1713-1720.
40. Kelly PM (1965) The martensite transformation in steels with low stacking fault energy. Acta Metall 13 (6): 653-646.
41. Mangonon PL & Thomas G (1970) The martensite phases in 304 stainless steel. Metall Trans 1 (6): 1577-1586.
42. Suzuki T, Kojima H, Suzuki K, Hashimoto T & Ichihara M (1977) An experimental study of the martensite nucleation and growth in 18/8 stainless steel. Acta Metall 25 (10): 1151-1162.
43. Murr LE, Staudhammer KP & Hecker SS (1982) Effect of strain state and strain rate on deformation-induced transformation in 304 stainless steel: Part II. Microstructural study. Metall Trans A 13 (4): 627-635.
44. Choi J-Y & Jin W (1997) Strain-induced martensite formation and itse effect on strain-hardening behaviour in tye cold-drawn 304 austenittic stainless steels. Scr Mater 36 (1): 99-104.
45. Tsakiris V & Edmonds DV (1999) Martensite and deformation twinning in austenitic steels. Mater Sci Eng A 273-275: 430-436.
46. Huang C, Yang G, Gao Y, Wu S & Li S (2007) Investigation on the nucleation mechanism of deformation-induced martensite in an austenitic stainless steel under severe plastic deformation. Jour Mater Res 22 (3): 724-729.
47. Nakada N, Ito H, Matsuoka Y, Tsuchiyama T & Takaki S (2010) Deformation-induced martensitic transformation behaviour in cold-rolled and cold-drawn type 316 stainless steels. Acta Mater 58 (3): 895-903.
48. Dash J & Otte HM (1963) The martensite transformation in steinless steel. Acta Met 11 (10): 1169-1178.
49. Breedis J & Kaufman L (1971) The formation of hcp and bcc phases in austenitic iron alloys. Metall Mater Trans B 2 (9): 2359-2371.
62
50. Narutani T, Olson GB & Cohen M (1982) Constitutive flow relations for austenitic steels during strain-induced martensitic transformation. Jounal de Physique 43 (C4): 429-434.
51. Narutani T (1989) Effect of deformation-induced martensitic transformation on the plastic behaviour of metastable austenitic stainless steel. Mater Trans, JIM, 30 (1): 33-45.
52. Lishtenfeld JA, Mataya MC & van Tyne CJ (2006) Effect of strain rate on stress-strain behaviour of alloy 309 and 304L austenitic stainless steel. Metall Mater Trans A 37 (1): 147-161.
53. Lecroisey F & Pineau A (1972) Martensitic transformations induced by plastic deformation in the Fe-Ni-Cr-C system. Metall Mater Trans B 3 (2): 391-400.
54. Olson GB & Cohen M (1976) A general mechanism of martensitic nucleation. Part II: fcc→bcc and other martensitic transformations. Metall Mater Trans A 7 (11): 1905-1914.
55. Guimaraes J & De Oliveira SF (1979) Work-hardening and martensitic transformation in Fe-27%Ni-0.3%C at 263 K. Scr Met 13 (7): 537-542.
56. Yang H-S & Bhadeshia H (2009) Austenite grain size and the martensite-start temperature. Scr Mater 60 (7): 493-495.
57. Spencer K, Embury JD, Conlon KT, Véron M & Bréchet Y. Strengthening via the formation of strain-induced martesite in stainless steels (2004) Mater Sci Eng A 387-389: 873-881.
58. Staudhammer KP, Murr LE & Hecker SS (1983) Nucleation and evolution of strain-induced martensitic (B.C.C.) embyros and substructure in stainless steel: A transmission electron microscopy study. Acta Metall 31 (2): 267-274.
59. Gunter CJ & Reed RP (1962) The effect of experimental variables including the martensitic transformation on the low-temperature mechanical properties of austenitic stainless steels. Transactions of the ASM 55 (3): 399-419.
60. Huang GD, Matlock DK & Krauss G (1989) Martensite formation, strain rate sensitivity, and deformation behaviour of type 304 stainless steel sheet. Metall Trans A 20 (7): 1239-1246.
61. Fang XF & Dahl W (1991) Strain hardening and transformation mechanism of deformation-induced martensite transformation in metastable austenitic stainless steels. Mater Sci Eng A 141 (2): 189-198.
62. Talonen J (2007) Effect of strain-induced α’-martensite transformation on mechanical properties of metastable austenitic stainless steels. PhD thesis. Helsinki, Helsinki university of Technology.
63. Powell GW, Marshall ER & Backofen WA (1958) Strain hardening of austenitic stainless steel. Transactions of the ASM 50 (1): 478-497.
