Investigation of the Effect of Different

“Q&P” Parameters on the Mechanical

Properties of AHSS

Borasi Luciano

Mechanical Engineering, master's level (120 credits)

2018

Luleå University of Technology

Department of Engineering Sciences and Mathematics
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Acknowledgments
I would like to express my gratitude to my supervisors Ms. Farnoosh Forouzan
and Prof. Esa Vuorinen for their support and guidance, which made this thesis
possible. In addition, discussions with my colleagues during meetings at the
Department of Materials Science at LTU are greatly appreciated.

I would like to extend my gratitude to Johnny Grahn and Lars Frisk for their
daily assistance to carry out experiments.

I gratefully acknowledge the financial support from European School of

Materials (EUSMAT) in order to study the Advanced Materials Science and
Engineering Master Programme.

Last but not least, I would like to thank my family, girlfriend and friends. Even
though most of them are physically far away, their support, chats and daily
encouragement to keep focusing on what I like have been invaluable to the
culmination of my studies.

i
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Abstract
In the present study, the influence of the quenching temperature and
partitioning conditions (temperature and time) have been investigated on a
0.6%C-1.2%Mn-1.6%Si-1.75%Cr alloy. Maps of hardness, impact toughness and
amount of retained austenite have been developed for three quenching
temperatures as a function of partitioning temperature and partitioning time.
Results demonstrate that, in this material, the carbon depletion of the
martensite and the stabilization of austenite can be achieved significantly faster
at high partitioning temperatures, promoting higher retained austenite
fractions, lower hardness, and maximizing the energy absorbed in a
Charpy V-Notch test. In addition, the effect of the partitioning time was also
analysed, presenting different behaviour at high and low partitioning
temperatures. Whereas an increment of time at high partitioning temperatures
(>400 ºC) leads to an austenite consumption, at low partitioning temperatures it
is effective to retain a higher amount of austenite.

Furthermore, tensile properties are shown to be better than in conventional

alloys utilized in industry. Whilst, for example, the AISI 52100 alloy achieves
~2 GPa of tensile strength and 1-2% of fracture deformation, in the present
study the notable combination of ~2.5 GPa of tensile strength and 5.7 % of
fracture deformation was achieved in samples quenched until room
temperature. Untempered martensite transformed during final cooling in
samples quenched until higher temperatures was shown to be detrimental for
tensile properties.

A comparison between the Q&P process and the austempering process on this
alloy has been carried out. Results reveal that the quenching and partitioning
heat treatment is presented as a promising alternative to reach higher hardness
(>700 HV) and similar specific wear rates in dry conditions performing a
shorter heat treatment.

Finally, a complementary study about the effect of micro-segregation on the

Q&P process and an optimization method to minimize the inhomogeneity of
the structure by a correct selection of the quenching temperature were
established.

ii
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Content
Acknowledgments .......................................................................................................... i
Abstract ........................................................................................................................... ii
1. Literature Review .................................................................................................. 1
1.1 Introduction to Advanced High Strength Steels (AHSS)............................... 1
1.2 Introduction to Quenching and Partitioning ................................................... 3
1.3 Design of the Q&P Heat Treatment .................................................................. 5
1.4 Stability of Retained Austenite .......................................................................... 8
1.5 Chemical Composition: Consideration in Q&P Process .............................. 10
1.6 Microstructure of Q&P Steel and Competing Reactions ............................. 12
1.7 Mechanical Properties of Steels Treated by Q&P ......................................... 14
1.8 A Brief Introduction to Carbide Free Bainitic Steels..................................... 16
1.8.1 Wear Performance in CFB steels .................................................................. 21
1.9 A Comparison between CFB and Q&P .......................................................... 23
2. Objectives .............................................................................................................. 25
3. Materials and Methods ........................................................................................... 26
3.1 Base Metal ........................................................................................................... 26
3.2 Heat Treatment and Samples Nomenclature ................................................ 27
3.3 Metallography – Microhardness and Impact Toughness ............................ 30
3.4 XRD Analysis ..................................................................................................... 32
3.5 Tensile Test ......................................................................................................... 33
3.6 Rolling/Sliding Wear Test ............................................................................... 34
4. Results ....................................................................................................................... 35
4.1 Hardness ............................................................................................................. 35
4.2 Impact Test ......................................................................................................... 36
4.3 X-Ray Diffraction ............................................................................................... 37
4.4 Tensile Test ......................................................................................................... 40
4.4.1 Fracture Analysis ........................................................................................ 41
4.5 Rolling/Sliding Wear Test ............................................................................... 43
5. Discussion ................................................................................................................. 45
5.1 Hardness-XRD-Impact Toughness ................................................................. 45

iii
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

5.2 Tensile Test ......................................................................................................... 48

5.3 Rolling/Sliding Wear Test ............................................................................... 48
6. Conclusions .............................................................................................................. 50
6.1 Comparison between CFB and Q&P .............................................................. 51
7. Future Work ............................................................................................................. 52
8. References ................................................................................................................. 53
Appendix A: An Optimization of Quenching Temperature to Minimize the
Banding Phenomena in Q&P Steels .......................................................................... 59
Introduction .............................................................................................................. 59
Materials and Methods ........................................................................................... 59
Results ....................................................................................................................... 60
Discussion ................................................................................................................. 64
Conclusions .............................................................................................................. 66
References ................................................................................................................. 67
Appendix B – Problems with Tensile Test Specimens ........................................... 68

iv
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 1. Different stress-strain curves for different steels and their application in
the automobile body structure according to their mechanical properties [1]. ..... 2
Fig. 2. Combinations of tensile strength and elongation for different generations
of steels [1]. ..................................................................................................................... 3
Fig. 3. Schematic Temperature vs Time representation of the quenching and
partitioning thermal cycle with the microstructure evolution [5]. ......................... 4
Fig. 4. A theoretical diagram where the fraction of different phases and the
carbon content in austenite are represented as a function of the quenching
temperature in a 0.25%C-1.58%Si-1.24%Mn-1.7%Cr-0.14Mo-0.11V steel. ............ 6
Fig. 5. Comparison between the retained austenite fraction measured by X-ray
diffraction analysis after 10 s at 500 ºC and what is predicted according to the
theoretical model in a 0.6%C–0.95%Mn–1.96%Si (wt.%) steel [8]. ......................... 7
Fig. 6. Austenite Phase Fraction expected as a function of Quench Temperatures
for steels with different carbon content [8]. ............................................................... 7
Fig. 7. Evolution of blocky austenite (left-bright field) and film austenite (right-
dark field) in a 0.22C–1.40Si–1.8Mn (wt.%) steel as a function of strain a-b) 0% c-
d)2% e-f)12% [14]. ........................................................................................................ 10
Fig. 8. Schematic C-curve emphasising the delaying effect of silicon on the
kinetic of carbides formation [8]. ............................................................................... 11
Fig. 9. Microstructure observed by EBSD analysis on a 0.2%C-1.5Si-2.5%Mn-
1.47%Ni-1%Cr steel a) Combination of band contrast map and colour-coded
phase map where blue corresponds to martensite (bcc) and red corresponds to
retained austenite (fcc)-Darker bands are associated with fresh martensite. b)
Combination of inverse pole-figure map and band contrast map on fcc- Bright
areas correspond to martensite, dark areas are associated with fresh martensite
and austenite is shown in colours depending on its orientation [20]. ................. 12
Fig. 10. Microstructure evolution of a 0.2 %C-1.5 %Si-3.5 %Mn steel; quenching
at 240 ºC, partitioning for 1000 s at 350 ºC- M refers to martensite, RA denotes
retained austenite and F and TM indicate the presence of ferrite and tempered
martensite respectively. a)Initial microstructure b) Analysis after austenitization
at 770 ºC c) Final microstructure after Q&P [16]. .................................................... 12
Fig. 11. Analysis of volume fraction of austenite as a function of partitioning
time for quenching temperatures of 180 ºC (above) and 150 ºC (below). The
presence of a second peak on the austenite volume fraction was associated with
a re-solution of transitional carbides [8]. .................................................................. 13
Fig. 12. Presence of a swing-back phenomenon (acceleration of austenite
decomposition close to Ms temperature) on two different steels [22].................. 14
Fig. 13. Combination of total elongation (%) and Ultimate Tensile Strength for
different steel grades. TRIP: transformation induce plasticity, DP: dual phase,
M: Martensitic, Q&P: quenching and partitioning [25]. ........................................ 15

v
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 14. Comparison of the Engineering Stress-Strain Curve of a 0.20C–1.57Mn–

1.55Si (wt.%) steel treated by Q&T (QT=25 ºC, Tempering Temperature=400 ºC)
and Q&P (QT=300 ºC, PT=400 ºC). A substantial different of hardening after
1.9% strain can be observed [29]................................................................................ 16
Fig. 15. Schematic illustration of an austempering thermal cycle. ....................... 17
Fig. 16. Effects of austempering temperature on the microstructure and
mechanical properties of CFB steels. Left) Effect of transformation temperature
on the bainite thickness and hardness [38]. Right) Stress-Strain curves of an
alloy isothermally transformed at different temperatures [42]. ........................... 18
Fig. 17. Schematic illustration of To curve on the phase diagram. Free energy of
both austenite (γ) and ferrite (α) at a certain temperature (T1) and as a function
of carbon concentration. ............................................................................................. 20
Fig. 18. Specific wear rate (SWR) as a function of hardness (HV) of different
alloys and temperature of austempering. In red it is highlighted the behaviour
of CFB alloys [37]. ........................................................................................................ 22
Fig. 19. a) Hardness profile from worn surface for different test duration. b)
Hardened layer extension as a function of the test cycles [49]. ............................ 23
Fig. 20. Phase diagram of 06CV simulated in ThermoCalc. .................................. 26
Fig. 21. The fully perlitic microstructure of the material before Q&P (base
metal). Nital etching. ................................................................................................... 27
Fig. 22. Schematic illustration of the Q&P conditions evaluated. PT and Pt refer
to partitioning temperature and partitioning time respectively........................... 27
Fig. 23. Nomenclature of the samples. .................................................................... 28
Fig. 24. Configuration of the furnace and two salt baths utilized to perform heat
treatments. .................................................................................................................... 28
Fig. 25. Real Temperature-Time profile measured in samples during heat
treatment. Quenching until 165 ⁰C and partitioning at 400 ⁰C. ............................ 29
Fig. 26. Hardness (HV0.1) as a function of the distance from the surface (µm). A
drop in hardness close to the surface is associated with a decarburization
during the heat treatment. .......................................................................................... 30
Fig. 27. Charpy V-Notched dimensions according to ASTM Standard E 23-
“Standard Test Methods for Notched Bar Impact Testing of Metallic Materials”.
........................................................................................................................................ 32
Fig. 28. Shape and layout of tensile test samples a) As-machined b) After heat
treatment c) After grinding d) Layout of specimens. ............................................. 34
Fig. 29. Disposition of the tensile test specimen and extensometer in Gleebe
3800. ............................................................................................................................... 34
Fig. 30. a) Disc of 44.6mm diameter tested in dry rolling/sliding condition. b)
Schematic illustration of the set-up utilized to perform the rolling/sliding test
[58]. ................................................................................................................................ 35

vi
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 31. Colour maps of the hardness for each QT. The change in hardness is
denoted by alterations in colours as a function of the partitioning time (s) and
partitioning temperature (ºC). ................................................................................... 36
Fig. 32. Colour maps of the impact toughness (J) for each QT. The change in
energy absorbed is denoted by alterations in colours as a function of the
partitioning time (s) and partitioning temperature (ºC). ....................................... 37
Fig. 33. Colour maps of the amount of retained austenite (%) for each QT. The
change in austenite quantity is denoted by alterations in colours as a function of
the partitioning time (s) and partitioning temperature (ºC).................................. 38
Fig. 34. Effect of an increment in time in the final amount of retained austenite
a) At high partitioning temperatures (500 ºC) b) At low partitioning
temperatures (<280 ºC). .............................................................................................. 39
Fig. 35. Results of Tensile Strength vs Fracture Deformation achieved for the
different Q&P conditions. ........................................................................................... 41
Fig. 36. Macroscopic examination of fractures in tensile via SEM. ...................... 42
Fig. 37. A detailed analysis of a fracture at higher magnification. ....................... 42
Fig. 38. Map of initial roughness evaluated via optical profilometer in sample
QT165-400-30. ............................................................................................................... 43
Fig. 39. Plastic deformation close to the worn surface. Analysis via OM and
SEM in sample QT25-Pt400-90. .................................................................................. 44
Fig. 40. Microhardness evaluation from the worn surface to the interior of the
discs in the radial direction. ....................................................................................... 44
Fig. 41. Analysis of wear mechanisms on the worn surface. Detection of cracks
performing an examination with both backscatter electron and secondary
detectors (BED and SED) is possible in samples QT165-400-30 (left) and 165-400-
90 (right) respectively.................................................................................................. 45
Fig. 42. Colour maps for a quenching temperature of 165 ºC. The hardness (a),
impact toughness (b), and % of retained austenite (c) are shown as a function of
the partitioning conditions. ........................................................................................ 46
Fig. 43. Diffusion distance of a carbon atom for the partitioning conditions
analysed. ....................................................................................................................... 47
Fig. 44. XRD patterns before and after rolling/sliding test for the sample QT165-
PT400-Pt90. ................................................................................................................... 49
Fig. 45. Specific Wear Rates vs Hardness reported in the literature [37] for CFB
alloys and the Q&P data obtained in the present project. ..................................... 49

vii
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

1. Literature Review

1.1 Introduction to Advanced High Strength Steels (AHSS)

The automotive industry requirements have been increasing during the last
decade. The main challenges have been the reduction of the car weight in order
to reduce the fuel consumption and gas emissions and the improvements in the
safety of vehicles [1-3]. Considering the required mechanical properties by
automakers, the cost, formability and durability are also critical aspects that
must be optimised. These industry requirements concluded in the development
of a steel generation which was called “Advanced High Strength Steels”
(AHSS). For this reason, currently, the enhancements of the AHSS grades do not
only represent a significant area of study but also a challenging one that must
be carefully analysed.

Numerous steels grades are used in a vehicle and the mechanical properties
depend mainly on the local necessity (Fig. 1) [1]. The conventional low to high
strength steels have been widely used in automobiles due to a relatively simple
microstructure and acceptable mechanical properties. This family includes the
bake-hardenable (BH), interstitial-free (IF) and high-strength low-alloy (HSLA),
steel grades that can achieve levels of strength up to 500-600 MPa [1]. On the
other hand, the AHSS can be divided into ultra-high-strength steels or
GigaPascal steels due to strengths higher than 750 MPa or 1000 MPa
respectively. The AHSS are based in a complex microstructure, a strict chemical
composition and carefully controlled heat treatments aiming to provide not
only higher levels of strength but also ductility for an excellent formability [3].

A brief introduction to AHSS family is given below.

1
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 1. Different stress-strain curves for different steels and their application in the automobile
body structure according to their mechanical properties [1].

The Advanced High Strength Steels have been evolving over time and three
generations can be well distinguished. While the First Generation of AHSS
contains low alloy steels as the martensitic (M), complex phase (CP) and
transformation induced plasticity (TRIP) [2,3], the Second Generation was
developed by adding higher amount of alloying elements, such as manganese,
which led to the introduction of the twinning induced plasticity (TWIP), or
austenitic stainless steel grades. Despite providing an excellent combination of
strength and ductility, the cost and difficulties in the industrial processing
characterise the main problems of the Second Generation and, consequently, its
use is limited[2]. It can be seen from Fig. 2 that the current area of research is
focused on the Third Generation of AHSS in order to expand the tensile
strength/ductility combinations by developing steels with lower alloy levels
[2]. At present, the effort is mainly dedicated to improve the processing of the
already developed steels using new heat treatments or low additions of special
elements. The modified TRIP steels, carbide free bainitic steels, or quench and
partitioned steels (Q&P) are examples of materials integrating the Third
Generation.

In particular, the quenching and partitioning process is thoroughly described in

the next section aiming to understand the basis of this heat treatment and why
it could represent a promising alternative for the immediate future.
Nevertheless, it should be mentioned that, because the Q&P heat treatment has
come up to fulfil the industrial requirement of automotive industry, the
investigations have been intensively focused in low carbon steels (≈0.2 %C),
whilst its application in high carbon steel still needs further investigation.

2
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 2. Combinations of tensile strength and elongation for different generations of steels [1].

1.2 Introduction to Quenching and Partitioning

The quenching and partitioning (Q&P) represents an original heat treatment to
achieve a martensitic matrix with a minor fraction of retained austenite. The
representation of the Q&P thermal profile and microstructure evolution is
presented in Fig. 3. It can be seen that Q&P process involves a full or partial
austenitization of the material, a subsequently controlled quenching between
the martensite start (Ms) and martensite-finish temperatures in order to obtain a
desired fraction of martensite in the microstructure, and finally a holding stage
called partitioning step, followed by the cooling down to room temperature.
The partitioning stage can be performed at the same temperature that the
quenching stage, in which case the process is called one-step Q&P, or at a
higher temperature (two-step Q&P). It should be noted that up to the
quenching step the carbon content in the martensite is equal to the carbon
content in the austenitic phase, but the partitioning promotes the carbon
enrichment in the retained austenite due to the carbon diffusion from
martensite to austenite [4]. Finally, the austenite that is not sufficiently enriched
in carbon transforms into martensite during the final quenching.

Assuming that the formation of any other microconstituent is avoided, the

characteristic microstructure reached by Q&P heat treatment consists of [5]:

 Martensite transformed during the first quenching (M1). This martensite

has low carbon content due to the partitioning step and it is usually
termed as “initial martensite”. Since the martensite formation is an
athermal phenomenon, the fraction of M1 must be controlled by
adjusting the quenching temperature.

3
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

 Retained austenite (RA or γR). The austenite which does not transform at
the first quenching is stabilized due to carbon enrichment during the
partitioning stage.

 Martensite transformed in the final cooling (M2). After the second and
final cooling, despite the austenite has been carbonenriched, if its Ms
temperature is above room temperature because of an insufficient
stabilization, the martensitic transformation can also take place. This
martensite is usually called “fresh martensite” or “untempered
martensite”. It has higher dislocation density and it is supersaturated in
carbon.

Fig. 3. Schematic Temperature vs Time representation of the quenching and partitioning

thermal cycle with the microstructure evolution [5].

While the required strength is intended to be achieved by a martensitic matrix,

at least a small fraction of retained austenite is critical to provide adequate
mechanical properties related to ductility and toughness [6]. Thus, a correct
amount of martensite and a sufficient stabilization of the retained austenite are
the key points in the Q&P processing in order to obtain the desired mechanical
properties.

It is important to mention that in Q&P steel an addition of elements in the

chemical composition of the material to inhibit/delay any carbide formation is
essential because the precipitation of carbides reduces the amount of carbon
available to stabilize the austenite (Section 1.5 Chemical Composition:
Consideration in Q&P Process). It should be noted that, contrary to quenching
and tempering steels (Q&T), for Q&P even any transitional carbide
precipitation is considered detrimental and thus, a deep study related to the
effect of alloying elements on the onset and composition of these transitional
carbides is still necessary [7].

4
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

1.3 Design of the Q&P Heat Treatment

Considering the major importance of the retained austenite presence in the
microstructure, a theoretical estimation to determine the optimum quench
temperature to obtain the maximum austenite amount was proposed and
extensively used in publications [7-9]. The basis of this method is described.

A schematic diagram should be constructed aiming to obtain the optimum

quench temperature (Fig.4). First, it is required to estimate the fraction of
martensite (and also austenite) at the quench temperature by Koistinen and
Marburger relationship (Eq. 1) [10]:

𝑓𝑀 = 1 − 𝑒 −𝛼𝑚 (𝑀𝑠 −𝑄𝑇) (Eq. 1)

𝛼𝑚 = 10−3 (27.2 − ∑𝑖 𝑆𝑖 𝑋𝑖 − 19,8 (1 − 𝑒 −1.56𝑥𝑐 ))(Eq. 2)

∑𝑖 𝑆𝑖 𝑥𝑖= 0.14𝑥𝑀𝑛 + 0.21𝑥𝑆𝑖 + 0.11𝑥𝐶𝑟 + 0.08𝑥𝑁𝑖 + 0.05𝑥𝑀𝑜 (Eq. 3)

𝑓𝑀 refers to the austenite fraction transformed into martensite at the quench

temperature (QT), αm is the rate parameter and depends on the chemical
composition (Eq.2-3), and Ms is the martensite start temperature that can be
estimated by Eq. 4-5:
𝑖
𝑀𝑠 (℃) = 565 − ∑𝑛=1 𝐾𝑛 𝑥𝑛 − 600(1 − 𝑒 −0.96𝑥𝐶 ) (Eq. 4)

∑𝑖𝑛=1 𝐾𝑛 𝑥𝑛 (℃) = 31𝑥𝑀𝑛 + 13𝑥𝑆𝑖 + 10 𝑥𝐶𝑟 + 18 𝑥𝑁𝑖 + 12𝑥𝑀𝑜 (Eq. 5)

where C, Mn, Cr, Ni, Mo, Si represent the weight percent of the respective
elements in the chemical composition of the steel.

Since the initial austenite after the quenching stage can be described as 1 − 𝑓𝑀 ,
assuming that all the carbon partitions from martensite to austenite and that
other competing reactions are precluded (bainite growing, carbide
precipitation, etc.), the carbon content in austenite (%𝐶 𝛾 ) after the partitioning
step is obtained by (Eq.6):

%𝐶 𝑎𝑙𝑙𝑜𝑦
%𝐶 𝛾 = (Eq. 6)
1 − 𝑓𝑀

At the final cooling, if the austenite is not completely stabilized (which means
that the carbon content is not sufficient and its Ms temperature is above room
temperature) part of this retained austenite transforms into “fresh martensite”
following Koistinen and Marburger relationship (Eq.1).

5
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

12

10

% Carbon Content
8

Fig. 4. A theoretical diagram where the fraction of different phases and the carbon content in
austenite are represented as a function of the quenching temperature in a 0.25%C-1.58%Si-
1.24%Mn-1.7%Cr-0.14Mo-0.11V steel.