64. Sanderson GP & Llewellyn DT (1969) Mechanical properties of standard austenitic stainless steels in the temperature range -196 to +800°C. Journal of the iron and steel institute 207 (8): 438-445.
65. Kraus Jr. G & Cohen M (1962) Strengthening and annealing of austenite formed by reverse martensitic transformation. Trans Met Soc AIME 224: 121-1221.
63
66. Lee S-J, Park Y-M & Lee Y-K (2009) Reverse transformation mechanism of martensite to austenite in a metastable austenitic alloy. Mater Sci Eng A 515 (1-2): 32-37.
67. Misra RDK, Zhang Z, Venkatasurya PKC, Somani MC & Karjalainen LP (2011) The effect of nitrogen on the formation of phase reversion-induced nanograined/ultrafine-grained structure and mechanical behavior of a Cr-Ni-N steel. Mater Sci Eng A 528 (3): 1889-1896.
68. Apple CA & Krauss G (1972) The effect of heating rate on the martensite to austenite transformation in Fe-Ni-C alloys. Acta Metall 20 (7): 849-856.
69. Montanari R (1990) Reversion of α’ martensite in 304 steel: a diffusion controlled stage. Z für Metall (Zeitschrift fuer Metallkunde) 21 (11): 114-118.
70. Leem D-S, Lee Y-D, Jun J-H & Choi C-S (2001) Amount of retained austenite at room temperature after reverse transformation of martensite to austenite in an Fe-13%Cr-7%Ni-3%Si martensitic stainless steel. Scr Mater 45 (7): 767-772.
71. Kaufman L, Clougherty EV & Weiss RJ (1963) The lattice stability of metals–III. Iron. Acta Metall 11 (5): 323-335.
72. Somani M. Karjalainen P, Juntunen P, Rajasekhara S, Ferreira P, Kyröläinen A, Taulavuori T & Aspegren P (2005) Submicron microstructure and mechanical properties achieved in a short annealing of cold-rolled austenitic stainless steels. Iron and Steel 40: 283-289.
73. Padilha AF, Plaut RL & Rios PR (2003) Annealing of cold-worked austenitic stainless steels. ISIJ Int 43 (2): 135-143.
74. Poulon-Quintin A, Brochet S, Vogt J-B, Glez J-C & Mithieux J-D (2009) Fine grained austenitic stainless steels: The role of strain inducedα’ martensite and the reversion mechanism limitations. ISIJ Int 49 (2): 293-301.
75. Misra RDK, Nayak S, Venkatasurya PKC, Ramuni V, Somani MC & Karjalainen LP (2010) Nanograined/ultrafine-grained structure and tensile deformation behaviour of shear phase reversion-induced 3011 austenitic stainless steel. Metall Mater Trans 41A (8): 2162-2174.
76. Behjati P, Kermanpur A, Karjalainen LP, Järvenpää A, Jaskari M, Samaei Bahdbadorani, Najafizadeh A & Hamada A (2015) Influence of prior cold rolling reduction on microstructure and mechanical properties of a reversion annealing high-Mn austenitic steel. Mater Sci Eng A 650 (5): 119-128.
77. Hiller M (1965) On the theory of normal and abnormal grain growth. Acta Mater 13 (3): 227-238.
78. Rajasekhara S & Ferreira PJ (2011) Martensite → austenite phase transformation kinetics in an ultrafine-grained metastable austenitic stainless steel. Acta Mater 59 (2): 738-748.
79. Sabooni F, Karimzadeh F & Enayati MH (2014) Thermal stability study of ultrafine grained 304L stainless steel produced by martensitic process. J Mater Eng Perform 23 (5): 1665-1672.
80. Sanley JK & Perrotta AJ (1969) Grain growth in austenitic stainless steels. Metallography 2 (4): 349-362.
64
81. German RM (1978) Grain growth in austenitic stainless steels. Metallography 11 (2): 235-239.
82. Gladman T (2004) Grain size control. London, Maney. 83. Cuddy LJ (1982) The effect of microalloying concentration on the recrystallization of
austenite during hot deformation. In: Proceedings of international conference on thermomechanical processing of microalloyed austenite TMS of AIME. Warrendale, USA: 129-140.
84. Johanssen DL, Kyrolainen A & Ferreira PJ (2006) Influence of annealing treatment on the formation of nano/submicron grain size AISI 301 austenitic stainless steels. Metall Mater Trans A 37 (8): 2325-2338.