Fig. 4 illustrates the predicted evolution of the microstructure according to this

methodology in a 0.25%C steel. The final amount of austenite is shown as solid
black-line. It should be noted that the optimum quench temperature
corresponds to the peak in the austenite fraction curve. Above the optimum
quench temperature, in spite of the high retained austenite content after the
quenching, the carbon available is insufficient to decrease Ms temperature
below room temperature and thus, fresh martensite is formed during the final
cooling. On the contrary, at temperatures below the optimum, although the Ms
temperature is under room temperature (which means that the austenite is
completely stabilized and fresh martensite could not transform), a large
austenite fraction is transformed/“consumed” during the first quenching.

Differences between the retained austenite fraction measured by X-ray

diffraction analysis and the predicted according to the model described above
in a steel grade 9260 (0.6%C–0.95%Mn–1.96%Si (wt.%)) can be observed in
Fig.5 [8]. Despite the lower amount of austenite achieved, the shape of the curve
and the optimum quench temperature to obtain the maximum austenite
fraction do not change radically, which means that the method can be suitable.
The alteration on the final austenite can be explained by considering the
presence of competing reactions which either decrease significantly the carbon
available to partition and stabilize the austenite or reduce the austenite volume

6
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

due to a transformation (carbide precipitation, bainite growing respectively) [8]

(Section 1. 6. Microstructure of Q&P Steel and Competing Reactions).

The effect of the carbon content in the alloy can also be analysed by this method
(Fig.6) [8]. Increasing the carbon content derives in a rise in final austenite
fraction and in a diminution in the optimum quench temperature due to the
changes in Ms temperature.

Fig. 5. Comparison between the retained austenite fraction measured by X-ray diffraction
analysis after 10 s at 500 ºC and what is predicted according to the theoretical model in a
0.6%C–0.95%Mn–1.96%Si (wt.%) steel [8].

Fig. 6. Austenite Phase Fraction expected as a function of Quench Temperatures for steels with
different carbon content [8].

It should be recognised that the method is mainly constructed by theoretical

assumptions, which only contemplate the situation after the partitioning time.
Hence, both the kinetics and the possible compositional gradients within each

7
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

phase before completing the partitioning process are not taken into account at
all.

1.4 Stability of Retained Austenite

The effect of retained austenite in the microstructure on the mechanical
properties has been considered in TRIP steels and in Carbide Free Bainitic steels
(CFB). Owing to the similar chemical compositions and constituents in these
steels with those treated by Q&P, the stability of the retained austenite is
expected to be affected in a similar manner [11].

It was shown in several publications that the stability of retained austenite is

mainly affected by 1) the chemical composition (predominantly local carbon
content) in the austenite [11-14] 2) the grain size [11-14] 3) the morphology of
the retained austenite [13,14] 4) the constraint generated by phases surrounding
the austenite [11] and also there is small effect of dislocation density [13] and
the crystallographic orientation of the austenite with respect to the loading
direction [11].

Evaluating the Ms equation (Eq.4) it is clear that carbon is the element which
has the strongest effect decreasing Ms temperature. Hence, it can be elucidated
that carbon is the element which provides austenite with the highest stability
[12]. The carbon content should be homogeneous in the austenite in order to
avoid the transformation to fresh martensite during cooling as was considered
in earlier sections. Moreover, other alloying elements must also be considered
(Section 1. 5 Chemical Composition: Consideration in Q&P Process).

However, high carbon content can be insufficient to ensure the stability of the
austenite. García Mateo et al. [13] found that in a 0.28%C-1.5%Si-2%Mn-1.5%Cr-
0.2%Mo-1.48%Co (wt.%) the retained austenite with higher carbon content
transforms to martensite at a higher rate. A similar result was achieved by
Xiong et al. [14] and the data suggest that besides the chemical composition,
other factors as morphology can strongly affect the austenite stability.

Morphologies as blocky retained austenite and film retained austenite can

usually be seen in publications (Fig.7) [13,14]. García Mateo et al. [13] and Xiong
et al. [14] found blocky austenite with lower and higher carbon content than
film-retained austenite respectively, but in both cases blocky retained austenite
presented lower mechanically stability than film austenite. The differences in
the carbon content between morphologies in each publication are due to the
heat treatment applied to each material. The blocky austenite with lower carbon
content than film-austenite was formed by an austempering thermal cycle and
thus, the main difference between blocky and film austenite is that the carbon

8
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

remains trapped in a much smaller volume in the film-like [13]. However, a

short intercritical step was performed by Xiong et al. [14] aiming to obtain a
non-uniform carbon profile within the austenite. Therefore, the austenite at the
α/γ interface had a higher carbon content and a Ms below room temperature
when the quenching was executed, remaining untransformed as blocky
austenite. On the other hand, part of the austenite which had an insufficient
stability (due to the carbon profile in prior austenite) transforms into martensite
and only film-like austenite remains between laths of martensite.

It can be seen from Fig. 7 that while the blocky austenite starts to transform into
martensite at 2% strain, at 12% strain it was still possible to note the presence of
film austenite [14]. These results were explained considering two different
mechanisms. As it is mentioned above, blocky austenite was adjacent to
proeutectoid ferrite (due to a partial austenitization) and film austenite was
located between martensite laths. As a result of the finer grain size, the higher
carbon content and the higher dislocation density of the lath martensite
compared to proeutectoid ferrite, the yield strength of the martensite may be
assumed to be significantly higher. Therefore, because the austenite must
deform in order to accommodate the volume expansion of the martensitic
transformation during TRIP effect, the upper yield strength surrounding the
film austenite could avoid/delay the martensite transformation in this
morphology [14]. The second possible mechanism to explain these results is
based on the compression caused in the film-austenite due to the residual
stresses i.e. volume expansion during martensitic transformation that compress
the retained austenite. Consequently, this hydrostatic pressure may inhibit or
postpone the TRIP effect as the martensitic transformation would involve a
volume expansion [14].

In addition, a constraint effect of the fresh martensite has also been reported in
the literature [11, 15]. On the contrary, in this case, as the fresh martensite has
higher carbon content dissolved (which means higher strength) and it is formed
adjacent to the initial martensite, it constrains the deformation of the initial
martensite and generates stress concentrations that then result in a martensitic
transformation from retained austenite at earlier elongation [12].

9
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 7. Evolution of blocky austenite (left-bright field) and film austenite (right-dark field) in a
0.22C–1.40Si–1.8Mn (wt.%) steel as a function of strain a-b) 0% c-d)2% e-f)12% [14].

With respect to the grain size, it is accepted that finer grain size/thinner film
austenite improve the austenite stability [11,12,14]. It is also important to
mention that finer austenite leads to a faster carbon homogenization due to a
decrease in the necessary diffusion distance.

1.5 Chemical Composition: Consideration in Q&P Process

Since the absence of carbide formation and other competing reactions are
fundamental in Q&P, the chemical composition must be carefully evaluated.
The target of the chemical composition design is focused on generating the
optimal conditions to achieve retained austenite, avoid the carbide formation or
any diffusive transformation and provide the possibility to attain the required
mechanical properties.

Silicon, aluminium and phosphorous represent a suitable alternative to inhibit

diffusive transformations and thus, avoid the cementite formation, which
would eliminate carbon available to enrich the austenite (Fig. 8) [5,7,16,17]. The

10
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

solubilisation of these elements in the cementite is not possible and hence, its
rejection is required for the cementite to grow and it represents the main cause
for the delay of this transformation [17]. Although silicon appears to be more
effective in this concern [5,8,17], the influence of Si, Al and P on the onset of the
transitional carbides formation has not been completely identified and further
investigations are needed.

Whilst the aluminium is usually added to suppress the carbide formation in

CFB steel, its utilization in steels for Q&P is limited because it is associated with
faster austenite decomposition into bainite [7]. However, steels with silicon
additions have shown problems related to galvanising process because of the
formation of an adherent oxide [17]. Therefore, it is necessary to replace Si by Al
or P in steels which need to be galvanised.

Fig. 8. Schematic C-curve emphasising the delaying effect of silicon on the kinetic of carbides
formation [8].
Other elements such as manganese and nickel are necessary to improve the
austenite stability [16] because they extend the gamma region in the
equilibrium phase diagram and delay the carbide precipitation reaction [18].
The use of manganese has been more popular because it is economically
convenient compared to nickel. On the other hand, the addition of Cr, which
decreases Ms, increases the hardenability and reduces C diffusivity in austenite
and slows down the austenite decomposition, has been recently reported as
more effective than Ni in retaining higher austenite content after Q&P [19].

A correct selection of the carbon content in the alloy may be critical for several
reasons. First, Ms temperature drops considerably with the addition of carbon
and thus an austenite stabilization can be reached easily. Moreover, the carbon
content affects significantly the mechanical properties due to its interstitial
nature in steel and the existence of a solid solution strengthening.

11
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

1.6 Microstructure of Q&P Steel and Competing Reactions

A final microstructure consisting of a martensite matrix and retained austenite
between martensite laths or plates is a goal of the Q&P process, but several
competing reactions during different stages can occur and need to be analysed.

A characteristic microstructure after two steps Q&P process consist of

martensite, retained austenite and frequently fresh martensite can be seen in
Fig. 9 [20]. In addition, if the austenitization is conducted in the intercritical
region (α+γ region in the Fe-Fe3C phase diagram) the presence of intercritical
ferrite can be detected after the quenching and partitioning heat treatment
(Fig. 10) [16]. An important improvement in the final amount of retained
austenite achieved by this heat treatment was largely reported in the literature.

Fig. 9. Microstructure observed by EBSD analysis on a 0.2%C-1.5Si-2.5%Mn-1.47%Ni-1%Cr steel

a) Combination of band contrast map and colour-coded phase map where blue corresponds to
martensite (bcc) and red corresponds to retained austenite (fcc)-Darker bands are associated
with fresh martensite. b) Combination of inverse pole-figure map and band contrast map on fcc-
Bright areas correspond to martensite, dark areas are associated with fresh martensite and
austenite is shown in colours depending on its orientation [20].

Fig. 10. Microstructure evolution of a 0.2 %C-1.5 %Si-3.5 %Mn steel; quenching at 240 ºC,
partitioning for 1000 s at 350 ºC- M refers to martensite, RA denotes retained austenite and F
and TM indicate the presence of ferrite and tempered martensite respectively. a)Initial
microstructure b) Analysis after austenitization at 770 ºC c) Final microstructure after Q&P [16].

However, not only martensite, austenite and fresh martensite or ferrite can
appear in the final microstructure of steel processed via Q&P process. Other
competing reactions were reported in numerous publications. These competing

12
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

reactions can be 1) bainite growing [6,21] 2) isothermal martensite [6,22,23] 3)

tempering of martensite/carbide precipitation [5,6,8,21].

Depending on the chemical composition, an almost immediate transitional

carbide precipitation after quenching could occur in martensite before the
partitioning stage. Nevertheless, as the formation of cementite is supposed to be
inhibited, an increase in the partitioning time or partitioning temperature may
lead to a re-solution of these transitional carbides providing at this point, a
“new” source of carbon to enrich the austenite carbon content. In Fig. 11 an
important increment in the austenite fraction is observed as a second peak
when different partitioning times are considered. According to Edmonds et al.
[8] the reason for this behaviour is the re-solution of transitional carbides
formed at lower temperatures.

Fig. 11. Analysis of volume fraction of austenite as a function of partitioning time for quenching
temperatures of 180 ºC (above) and 150 ºC (below). The presence of a second peak on the
austenite volume fraction was associated with a re-solution of transitional carbides [8].

When the partitioning is performed nearby the martensite start temperature, an

earlier austenite decomposition has been reported in publications (Fig. 12)
[22,23]. This behaviour takes place just above and below the Ms temperature for
hypereutectoid and hypoeutectoid compositions respectively. This anomalous
conduct was termed “swing-back” phenomenon. The faster austenite
decomposition into bainite was detected and attributed to the presence of plate
martensite surrounding the untransformed austenite [22,23]. Also a martensite
formation was associated with the swing-back phenomenon as a result of the
stresses generated by the bainite growing [22,23]. This non-athermal martensite
can usually be found in literature as “isothermal martensite” [6].

13
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 12. Presence of a swing-back phenomenon (acceleration of austenite decomposition close to

Ms temperature) on two different steels [22].

When the partitioning time is considerable, a bainitic transformation could also

occur at temperatures of partitioning far below from Ms temperature and the
kinetics of this lower bainite is also affected by the partitioning temperature,
being faster when the temperature involved increases [6]. Additionally, the
quenching temperature has an effect on the kinetics of the bainitic
transformation. Higher quenching temperatures involve a higher initial fraction
of austenite and thus, lower carbon content present within the austenite. Hence,
as a result of the lower carbon content stabilizing the austenite, the bainite
formation occurs faster when the quench temperature is increased [6].

Another important consideration should be mentioned. The fresh, high-carbon,

martensite has been reported as a high strength and brittle microconstituent
because of its carbon content [11,15]. Therefore, its low capability to deform
leads to potential sites for voids or cracks nucleation in the interfaces fresh
martensite/initial martensite or fresh martensite/austenite [11,24].

1.7 Mechanical Properties of Steels Treated by Q&P

The Q&P process corresponds to a new alternative heat treatment to reach high
levels of strength with a relatively high ductility, not only in low alloy steels but
also in high alloy steels as stainless steels.

The mechanical properties comparison between Q&P and other heat treatment
has been largely evaluated [4,25-27]. It can be seen from Fig. 13 that due to a
microstructure formed by a martensitic matrix combined with retained
austenite, the Q&P process is characterized by improved mechanical properties
as elongation and strength compared to martensitic or TRIP and DP grades
respectively [25].

14
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

An implementation of a one-step Q&P at 180 ºC during different partitioning

times in a 0.41%C-1.27%Si-1.3%Mn-1%Ni-0.5%Cr (wt.%) registered remarkable
mechanical properties (TS~2000 MPa and a ductility ~10%) [28]. On the
contrary, Yan et al. [29] performed a comparison of mechanical performance
between Q&P and Q&T processes in a 0.20C–1.57Mn–1.55Si (wt.%) steel. While
the tensile strength of Q&P and Q&T processed samples were in the range of
1220-1340MPa and 1310-1440 MPa respectively, the total elongation in Q&P
samples (11.8-15.7%) was significantly higher than in Q&T specimens
(8.7-11.1%). The origin of this improved elongation can be traced to a
divergence in the hardening after a plastic strain of about 1.9% (Fig. 14). The
gradual increase in the slope of the Q&P curves was attributed to the TRIP
effect of retained austenite and hence, because of this transformation the
necking comes about later and a higher and uniform elongation can be
reached [29].

Fig. 13. Combination of total elongation (%) and Ultimate Tensile Strength for different steel
grades. TRIP: transformation induce plasticity, DP: dual phase, M: Martensitic, Q&P: quenching
and partitioning [25].

15
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 14. Comparison of the Engineering Stress-Strain Curve of a 0.20C–1.57Mn– 1.55Si (wt.%)
steel treated by Q&T (QT=25 ºC, Tempering Temperature=400 ºC) and Q&P (QT=300 ºC,
PT=400 ºC). A substantial different of hardening after 1.9% strain can be observed [29].

Furthermore, an improvement in toughness has been informed in the literature

[30,31]. It was shown that increasing the amount of retained austenite, higher
values of toughness can be achieved [31]. In fracture toughness testing the
enhancement was associated with a martensitic transformation at the tip of the
crack, which not only generates compressive stresses in this region (due to the
volume expansion characteristic of the martensitic transformation), but also
consumes energy available for extending the crack [31]. Considering impact
toughness, a comparison between Q&P and Q&T steels with similar levels of
strength (~1150 MPa) shown an improvement of about 50 ºC in the 50% fracture
appearance transition temperature (FATT) [32]. It was also suggested that in
order to achieve better results in the ductile-brittle transition temperature
(DBTT) it should be necessary to reduce the presence of islands of fresh
untempered martensite [32].

1.8 A Brief Introduction to Carbide Free Bainitic Steels

The Carbide Free Bainitic steels (CFB) have also been developed to fulfil the
industry requirements of high strength and good ductility and thus, they
belong to the Third Generation of Advanced High Strength Steels. Its design
aims to obtain a nanoscaled microstructure constituted of carbide free bainite
and retained austenite [33,34].

The conventional upper and low bainite have been widely studied [35], a
distinction is made based on the distribution and location of carbides. However,
the development of CFB is based on the suppression of carbides by high silicon
content and then, the possibility to stabilize the residual austenite by carbon
partitioning into it after the bainite growth [33,34,36]. The heat treatment
consists of an austenitization followed by a cooling to a temperature between
bainite start temperature (Bs) and martensite start temperature (Ms), where the
16
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

material is held for some time and cooled down to room temperature
(Fig. 15) [37]. Thus, the final microstructure is a combination of thin plates of
bainite and carbon enriched austenite. Theoretically and ideally, the austenite
should be located between bainitic ferrite plates (film austenite) and the
martensitic transformation of the austenite during the final cooling should be
avoided. Hence, the absence of carbides leads to a high resistance to void
formation; the extremely fine microstructure improves not only the strength but
also the toughness and the TRIP effect of the retained austenite during
deformation enhances the ductility [33,37,38].

Fig. 15. Schematic illustration of an austempering thermal cycle.

Conventionally, bainite is in the micrometre range, but the possibility to add

elements of alloy to improve the hardenability and decrease both Bs and Ms
allows the austempering at a lower temperature and the refinement of the
bainite until the nanoscale [38,39]. For this reason, the austempering
temperature is critical to obtain the desired mechanical properties. A decrease
in the transformation temperature concludes in a stronger austenite and thus,
its displacive transformation into bainite takes place in smaller sub-units. It
leads to a finer bainite and an improvement in the mechanical properties
[40,41]. The effect of the transformation temperature is pointed out in Fig. 16
[38,42]. This ‘new’ phase is usually denominated low-temperature bainite
(LTB), and the minimum temperature where bainite can be obtained is a great
interest to design CFB steels [38].

17
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 16. Effects of austempering temperature on the microstructure and mechanical properties
of CFB steels. Left) Effect of transformation temperature on the bainite thickness and hardness
[38]. Right) Stress-Strain curves of an alloy isothermally transformed at different
temperatures [42].

Nevertheless, the volume content of bainite is thermodynamically limited and

low bainite content in the microstructure may lead to large pools of austenite
(blocky austenite), which are detrimental to achieve good levels of toughness.
The martensite transformed through deformation of the retained austenite
tends to crack and this tendency is related to the absolute size of the martensite.
Thus, coarser retained austenite promotes susceptibility of cracking when the
martensite transformation comes about. On the other hand, the martensite
which is formed during deformation in the fine retained austenite between
bainitic ferrite plates, is tolerated and does not compromise the ductility [33,38].
For this reason, a high volume fraction of bainite in the microstructure, the
design and the control of the amount, shape and composition of the retained
austenite are the key points to obtain good mechanical properties in CFB steels.
However, poor impact toughness seems to be an intrinsic properties of these
steels with the presence of untempered brittle martensite in the microstructure
[38]. This behaviour is associated with the fracture of martensite at high strain
rates, due to the lack of a phase that may accommodate the imposed strain. The
energy absorbed during Charpy impact test were reported between 5 J up to
25 J for relative high carbon steels [37,38].

Fig. 17 denotes the free energy of both austenite (γ) and ferrite (α) at a certain
temperature (T1) and as a function of carbon content. The displacive
transformation of austenite into bainite could only take place if it reduces the
free energy of the phase. It means that, if the carbon supersaturation of the
bainite is rejected into the residual austenite, there is a point corresponding to
certain carbon content in the austenite, where the austenite and ferrite have the

18
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

same free energy and thus, there is no driving force for the transformation of
the austenite into bainite and then, the transformation is arrested. This is
usually known as incomplete reaction phenomenon, because the transformation
stops before the phases achieve the equilibrium compositions.

An alteration in the temperature at which the transformation happens leads to a

change in the carbon content where the free energy of ferrite and austenite are
equal. For this reason, a To curve is defined as the locus of all points,
considering temperature against carbon content, where austenite and ferrite of
same composition have the same free energy. In addition, as bainitic ferrite has
a higher stored energy than the ferrite, a higher free energy for bainite
(α’+strain) and To’ curve are defined [35].

To’ curve imposes a thermodynamically limit for the carbide free bainitic
transformation, and it can be a severe limitation if large pools of austenite
remain in the microstructure when the transformation is arrested. A number of
solutions to control the T0’ curve, to allow the transformation of higher
volumes of bainite and preventing the formation of large quantities of blocky
austenite, were purposed [33,36]:

1. Reducing the mean carbon concentration. The critical carbon

concentration in the austenite is reached at a higher volume fraction of
bainite. Nevertheless, the impact of an increase the volume of bainite and
reduction in the carbon content in the strength should be carefully
analysed.

2. Increasing the substitutional alloying elements. It has been proved that

the To’ curve can be moved to higher carbon concentration by modifying
the elements of alloys.

3. Minimizing the transformation temperature. Fixing the nominal carbon

content, an austempering at a lower temperature means that the To’
condition is reached at a higher carbon content in the austenite. The
latter results in an increase in the final volume of bainite in the
microstructure and, as it is previously mentioned, better mechanical
properties.

19
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 17. Schematic illustration of To curve on the phase diagram. Free energy of both austenite
(γ) and ferrite (α) at a certain temperature (T1) and as a function of carbon concentration.

20
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

1.8.1 Wear Performance in CFB steels

Wear is defined as the damage originated by material removal from the surface
[43] and it has a great impact in cost. The purpose of this section is not to give a
comprehensive description of wear and tribology, but rather to establish a few
significant characteristic of wear and its terminology, along with the recent
investigations in the field of carbide free bainitic steels.

There are several applications such as railway wheels, rails, bearing, gears, etc.
that require excellent wear performance of the material [43-45], but it is
important to mention that “wear is not a material property, it is a system response”
[46]. For this reason, any changes in conditions such as environmental
parameters (humidity, atmosphere, temperature, etc.), material parameters
(surface properties, hardness, microstructure, etc.) and dynamic parameters
(force, speed, sliding, etc.) lead to changes in the wear performance of the
system [43,46].

The wear performance of nanostructured low-temperature bainitic steels was

extensively evaluated in Research Fund for Coal and Steel project (RFCS)
known as “Nanobain” [37]. Several fruitful publications came about and they
can be found in literature [47-50]. Fig. 18 [48] summarized the wear resistance of
different CFB steels and temperature of austempering, a significant
improvement in the specific wear rate (SWR) can be observed when compared
to Q&T steels or bainitic steels exhibiting similar HV. It should be noted that
SWR was evaluated using a twin-discs in a dry rolling test, and it is described
as the volume loss (mm3) divided by (the load applied multiple by the sliding
distance) (Eq. 7).