85. Padilha AF & Rios PR (2002) Decomposition of austenite in austenitic stainless steels. ISIJ Int 42 (4): 325-337.
86. Talonen J, Aspegren P & Hänninen H (2004) Comparison of different methods for measuring strain induced α´-martensite content in austenitic steels. Mater Sci Tech 20 (12): 1506-1512.
87. Miettunen I (2010) The anisotropic behaviour of mechanical properties in high strength austenitic stainless steels. Master’s theses. Oulu, University of Oulu.
88. Popa NC (1998) The (hkl) dependence of diffraction-line broadening caused by strain and size for all Laue groups in Rietveld refinement. J Appl Cryst 31 (2): 176-180.
89. Ravi Kumar B, Singh AK, Samar Das & Bhattacharya DK (2004) Cold rolling texture in AISI 304 stainless steel. Mater Sci Eng A 364 (1-2): 132-139.
90. Chowdhury SG, Das S & De PK (2005) Cold rolling behaviour and textural evolution in AISI 316L austenitic stainless steel. Acta Mater 53 (14): 3951-3959.
91. Ravi Kumar B, Mahato N, Bandyopadhyay, NR & Bhattacharya DK (2005) Comparison of rolling texture in low and medium stacking faults energy austenitic stainless steels. Mater Sci Eng A 394 (1-2): 296-301.
92. Lü Y, Hutchinson B, Molodov DA & Gottstein G (2010) Effect of deformation and annealing on the formation and reversion of ε-martensite in an Fe-Mn-C alloy. Acta Mater 58 (8): 3079-3090.
93. Shintani T & Murata Y (2011) Evaluation of the dislocation density and dislocation character in cold rolled Type 304 steel determined by profile analysis of X-ray diffraction. Acta Mater 59 (11): 4314-4322.
94. Escobar DP, Ferreira de Dafé SS & Santos DB (2015) Martensite reversion and texture formation in 17Mn-0.6C TRIP/TWIP steel after hot cold rolling and annealing. J Mater Res Technol 4 (2): 162-170.
95. Guy KB, Butler EP & West DRF (1983) Reversion of bcc α’martensite in Fe-Cr-Ni austenitic stainless steels. Metal Sci 17 (4): 167-176.
96. Sadeghpour S, Kermanpur A & Najafizadeh A (2013) Influence of TI microalloying on the formation of nanocrystalline structure in the 201L austenitic stainless steel during martensite thermomechanical treatment. Mater Sci Eng A 584: 177-183.
65
97. Kisko A, Hamada A, Karjalainen LP & Talonen J (2011) Microstructure and mechanical properties of reversion treated high Mn austenitic 204Cu and 201 stainless steel. The 1st International Conference on High Manganese Steels. Seoul, Korea, 15-18 May. Research Institute of Iron and Steel Technology, Yonsei University, B-19.
98. Hutchinson CR, Zurob HS, Sinclair CW & Brechet YJM (2008) The comparative effectiveness of Nb solute and NbC precipitates at impeding grain-boundary motion in Nb steels. Scr Mater 59 (6): 653-637.
99. Gómez M, Medina SF & Valles P (2005) Determination of driving and pinning forces for static recrystallization during hot rolling of a niobium microalloyed steel. ISIJ Int 45 (11): 1711-1720.
100. Spruiell JE, Scott JA, Ary CS & Hardin RL (1973) Microstructural stability of thee thermal-mechanically pretreated type 316 austenitic stainless steel. Metall Mater Trans B 4 (6): 1533-1544.
101. Ashby MF & Ebeling R (1966) On the determination of the number, size, spacing and volume fraction of spherical second-phase particles from extraction replicas. Trans Met Soc AIME 236: 1396-1404.
102. Maréchal D (2001) Linkage between mechanical properties and phase transformations in a 301LN austenitic stainless steel. PhD thesis. Vancouver, University of British Columbia.
103. Takaki S, Fukanaga K, Syarif J & Tsuchiyama T (2004) Effect of grain refinement on thermal stability of metastable austenitic steel. Mater Trans 45 (7): 2245-2251.
104. Jin J-E, Jung Y-S & Lee Y-K (2007) Effect of grain size on the uniform ductility of a bulk ultrafine-grained alloy. Mater Sci Eng A 449-451: 786-789.
105. Jun J-H & Choi C-S (1998) Variation of stacking fault energy with austenite grain size and its effect on the Ms temperature of γ→ε martensitic transformation in Fe-Mn alloy. Mater Sci Eng A 257 (2): 353-356.
106. Lee Y-K & Choi C-S (2000) Driving force for γ→ε martensitic transformation and stacking fault energy of γ in Fe-Mn binary system. Metall Mater Trans A 31 (2): 355-360.