𝑚𝑚3 𝑉𝑜𝑙𝑢𝑚𝑒 𝑙𝑜𝑠𝑠 (𝑑𝑖𝑠𝑐 1 + 𝑑𝑖𝑠𝑐 2) (𝑚𝑚3 )

𝑆𝑊𝑅 ( )= (Eq. 7)
𝑁. 𝑚 𝐿𝑜𝑎𝑑 (𝑁). 𝑆𝑙𝑖𝑑𝑖𝑛𝑔 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒 (𝑚)

The conditions of the experiment are described in following sections (see.

Materials and Methods). The mass loss on each disc after the test is measured
and converted into volume loss using the conventional density of steels
(7.84 mg/mm3).

21
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 18. Specific wear rate (SWR) as a function of hardness (HV) of different alloys and
temperature of austempering. In red it is highlighted the behaviour of CFB alloys [37].

The excellent wear resistance of CFB steels was attributed, besides to the
hardness, to the fine scale nanostructure and the role of austenite, which under
stress or strain can transform into martensite generating not only a hardened
layer but also avoiding the crack propagation [46-48]. Therefore, structures
containing retained austenite result promising regarding wear performance and
because of these reasons, in this study some Q&P conditions are analysed and
compared with CFB steels.

In Fig. 19a [49] the hardness profile from the worn surface is shown for three
different test duration (cycles). A strengthen close to the surface is associated
not only with the plastic deformation on the surface, which increase the
dislocation density, but also with the transformation of the fine high-carbon
austenite into hard martensite. On the other hand, the depth of the deformed
layer is based in the duration of the experiments and the mean hardness of the
material (Fig. 19b) [49].

22
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 19. a) Hardness profile from worn surface for different test duration. b) Hardened layer
extension as a function of the test cycles [49].

The reduction of the austenite peaks on the surface was confirmed via XRD
analysis. However, the total amount of austenite does not transform into
martensite (austenite diminished from 27.9% to 17.3%) and it is associated with
the high stability of the austenite and the phenomenon known as mechanical
stabilization [49]. A severe plastic deformation prevents the austenite-
martensite transformation due to the fact that it needs the presence of a glissile
interface.

1.9 A Comparison between CFB and Q&P

Besides the differences in the thermal cycle, CFB and Q&P were developed in
order to reach the same mechanical properties: advanced high strength
(>1 GPa) conserving good ductility and toughness. Indeed, the final
microstructure is based on a similar design, a strong matrix with retained
austenite as the key point to enhance ductility [25]. Consequently, the volume
of retained austenite after the heat treatment and its stability, which is
determined in both cases for the same factors (composition, shape, etc.) [11-13],
have a fundamental effect to increase the strain rate and the tensile elongation.
Furthermore, as it is mentioned before, large pools of austenite represent a
problem in both microstructures when they transform to martensite under
deformation, because this large untempered martensite is likely to crack and
decrease the toughness [38]. The design of the chemical composition of the alloy
can be similar in some cases. High silicon content (>1.5 wt.%) is added to delay
carbide precipitation [33]. However, in CFB the addition of Al and/or Co is
favourable to accelerate the bainite formation and this is an unfavourable effect
in Q&P, due to the bainite growth consumes the available austenite during the
partitioning stage. Theoretically, the microstructures are formed by bainite+RA
(CBF) and martensite+RA (Q&P). Nevertheless, sometimes, the formation of
other microconstituents is inevitable. In CFB steels, if the retained austenite is
not completely stabilized during the austempering, which means that M s is not

23
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

below room temperature, fresh martensite can appear during final cooling. This
behaviour and the possibility of other competing reactions during the
partitioning are also expected in Q&P. It means that the microstructure can be
different to the one that was designed, and the consequences of the heat
treatment temperatures and times should be evaluated for each chemical
composition in particular.

The differences between CFB and Q&P steels come about from the thermal
cycle applied and the microstructure achieved. The martensitic transformation
in low alloy steels requires a rapid cooling and thus, the introduction of
residual stresses into the component is most likely. On the other hand, the
thermal cycle for CFB is simple and the bainite reaction is known to have a
superior capacity for control than the martensite [38], but the isothermal stage
may take much longer than the Q&P heat treatment. The researchers report
holding times at the austempering temperature between 12h up to more than 60
days to complete the bainite transformation [34,37,38]. However, as there is no
requirement for a rapid cooling during CFB thermal cycle, the component can
be gently transferred from the austenitization temperature into an oven at the
austempering temperature. The industrial advantage of a less severe cooling
relies on the capability to heat treat large component [38]. Furthermore, any
residual stress that can evolve from the cooling down is removed during the
long austempering. Despite the thermal cycle for CFB does not seem to have
several complications, as it is mentioned before, the mechanical properties
achieved after heat treatment are totally related to the temperature of the
isothermal stage [37] and a careful design of the chemical composition and
austempering conditions is necessary [36,42]

Theoretically, in both heat treatments, the final retained austenite content can be
estimated by thermodynamic concepts [33,51]. But, because of the incomplete
reaction phenomenon described above, the austenite content reported by CFB
steels is usually larger than those achieved by Q&P steels of similar chemical
composition [7,17,36]. In addition, the bainite is usually softer than the
martensite and hence, a bainitic microstructure does not reach the same levels
of hardness that a martensitic steel with the same composition [8,36]. Although
to reach levels of hardness in martensite around 800 HV, the dissolved carbon
in martensite should be high which in turns implies an extremely brittle
microstructure. On the contrary, the reported values in low-temperature bainite
are around 600-670 HV [38,52]

The microstructure in CFB is finer than the martensite in Q&P steels.

Additionally, the bainite transformation rejects carbon to the austenite that
remains there soon after the growing and thus, the final retained austenite has
24
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

high carbon content and a high stability when it is as film austenite. On the
other hand, the complete carbon partitioning from martensite to austenite and
its chemical homogenization is more difficult to achieve by Q&P, and along
with the most likely presence of untempered martensite, it results in relatively
better toughness for CFB than Q&P steels.

Accordingly, higher strength but less elongation should be expected in Q&P

steels when compared with CFB of a similar chemical composition.
Nevertheless, because the design of these heat treatments are not
straightforward, the selection of temperatures and times conditions plays a
critical role in the mechanical properties, this rule is not completely correct.

2. Objectives
As it is mentioned in previous sections, the Q&P heat treatment has come up to
fulfil the industrial requirement of automotive industry and the investigations
have been intensively focused on low carbon steels (≈0.2%C). In the present
study, the application of Q&P on high carbon steel is intended to be analysed.
The goals of the present study are clearly defined:

 To investigate the effect of different quenching and partitioning

parameters, i.e. quenching temperature, partitioning temperature, and
partitioning time, on the microstructure and mechanical properties.

 To elaborate maps of the possible mechanical properties that a

quenching and partitioning heat treatment can achieve for a particular
chemical composition.

 In addition, based on previous results obtained in carbide free bainite

condition for the particular alloy analysed, a comparison between the
properties achieved in CFB and Q&P is highly desirable.

25
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

3. Materials and Methods

3.1 Base Metal

The steel used in the following study is known as 06CV, it is manufactured by
ArcelorMittal® and it was evaluated as potential CFB steel by “Nanobain”
project [37]. The chemical composition is listed in Table 1. According to Eq.4-5,
the nominal Ms temperature is 223 ºC.

Table 1. The chemical composition of steel 06CV

%C %Si %Mn %Cr %Mo %V
0.6CV 0.6 1.6 1.25 1.75 0.15 0.12

The thermodynamic diagram Fe-FeC3 was simulated by ThermoCalc ® and it is

shown in Fig. 20. It can be seen that a 0.6%C content is close to the eutectoid
carbon composition (red arrow) and examining the microstructure of the as-
received base metal by optical microscopy, it is possible to observe a fully
perlitic steel with a fine perlitic interspacing (Fig. 21). In addition, it should be
noted that, for a carbon content of 0.6%, the minimum temperature to reach the
austenitic field is ≈815 ºC.

Fig. 20. Phase diagram of 06CV simulated in ThermoCalc.

The evaluation of Vickers hardness reveals a value of 346 ± 7.6 HV0.5, which
corresponds well to a pearlitic structure. Thus, it is possible to conclude that the
condition of the base metal is regular.

26
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 21. The fully perlitic microstructure of the material before Q&P (base metal). Nital etching.

3.2 Heat Treatment and Samples Nomenclature

Three different quenching temperatures (190 ºC-165 ºC-25 ºC) and different
partition conditions were analysed, groups of samples are divided according to
the quenching temperature. Five partitioning temperatures were applied
(500 ºC-400 ºC-280 ºC-250 ºC-220 ºC) with time alterations (from 2 s up to 900 s),
aiming to achieve similar levels of carbon enrichment in the austenite. Thus,
while decreasing the partitioning temperature, an increase in time was carried
out. A total of 48 samples were analysed (Fig. 22).

Fig. 22. Schematic illustration of the Q&P conditions evaluated. PT and Pt refer to partitioning
temperature and partitioning time respectively.

27
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

The nomenclature of the samples is explained in Fig. 23 where QT, PT, and Pt
refer to the quenching temperature, partitioning temperature and partitioning
time respectively, while X and Y denote the temperature (in Celsius) and time
(in seconds) used during the different step of the heat treatment.

Fig. 23. Nomenclature of the samples.

The austenitization was carried out considering data from literature [37] and
the thermodynamic diagram simulated; samples were austenitized at 890 ºC for
1h using a Nabertherm N11/H batch furnace. The quenching stage was
performed in either Therm Concept salt bath (for QT190 and QT165) or oil (for
samples quenched until room temperature). Samples quenched until 190 ºC and
165 ºC were held in the salt bath for 1min, in order to reach the real temperature
of the salt bath and immediately subsequent, they were partitioned at the
respective temperature in a second salt bath. The configuration of the
equipment can be seen in Fig. 24. On the other hand, for samples quenched
until room temperature, the oil was removed before they were partitioned in
the salt bath. All the samples were cooled down in calm air after the
partitioning.

Fig. 24. Configuration of the furnace and two salt baths utilized to perform heat treatments.

28
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

In order to analyse the real thermal profile achieved in the samples, a

thermocouple K-type was welded on the centre of the longitudinal surface of a
10x10x75 mm sample using DSI Thermocouple Welder (35 V). A complete heat
treatment with a quenching temperature of 165 ºC and a partitioning at 400 ºC
was performed. The temperature was recorded along the time and it can be
observed in Fig. 25.

1000
900
800
Temperature (ºC)

700
600
500
400
300
200
100
0
0 20 40 60 80 100 120 140 160
Time (s)

Fig. 25. Real Temperature-Time profile measured in samples during heat treatment. Quenching
until 165 ⁰C and partitioning at 400 ⁰C.

An estimation of the distance that a carbon atom can diffuse at a certain

temperature and a given time can be evaluated through Eq. 8 [53]:
𝑟 = 2.4√𝐷 ∗ 𝑡 (Eq. 8)

where D is the diffusion coefficient of carbon in austenite (Eq. 9)[54]:

𝑐𝑚2 −31350
𝐷 ( ) = (0.04 + 0.08𝐶)𝑒𝑥𝑝( 𝑅𝑐∗𝑇 ) (𝐸𝑞. 9)
𝑠
𝑐𝑎𝑙
Rc and T are the ideal gas constant (1.987 ) and the temperature in Kelvin
𝐾.𝑚𝑜𝑙
respectively.

It should be noted that it takes around 20 s to reach the real temperature of the
salt bath after the first quenching (from 60 s to 80 s in Fig 25). The distance
diffused by carbon, considering the thermal profile of Fig. 25 from the 60 s to
80 s, can be estimated as just the 33% of the distance diffused if the temperature
is 400 ºC during 20 s. For this reason, it was decided to introduce samples in the
partitioning salt bath and start counting the partitioning time after the first 20 s.
Thus, Pt refers to the real time of the sample at the partitioning temperature.

Samples of 10x10x75 mm were heat treated to evaluate the microstructure,

hardness and impact toughness, while tensile test and wear specimens were

29
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

machined and heat treated in a second round using equal parameters and
equipment.

It should be mention that the austenitization was performed in an air

atmosphere and a surface decarburization, which leads to a decrease in
hardness, of around 250 µm was detected after heat treatment (Fig. 26). For this
reason, the diameter of tensile test and twin-disc specimens were ground and
reduced 400 µm after heat treatment in order to remove the decarburized layer.

Fig. 26. Hardness (HV0.1) as a function of the distance from the surface (µm). A drop in
hardness close to the surface is associated with a decarburization during the heat treatment.

3.3 Metallography – Microhardness and Impact Toughness

Results of the project work “Microstructural and Mechanical Investigations of
Q&P Advanced High Strength Steel”, which was carried out during autumn
semester 2017 by the author of this report, are also shown in the present study
in order to understand and analyse in depth the effects of the heat treatment
conditions.

A polished surface was prepared by conventional manual grinding and

polishing in order to reveal the microstructure and evaluate the microhardness.
The specimens were ground using Buehler MetaServ 250 (P220-P600-P1200),
while the polishing was completed in Struers LaboPol-5. Diamond suspensions
of 9 µm - 6 µm -1 µm - 0.25 µm and a final stage of colloidal silica of 0.06 µm
were utilized.

A Matsuzawa MXT microhardness tester was utilized to study the Vickers

Hardness (HV) variations of the samples, alterations that may be attributed to
the change in heat treatment conditions. A load of 500 g (HV0.5) was applied to

30
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

obtain an indentation size with a diagonal bigger than 25 µm that could include
all the phases present in the microstructure.

The optical microscopy (OM) analysis was carried out using Nikon Eclipse
MA200. LePera colour etching (Table 2) was utilized to reveal and distinguish
initial martensite (coloured in blue) from bainite (brown), while either retained
austenite or fresh martensite do not etch and can be observed in white [55].
Samples QT165 and QT200 were etched during 15-20 s, while 30 s were
necessary in order to reveal properly the microstructure in specimens quenched
until room temperature. The difference in time is attributed to the variation in
the amount of initial martensite.

On the other hand, in order to resolve and analyse the fine microstructure
generated by Q&P, an evaluation by scanning electron microscopy (SEM) in
JEOL JSM-IT300LV microscope was performed. In addition, Oxford X-Max
energy dispersive spectroscopy (EDS) attached to the SEM was utilized when
necessary. For SEM-EDS analysis, an etching with the conventional 2%-Nital
during 15 s was performed (Table 2).

Table 2 Etchant to reveal the microstructure, its effects, and observations

Etching Solution Effect on Observations
microstructure
*LePera Ferrite Austenite and fresh
Equal portions of solutions: yellow/blue; martensite not
(a) 1% Na2S2O5 in bainite brown, separately
aqueous dilution retained identifiable.
(b) 4% picric acid in austenite and
ethanol fresh martensite
white
**2%-Nital Reveals alpha The most common
(a) 98ml ethanol 99% grain etchant for iron,
(b) 2ml nitric acid 65% boundaries and carbon, alloyed
constituents. steels, and cast iron.
The 2 or 4% solution
is commonly used.
Use by immersion of
sample for
up to 60 s.
*[55,56]

The length of the samples was cut from 75 mm down to 55 mm in the Struers
Discotom-100 cutting machine. Thus, the dimensions are according to ASTM
Standard E 23-“Standard Test Methods for Notched Bar Impact Testing of
Metallic Materials” (Fig. 27). The Charpy Impact Tests were carried out at room
temperature and the energy absorbed during test is reported in joule (J).
31
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 27. Charpy V-Notched dimensions according to ASTM Standard E 23-“Standard Test
Methods for Notched Bar Impact Testing of Metallic Materials”.

3.4 XRD Analysis

A transversal cutting was executed in Struers Secotom-10 using a feed speed of
0.05 mm/s aiming to introduce the lowest amount of plastic deformation. A
final cutting parallel to the first were performed in order to obtain samples of
10x10x5 mm.

The conventional grinding (P220-P600-P1200-P2500) and polishing (6 µm-3 µm-

1 µm-0.25 µm) step by step were performed until the last stage of Colloidal
Silica (0.06 µm). Panalytical Empyrean diffractometer was operated to obtain
XRD patterns, collecting data from 40º to 100º (2ϴ) with a step size of
0.0131303º, and a scan speed of 0.026796 º/s.

HighScore Plus v3.0.4 software was utilized to perform semi-automatic Rietveld

analysis and determinate the phases fraction and its lattice parameter. It should
be mentioned that, due to the limited database of patterns, only austenite and
‘ferrite’ can be recognized, which means that both martensite and bainite are
being considered as the same phase. Furthermore, results were double-checked
in HighScore Plus v4.7 performing an automatic Rietveld simulation. The
penetration of X-Rays, depending on the angle of incidence, was calculated also
in HighScore Plus with results varying between 3 µm and 9 µm.

Table 3 describes different equations considered to measure the carbon content

in austenite:

32
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Table 3. Equations to determine the austenite carbon content

Reference Equation
De Knijf [11] 𝑎(𝑛𝑚) = 0.3556 + 0.00453%𝐶 + 0.000095%𝑀𝑛
Purdy [23] (𝑎(𝑛𝑚) − 0.3578)
%𝐶𝛾 =
0.0033
Van Dijk [12] 𝑎(𝐴̇) = 3.556 + 0.0453%𝐶 − 0.00095%𝑀𝑛 − 0.0056%𝐴𝑙
Yan [29] 𝑎 (𝑛𝑚) = 0.35467 + 0.00467%𝐶
Vuorinen 𝑎(𝑛𝑚) = 0.03578 + 0.0033%𝐶 + 0.000095𝑀𝑛 − 0.00002𝑁𝑖
[57] + 0.00006%𝐶𝑟 + 0.00031%𝑀𝑜 + 0.00018%𝑉
Santofimia 𝑎(𝑛𝑚) = 0.3556 + 0.00453%𝐶 + 0.000095%𝑀𝑛 + 0.00056%𝐴𝑙
[20] + 0.0006%𝐶𝑟 − 0.0002%𝑁𝑖
Li [28] 𝑎 = 0.3573 + 0.0033𝐶 + 0.000095𝑀𝑛 − 0.00002𝑁𝑖 + 0.00006𝐶𝑟
+ 0.00031𝑀𝑜 + 0.00018𝑉

3.5 Tensile Test

The extremely high hardness of specimens (above 700 HV) and consequently
high strength has brought problems regarding the machining, grinding and
testing. The complications are described in detail in “Appendix B – Problems
with Tensile Test Specimens”. Finally, a sub-size specimen was machined and
the shape in the as-machined, after heat treatment, and after grinding condition
can be observed in Fig. 28a-c respectively. In addition, the dimensions are also
specified in Fig. 28d. It should be noted that the gage length is 15 mm with a
diameter of 3 mm (L/D=5). The tensile test was carried out at room
temperature in Gleebe 3800, with a longitudinal glass extensometer (HZT
extensometer), which is designed to operate at high temperature. The lack of a
correct and consistent extensometer was the main problem to measure ductility.
The arrangement of the specimen in Gleebe 3800 before the test is shown in
Fig. 29. The strain rate applied was 0.005 s-1. In addition, the extensometer
slipped during the test and did not correctly record the strain; the reported
strain is the fracture deformation calculated measuring the distance between
two lines in the gage length before and after the test.

Due to distortion during the heat treatment, the subsequent grinding could not
guarantee the circular shape of the 3 mm diameter. For this reason, after the
tensile test, a transversal cutting of each specimen was inspected via SEM and
Nikon SMZ 1270 Stereomicroscope. Each particular area was calculated
processing the pictures with Image J Software and the stress was recalculated
considering the force recorded during the experiment.

It should also be mentioned that due to lack of available material and funding,
only 12 Q&P conditions were carefully selected to be tested in tensile. The
selection was performed considering the previous results of hardness and XRD
analysis.

33
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

A macro and micro examination of fracture surfaces via Stereomicroscopy and

SEM were performed in order to reveal the fracture mechanisms and evaluate
the fracture initiation in some specimens.

a) b) c) d)

Fig. 28. Shape and layout of tensile test samples a) As-machined b) After heat treatment c) After
grinding d) Layout of specimens.

Fig. 29. Disposition of the tensile test specimen and extensometer in Gleebe 3800.

3.6 Rolling/Sliding Wear Test

The wear performance of 3 different Q&P conditions was evaluated in
rolling/sliding conditions. Tests were carried out in a UTM 2000 twin disc
machine, considering the same parameters utilized in Nanobain project [37] in
order to be able to compare results. Thus, cylindrical discs of ≈44.6 mm and
10 mm thickness were tested (Fig. 30) in dry conditions at a speed of 100 rpm
and 95 rpm, which in turn imply 5 % of slip between discs. The test was run
during 30000 cycles (5 h duration) with an applied load of 300 N. A schematic
illustration can be observed in Fig. 30b [58].

34
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 30. a) Disc of 44.6 mm diameter tested in dry rolling/sliding condition. b) Schematic
illustration of the set-up utilized to perform the rolling/sliding test [58].

Before rolling/sliding test, the as-machined roughness (Ra) was evaluated in a

Wyko 1100NT 3D Optical Profiler. In order to determine the volume loss and
the specific wear rate (SWR-Eq. 7), the mass of each disc was weighted three
times before and after test in a Mettler Toledo AX205 balance, with an accuracy
of 0.01 mg. The mean mass, the standard deviation (SD) and the mass loss were
calculated. After the test, the deformed layer was examined by SEM, OM, and
microhardness (HV0.05). On the other hand, the worn surface was inspected by
SEM. In addition, an XRD analysis to evaluate the amount of austenite on the
worn surface and in the bulk material was performed. The description of
equipment has been described in previous sections.

4. Results

4.1 Hardness
The evaluation of the Vickers hardness is summarized in three different colour
maps. The levels of hardness are denoted by alterations in colours as a function
of the partitioning time and partitioning temperature, for each quenching
temperature (Fig 31).