107. Jung Y-S, Lee Y-K, Matlock DK & Mataya MC (2011) Effect of grain size on strain-induced martensitic transformation start temperature in an ultrafine grained metastable austenitic steel. Metal Mater Int 17 (4): 553-556.
108. Matsuoka Y, Iwasaki T, Nakada N, Tshuchiyama T & Takaki S (2013) Effect of grain size on thermal and mechanical stability of austenite in metastable austenitic stainless steel. ISIJ Int 53 (7): 1224-1230.
109. Hedström P, Lienert U, Almer J & Odén M (2007) Stepwise transformation behavior of the strain-induced martensitic transformation in a metastable stainless steel. Scr Mater 56 (3): 213-216.
110. Curtze S, Kuokkala V.T, Oikari A, Talonen J & Hänninen H (2011) Thermodynamic modeling of the stacking fault energy of austenitic steels. Acta Mater 59 (3): 1068-1076.
111. Saeed-Akbari A, Imlau J, Prahl U & Bleck W (2009) Derivation and variation in composition-dependent stacking fault energy maps based on subregular solution model in high-manganese steels. Metall Mater Trans 40 (13): 3076-3090.
66
112. Moallemi M, Najafizadeh A, Kermanpur A & Rezaee A (2011) Effect of reversion annealing on the formation of nano/ultrafine grained structure in 201 austenitic stainless steel. Mater Sci Eng A530: 378-381.
113. Rezaee A, Kermanpur A, Najafizadeh A & Moallemi M (2011) Production of nano/ultrafine grained AISI 201L stainless steel through advanced thermo-mechanical treatment. Mater Sci Eng A 528 (15): 5025-5029.
114. Behjati P, Kermanpur A, Najafizadeh A & Baghbadorani HS (2014) Effect of annealing temperature on nano/ultrafine grain of Ni-free austenitic stainless steel. Mater Sci Eng A 592: 77-82.
115. Baghbadorani HS, Kermanpur A, Najafizadeh A, Bejahti P, Rezaee A & Moallemi M (2015) An investigation on microstructure and mechanical properties of a Nb-microalloyed nano/ultrafine grained 201 austenitic stainless steels. Mater Sci Eng A 636: 593-599.
116. Davis JR (1998) Metals Handbook desk edition. ASM International. 117. Somani MC, Karjalainen LP, Misra RDK, Juntunen P & Kyröläinen A (2008)
Reversion transformation of cold rolled martensite to austenite in EN 1.4318 metastable austenitic steel; microstructures and mechanisms. In: 6th European stainless steel conference Science and Market. Helsinki, Finland: 519-524.
118. Hamada AS, Kisko AP, Sahu P & Karjalainen LP (2015) Enhancement of mechanical properties of a TRIP-aided austenitic stainless steel by controlled reversion annealing. Mater Sci Eng A 628: 154-159.
119. Kisko A, Rovatti L, Miettunen I, Karjalainen P & Talonen J (2011) Microstructure and anisotropy of mechanical properties in reversion-treated high-Mn type 204Cu and 201 stainless steels. 7th European Stainless Steel Conference Science and Market. Como, Italy, 21-23 September.
120. Di Schino A & Kenny JM (2003) Grain size dependence of the fatigue behaviour of a ultrafine-grained AISI 304 stainless steel. Mater Lett 57 (21): 3182-3185.
121. De las Cuevas F, Reis M, Ferraiuolo A, Pratolongo G, Karjalainen LP, Alkorta J & Gil Sevillano J (2010) Hall-Petch relationship of a TWIP steel. Key Eng Mater 423: 147-152.
67
Original publications
I Kisko A, Hamada A, Karjalainen LP & Talonen J (2011) Microstructure and mechanical properties of reversion treated high Mn austenitic 204Cu and 201 stainless steels. The 1st International Conference on High Manganese Steels. Seoul, Korea, 15-18 May. Research Institute of Iron and Steel Technology, Yonsei University, B-19.
II Kisko A, Rovatti L, Miettunen I, Karjalainen P & Talonen J (2011) Microstructure and anisotropy of mechanical properties in reversion-treated high-Mn type 204Cu and 201 stainless steels. 7th European Stainless Steel Conference Science and Market. Como, Italy, 21-23 September.
III Kisko A, Rovatti L, Misra D, Sahu P, Talonen J & Karjalainen P (2013) Studies on martensite transformation in a metastable austenitic Cr-Mn stainless steel. Materials Science Forum 762: 424-430.