Two important features showed in Fig. 31 should be highlighted:

a. Even though samples quenched until room temperature (QT25-Fig. 31c)

have a higher amount of initial martensite (which is the martensite
generated during the first controlled quenching), the highest levels of
hardness are achieved in QT165 and QT190 (Fig. 31a-b). In addition, a
hardness of ≈800 HV is related to the hardness of an untempered
martensite containing ≈0.6 %C. This behaviour confirms the presence of
fresh martensite, generated during final cooling, in QT165 and QT190.
Due to the high carbon content distorting the body central tetragonal

35
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

(bct) unit cell of the martensite, plus the high dislocation density, the
fresh martensite represents a hard phase in the microstructure.
b. The location of the lowest levels of hardness has a clear tendency. For
the three QT, the lowest levels of hardness are reached when the
highest partitioning temperatures are applied. This behaviour can be
seen in blue in the lower-right corner of Fig. 31a-c.

Fig. 31. Colour maps of the hardness for each QT. The change in hardness is denoted by alterations in
colours as a function of the partitioning time (s) and partitioning temperature (ºC).

4.2 Impact Test

Results of the energy absorbed in a Charpy V-Notch Test are presented in
different colours maps according to the quenching temperature (Fig. 32). In this
case, the impact toughness (J) is shown by alterations in colours as a function of
the partitioning time and partitioning temperature.

The values of impact toughness fluctuate from 2,8 J to 11,7 J. Furthermore, it

should be underlined that, in general, the highest energies absorbed are also
reached when the highest partitioning temperatures are applied.

36
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 32. Colour maps of the impact toughness (J) for each QT. The change in energy absorbed is denoted
by alterations in colours as a function of the partitioning time (s) and partitioning temperature (ºC).

4.3 X-Ray Diffraction

The alterations of volume percent (%) of retained austenite, analysed via
Rietveld simulation in HighScore Plus v4.7, for different partitioning conditions
are shown in colours maps in Fig. 33a-c.

37
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 33. Colour maps of the amount of retained austenite (%) for each QT. The change in austenite
quantity is denoted by alterations in colours as a function of the partitioning time (s) and partitioning
temperature (ºC).

Moreover, the effect of an increment in time in the final amount of retained

austenite is shown in Fig. 34a-b. The effect of time is linked with the
partitioning temperature applied. At a “high partitioning temperature” (500 ºC)
an increment in the partitioning time results in an austenite reduction (Fig. 34a).
But, in contrast, at “low partitioning temperatures” (≤280 ºC) an extension of
the partitioning time leads to an increase in the quantity of retained austenite
(Fig.34b).

In Fig. 34a faster austenite decomposition can also be observed when the
quenching temperature is higher. Additionally, in Fig. 34b the data clearly
suggest a higher quantity of retained austenite at lower quenching temperature
but, considering a fixed QT and Pt, at higher partitioning temperature allows
achieving a higher amount of retained austenite.

38
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 34. Effect of an increment in time in the final amount of retained austenite a) At high
partitioning temperatures (500 ºC) b) At low partitioning temperatures (<280 ºC).

The calculation of carbon content in austenite, according to the equation

reported in the literature by Yan [29] and considering the unit cell acquired via
Rietveld simulation, is shown in Table 4. The data do not show a specific
tendency. Nevertheless, it can be observed that the level of carbon content in
austenite increases when the quenching temperature is reduced. Carbon
content for QT25 is around 0.99-1.44 %wt., being 0.38-1.32 %wt. and 0.67-
0.89 %wt. for QT165 and QT190 respectively.

39
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Table 4. Calculation of carbon content in austenite considering the unit cell (measured by XRD
and Rietveld simulation) and using Yan equation.
Partitioning QT190 QT165 QT 25
Partitioning
Temperature %C %C %C
Time (s)
ºC (wt.%) (wt.%) (wt.%)
500 2 0.90 1.42
500 5 0.82 1.33 1.34
500 20 0.78 1.07 1.41
400 30 0.76 0.97 1.44
400 60 0.83 1.07 1.43
400 90 0.68 0.95 1.45
400 150 0.89 0.76
280 180 0.83 0.39 1.07
280 600 0.77 0.93 1.04
280 900 0.86 0.44 1.00
250 180 0.76
250 600 0.81 0.67 1.18
250 900 0.82 0.63 1.04
220 180 0.70 0.56
220 600 0.83 0.54

4.4 Tensile Test

The results of tensile strength and fracture deformation for the 12 Q&P
conditions analysed are listed in Table 5 and Fig. 35.

Table 5. Results of Tensile Strength (UTS) and Fracture Deformation.

Sample UTS (MPa) Fracture Deformation

QT25-PT400-Pt30 2407 5.7%

QT25-PT280-Pt180 2544 4.8%

QT25-PT280-Pt600 1931 2.9%

QT165-PT400-Pt30 1230 1.1%

QT165-PT400-Pt60 1520 1.2%

QT165-PT400-Pt90 1575 1.3%

QT165-PT280-Pt180 1426 3.9%

QT165-PT280-Pt600 1142 0.9%

QT165-PT250-Pt600 1393 1.3%

QT190-PT280-Pt180 1515 1.5%

QT25-PT280-Pt2h 2432 3.8%

40
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

The values of UTS rise up from 1.1 GPa to 2.5 GPa, while fracture deformation
is almost negligible in some specimens (≈0.9 %) and increases up to 5.7 %.

Samples quenched until room temperature achieve not only the highest UTS
but also the largest fracture deformation (red circle in Fig. 35).

Fig. 35. Results of Tensile Strength vs Fracture Deformation achieved for the different Q&P
conditions.

4.4.1 Fracture Analysis

An examination of the general aspects of the fractures was performed via SEM
(Fig. 36-37). All samples exhibit a brittle fracture with evidences of a
transgranular mechanism, which is in accordance with the levels of strain
achieved in the tensile tests. In addition, the fracture initiation takes place on
the surface of the specimen in all cases, suggesting the presence of stress-
concentration marks provided by the grinding. A detailed observation suggests
the presence of shear lips characterized by dimples/voids along the perimeter
(Fig. 37). A macroscopic examination shows that the fracture comes about close
to the initiation of the gage length in almost all the specimens, where a grinding
defect could be perceived. Indeed, these defects could also reduce the plastic
deformation of the specimens.

41
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 36. Macroscopic examination of fractures in tensile via SEM.

Fig. 37. A detailed analysis of a fracture at higher magnification.

42
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

4.5 Rolling/Sliding Wear Test

The data of the Q&P conditions tested are summarized in Table 6. Discs were
tested as-grinded and the initial roughness evaluated via optical profilometer
shows Ra values in a narrow range of 0.27-0.45 µm (Fig. 38). The specific wear
rate (mm3/N.d) was calculated according to Eq. 7. Furthermore, RA and RA on
worn surface are related to the retained austenite content measured via XRD
analysis on the surface previous and after the wear test respectively.

Table 6. Summary of specimens tested in dry rolling/sliding conditions and their respective
results.
RA
RA on worn
SWR Ra
Conditions Tested HV surface
(%) (mm3/N.m) (µm)
(%)

QT25-PT400-Pt30 710 20.1 12 2.42E-5 0.43-0.45

QT165-PT400-Pt30 769 14.3 10.6 2.28E-5 0.27-0.34

QT165-PT400-Pt90 777 16.5 0.7 5.64E-5 0.32-0.34

Fig. 38. Map of initial roughness evaluated via optical profilometer in sample QT165-400-30.

A severe plastic deformation on the worn surface in the rolling direction was
detected via OM and SEM in a transversal cutting in the radial direction of the
discs (Fig. 39). In addition, Fig. 40 shows the microhardness evaluation from the
worn surface to the interior of the discs in the radial direction. A hardened layer

43
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

of ~200 µm is detected close to the external worn-surface, the straight lines

represent the bulk hardness of the material.

Fig. 39. Plastic deformation close to the worn surface. Analysis via OM and SEM in sample
QT25-Pt400-90.

1050
1000 QT165-PT400-Pt90
QT165-PT400-Pt30
950 QT25-PT400-Pt30
900
HV0.05

850
800
750
700
650
0 100 200 300 400
Distance from worn surface (m)
Fig. 40. Microhardness evaluation from the worn surface to the interior of the discs in the radial
direction.

Even though the wear mechanisms were not evaluated in detail, the worn
surfaces were analysed by SEM (Fig. 41). The main observed feature is the
presence of cracks, which was reported as a characteristic of delamination wear
mechanism [49].

44
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 41. Analysis of wear mechanisms on the worn surface. Detection of cracks performing an
examination with both backscatter electron and secondary detectors (BED and SED) is possible
in samples QT165-400-30 (left) and 165-400-90 (right) respectively.

5. Discussion

5.1 Hardness-XRD-Impact Toughness

A general tendency for the partitioning temperature can be drawn; Fig. 42
shows the hardness, % of austenite and impact toughness data for a quenching
temperature of 165 ºC. It is possible to observe that the location in temperature
of the lowest levels of hardness (Fig. 42a) correlates with the partitioning
temperature applied when the highest quantities of retained austenite are
reached (Fig. 42c). Furthermore, these partitioning temperatures also maximize
the energy absorbed in a Charpy V-Notch test (Fig. 42b).

45
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 42. Colour maps for a quenching temperature of 165 ºC. The hardness (a), impact toughness
(b), and % of retained austenite (c) are shown as a function of the partitioning conditions.

The reason of this behaviour relies on the basis of the carbon diffusivity and
distance diffused by carbon atoms in austenite. The diffusivity of carbon has an
exponential increase when the temperature rises up (Eq. 9). In addition, Fig. 43
describes the distance diffused by carbon (Eq. 8) for the partitioning conditions
considered in this study. High partitioning temperatures (>400 ºC) have a
diffusion distance two orders of magnitude higher than low partitioning
temperatures (<280 ºC). Thus, it is possible to conclude that high partitioning
temperatures lead to faster carbon depletion in martensite and, not only a larger
amount of carbon is available to partition into austenite but also there is a faster
carbon enrichment and carbon content homogenization of the austenite.
Consequently, as there is a lower quantity of carbon distorting the unit cell of
martensite and the austenite stabilization is more effective, high partitioning
temperatures bring about a lower hardness and higher amount of retained
austenite. The maximization of the impact toughness is a result of the last two
factors, which make the material more tenacious.

46
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Diffusion Distance (m)

900 118
800 103

Partitioning Time (s)

700
88,8
600
74,1
500
59,5
400
44,9
300
30,3
200
100 15,6
1,00
250 300 350 400 450 500
Partitioning Temperature (ºC)
Fig. 43. Diffusion distance of a carbon atom for the partitioning conditions analysed.

The diffusivity of carbon can also explain the effect of time on the final amount
of retained austenite (Fig 34). When the partitioning is performed at high
temperatures, the carbon stabilization of austenite is accomplished fast and
then, an increment in time just leads to austenite consumption because of
competing reactions. Furthermore, the reduction of austenite is faster when the
quenching temperature is higher because the initial quantities of retained
austenite and initial martensite after quenching. When QT is increased, the
fraction of martensite transformed is lower, which in turn implies a lower
amount of carbon available to stabilize austenite during partitioning and larger
pools of austenite to distribute this carbon. The last two factors decrease the
stability of this austenite and are the main reason of the faster reduction. It can
be observed that higher carbon content is present in QT25 than in QT165 (0.99-
1.44 %wt and 0.38-1.32 %wt respectively). On the other hand, an increment in
time at partitioning temperatures below 280 ºC induces a larger carbon
enrichment in austenite and for this reason, extending the time is possible to
achieve higher amount of austenite. Similarly, fixing the time and the
quenching temperature, the higher carbon diffusivity explains the increment in
austenite fraction when the partitioning temperature is increased from 220 ºC
up to 280 ºC.