IV Kisko A, Hamada A.S, Talonen J, Porter D & Karjalainen LP (2016) Effects of reversion and recrystallization on microstructure and mechanical properties of Nb-alloyed low-Ni high-Mn austenitic stainless steels. Materials Science & Engineering A 657: 359-370.
V Kisko A, Talonen J, Porter D & Karjalainen LP (2015) Effect of Nb microalloying on reversion and grain growth in a high-Mn 204Cu austenitic stainless steel. ISIJ International 55(10): 2227-2234.
VI Kisko A, Misra D, Talonen J & Karjalainen LP (2013) The influence of grain size on the strain-induced martensite formation in tensile straining of an austenitic 15Cr-9Mn-Ni-Cu stainless steel. Materials Science & Engineering A 578: 408-416.
Reprinted with permission from Yonsei University, Seoul, Korea (I), AIM Editorial
Office (II), Trans Tech Publications Inc. (III), Elsevier Ltd. (IV), The Iron and Steel
Institute of Japan (V) and Elsevier Ltd. (VI).
Original publications are not included in the electronic version of the dissertation.
68
A C T A U N I V E R S I T A T I S O U L U E N S I S
Book orders:Granum: Virtual book storehttp://granum.uta.fi/granum/
S E R I E S C T E C H N I C A
552. Karjalainen, Mikko (2015) Studies on wheat straw pulp fractionation :fractionation tendency of cells in pressure screening, hydrocyclone fractionationand flotation
553. Nelo, Mikko (2015) Inks based on inorganic nanomaterials for printed electronicsapplications
554. Kursu, Olli-Erkki (2015) Micromotion compensation and a neural recording andstimulation system for electrophysiological measurements
555. Hallman, Lauri (2015) Single photon detection based devices and techniques forpulsed time-of-flight applications
556. Omran, Mamdouh (2015) Microwave dephosphorisation of high phosphorus ironores of the Aswan region, Egypt : developing a novel process for high phosphorusiron ore utilization
557. Kiljander, Jussi (2016) Semantic interoperability framework for smart spaces
558. Aula, Matti (2016) Optical emission from electric arc furnaces
559. Ferdinand, Nuwan Suresh (2016) Low complexity lattice codes forcommunication networks
560. Xue, Qiang (2016) Analysis of near-optimal relaying schemes for wireless tandemand multicast relay networks
561. Rautio, Anne-Riikka (2016) On the stability of carbon nanotube and titaniananowire based catalyst materials : from synthesis to applications
562. Kuokkanen, Ville (2016) Utilization of electrocoagulation for water andwastewater treatment and nutrient recovery : techno-economic studies
563. Huusko, Jarkko (2016) Communication performance prediction and linkadaptation based on a statistical radio channel model
564. Nguyen, Vu Thuy Dan (2016) Transmission strategies for full-duplex multiuserMIMO communications systems
565. Keränen, Pekka (2016) High precision time-to-digital converters for applicationsrequiring a wide measurement range
566. Koivuranta, Elisa (2016) Optical monitoring of flocs and filaments in the activatedsludge process
567. Lohikoski, Päivi (2016) Information processing in global virtual NPD projects
568. Kauppila, Osmo (2016) Integrated quality evaluation in higher education
C569etukansi.kesken.fm Page 2 Wednesday, April 27, 2016 2:24 PM
UNIVERSITY OF OULU P .O. Box 8000 F I -90014 UNIVERSITY OF OULU FINLAND
A C T A U N I V E R S I T A T I S O U L U E N S I S
Professor Esa Hohtola
University Lecturer Santeri Palviainen
Postdoctoral research fellow Sanna Taskila
Professor Olli Vuolteenaho
University Lecturer Veli-Matti Ulvinen
Director Sinikka Eskelinen
Professor Jari Juga
University Lecturer Anu Soikkeli
Professor Olli Vuolteenaho
Publications Editor Kirsti Nurkkala
ISBN 978-952-62-1214-2 (Paperback)ISBN 978-952-62-1215-9 (PDF)ISSN 0355-3213 (Print)ISSN 1796-2226 (Online)
U N I V E R S I TAT I S O U L U E N S I SACTAC
TECHNICA
U N I V E R S I TAT I S O U L U E N S I SACTAC
TECHNICA
OULU 2016
C 569
Anna Kisko
MICROSTRUCTURE AND PROPERTIES OF REVERSION TREATED LOW-Ni HIGH-Mn AUSTENITIC STAINLESS STEELS
UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF TECHNOLOGY
C 569
ACTA
Anna K
iskoC569etukansi.kesken.fm Page 1 Wednesday, April 27, 2016 2:24 PM