It should be mentioned that the impact toughness achieved for all conditions
are relatively low (2 J to 10 J). Poor impact toughness is thought to be an
intrinsic property of alloys in which large pools of blocky austenite and brittle
martensite are present. Blocky austenite may transform to martensite under
deformation easily when its size is large and the tendency of the martensite to
crack also increases with its size. Therefore, the size of austenite’s block should

47
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

be reduced below the size of phases that can initiate fracture, i.e. non-metallic
inclusions [36, 38].

5.2 Tensile Test

Best results regarding both UTS and fracture deformation were reached in
samples quenched until room temperature. The formation of fresh/untempered
martensite in QT165 and QT190 was confirmed via hardness analysis. This
martensite, containing at least 0.6 %C dissolved and distorting the unit cell,
results extremely brittle and restricts the plastic deformation of the samples
and, thus, the fracture takes place before macroscopic yield due to microcracks
that come up in the microstructure. On the other hand, the remarkable
combination of UTS ≈ 2.4-2.5 GPa and 4.8-5.7 % of fracture deformation are
reached performing the partitioning at high temperature (400 ºC).

It is well know that quenching and tempering steels containing high carbon
content, such as 52100 used in bearings, are brittle and that the elongation is
around 1-2%. This significant information is difficult to achieve and hardness,
compression or bending strength is reported instead [59].

Unfortunately, due to the inconveniences with the available extensometer, a

complete study about the effect of partitioning conditions and the analysis of
stability of austenite and incremental strain hardening coefficient could not be
carried out. In addition, it should be mentioned that fracture initiation took
place in pre-existing grinding defects, which compromises the results.

5.3 Rolling/Sliding Wear Test

A severe plastic deformation on the worn surface in the rolling direction was
the responsible effect of the hardening close to the worn surface, not only
because the dislocation density rises up but mainly because the plastic
deformation provides the driving force for the austenite-martensite
transformation and thus the hardness is sharply elevated. The austenite-
martensite transformation close to the worn surface can be confirmed via XRD
(Fig 44). The intensity of austenite peaks is reduced on the surface, suggesting
its transformation to martensite. This behaviour was also detected in CFB alloys
and it explains the better wear performance of the alloys compared with
conventional bainitic and perlitic steels.

48
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 44. XRD patterns before and after rolling/sliding test for the sample QT165-PT400-Pt90.

Fig. 45 shows the SWR as a function of hardness for different nanobainitic

alloys evaluated considering the same test conditions utilized in the present
study [37]. It is possible to observe that the wear performance of 06CV in CFB
and Q&P condition is similar. Moreover, the data obtained in the present study
for the three different Q&P conditions fits with the general tendency of SWR vs
HV established in “Nanobain”.

Considering the higher hardness acquired via Q&P, due to the fact that
martensite is harder than bainite, further investigations of wear performance of
different alloys and the analysis of the influence of quenching and partitioning
parameters (QT,PT,Pt) result promising.

1C 250
1CSi
0,8 220
2,0x10-4 08C250
06C
06CMo
06CNb
1,6x10-4
1Cr Si 220
06CV 220
SWR (mm3/N.m)

06CV 280
1,2x10-4 06CV-QT165-PT400-Pt30
06CV-QT165-PT400-Pt90
06CV-QT25-PT400-Pt30
8,0x10-5

4,0x10-5

0,0
510 540 570 600 630 660 690 720 750 780
HV
Fig. 45. Specific Wear Rates vs Hardness reported in the literature [37] for CFB alloys and the
Q&P data obtained in the present project.

49
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

6. Conclusions
The effect of the heat treatment parameters such as quenching temperature,
partitioning temperature and partitioning time on the microstructure and
mechanical properties have been investigated in a 06CV alloy. The influence of
these parameters have been described by analysing results of hardness, impact
toughness and XRD measurements. Maps of hardness, retained austenite, and
impact toughness for each quenching temperature as a function of partitioning
temperature and time were possible to assess. In addition, tensile properties for
some Q&P conditions were measured. Furthermore, the wear performance in a
dry rolling/sliding test was evaluated and the specific wear rate achieved by
Q&P is promising. The following conclusions can be drawn:

 Maps of hardness were developed for the three quenching temperatures

analysed. Even though samples quenched until room temperature have a
higher amount of initial martensite (which is the martensite generated
during the first controlled quenching), the highest levels of hardness are
achieved after quenching to 165 and 190 ℃. This behaviour confirms the
presence of fresh/untempered martensite after Q&P treatments with
using the last two quenching temperatures mentioned.

 High partitioning temperatures lead to faster carbon depletion in

martensite and thus, a larger amount of carbon is available to stabilize
austenite. In addition, the carbon enrichment and stabilization of
austenite is also faster due to the higher C diffusivity. Therefore, high
partitioning temperatures, for the three quenching temperatures
analysed, are able to retain a higher quantity of austenite, achieving
lower levels of hardness. The combination of a softer microstructure with
higher retained austenite content in these partitioning conditions
maximizes the energy absorbed in a Charpy V-Notch test.

 Regarding the effect of time, at high partitioning temperatures the

carbon stabilization of austenite is completed fast and thus, an increment
in time induces austenite decomposition through different competing
reactions. This austenite consumption is faster when the quenching
temperature is higher, due to the fact that an increase in quenching
temperature rises up the initial austenite to be stabilized during
partitioning, decreasing the carbon content available to do that (because
less fraction of martensite is present) and thus, decreasing the stability of
austenite against possible competing reactions. On the other hand, an
increment in time at partitioning temperatures below 280 ºC induces a

50
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

larger carbon enrichment in austenite and, for this reason, extending the
time is possible to achieve a higher amount of austenite.

 The best results of both tensile strength and fracture deformation are
reached in samples quenched until room temperature (QT25), showing
UTS as high as 2.5 GPa with fracture deformation of 5.7 % in the best
case. The presence of fresh martensite restricts the macroscopic yield
deformation in QT165 and QT190.

 The rolling/sliding wear performance of three Q&P conditions was

evaluated. The specific wear rates achieved are similar to the SWR
reported for 06CV in CFB state. In addition, quenching and partitioning
data fits with the general tendency of SWR vs. HV described in the
literature for CFB alloys.

6.1 Comparison between CFB and Q&P

CFB and Q&P are based on the same concept, a combination of a phase that
provides the strength (either carbide free nanobainite or martensite
respectively) and retained austenite in order to enhance ductility are achieved
via austempering at low temperature or a quenching step with the subsequent
partitioning. In Table 7 the results achieved in the present study via Q&P are
compared with previous results reported in the literature in CFB state for the
06CV alloy.

Table 7. Properties achieved in 06CV in CFB and Q&P state.

CFB 06CV
Q&P 06CV
[37]
Retained
24-31 7-27
Austenite (%)
Hardness
530-630 700-800
(HV)
Impact
5-25 2-11
Toughness (J)
UTS
2000 2500
(MPa)
Fracture
Deformation 12.4 5.7
(%)
SWR
5.42E-5 - 1.2E-5 5.64E-5 - 2.28E-5
(mm3/N.m)
Time of heat
12-24 h 2-10 min
treatment

The combination of a long time required in order to complete the bainite

formation and the incomplete reaction phenomenon leads to the possibility to

51
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

reach a higher quantity of retained austenite in CFB condition. On the other

hand, as it is mentioned in previous section, martensite is usually harder than
bainite and thus, Q&P may result in a stronger microstructure comparing with
the results reported in “nanobain”, with hardness levels around 700-800 HV
and UTS as high as 2.5 GPa. Nevertheless, the long austempering performed at
low temperatures involved in the CFB treatment leads not only to a finer
structure with higher retained austenite fraction but also it relieves any residual
stress in the material. For this reason, both energies absorbed in a Charpy
V-Notch test and fracture deformations are higher in CFB than Q&P.

The SWR achieved in a dry rolling/sliding test are similar for both CFB and
Q&P due to the presence of a substantial retained austenite content that leads to
a similar wear behaviour.

In conclusion, Q&P heat treatment may achieve an increase of about 500 MPa in
tensile strength when compared with CFB, whilst the latter could double both
fracture deformation and impact toughness.

7. Future Work
The future works that emerge from the present study are listed below:

 Further evaluations of the wear performance of different alloys treated

via Q&P and the influences of Q&P parameters could be a motivating
and relevant topic to work with.

 The elaboration of tensile property maps for 06CV reconsidering the

specimens design could be studied in parallel with the influences of
partitioning conditions on the austenite stability and the incremental
strain hardening coefficient.

 A refining of the prior austenite grain sizes (PAGs) by Q+Q&P, which in

turn would imply a refining of the retained austenite pools and
martensite packages and blocks, results in an interesting alternative that
could elevate the impact toughness.

52
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

8. References
[1] Tamarelli, C. M. 2011. The evolving use of advanced high-strength steels for
automotive applications. Southfield (MI): Steel Market Development Institute
(SMDI)

[2] Speer, J. G., De Moor, E., Findley, K. O., Matlock, D. K., De Cooman, B. C.,
and Edmonds, D. V. 2011. Analysis of microstructure evolution in quenching
and partitioning automotive sheet steel. Metallurgical and Materials
Transactions A. 42:3591

[3] Bouaziz, O., Zurob, H., and Huang, M. 2013. Driving force and logic of
development of advanced high strength steels for automotive applications. steel
research international. 84:937-947

[4] Speer, J. G., De Moor, E., and Clarke, A. J. 2015. Critical assessment 7:
quenching and partitioning. Materials Science and Technology. 31:3-9

[5] B. Kim;Sietsma, J., and Santofimia, M. J.;(2017). - The role of silicon in carbon
partitioning processes in martensite/austenite microstructures - Materials &
Design. 127:336-345

[6] Somani, M. C., Porter, D. A., Karjalainen, L. P., and Misra, R. 2014. On
various aspects of decomposition of austenite in a high-silicon steel during
quenching and partitioning. Metallurgical and Materials Transactions A.
45:1247-1257

[7] Speer, J. G., Assuno, F. C. R., Matlock, D. K., and Edmonds, D. V. 2005. The"
quenching and partitioning" process: background and recent progress.
Materials Research. 8:417-423

[8] Edmonds, D. V., He, K., Rizzo, F. C., De Cooman, B. C., Matlock, D. K., and
Speer, J. G. 2006. Quenching and partitioning martensite—A novel steel heat
treatment. Materials Science and Engineering: A. 438:25-34

[9] Clarke, A. J., Speer, J. G., Matlock, D. K., Rizzo, F. C., Edmonds, D. V., and
Santofimia, M. J. 2009. Influence of carbon partitioning kinetics on final
austenite fraction during quenching and partitioning. Scr.Mater. 61:149-152

[10] Koistinen, D. P., and Marburger, R. E. 1959. A general equation prescribing

the extent of the austenite-martensite transformation in pure iron-carbon alloys
and plain carbon steels. acta metallurgica. 7:59-60

[11] De Knijf, D., Petrov, R., Fjer, C., and Kestens, L. A. 2014. Effect of fresh
martensite on the stability of retained austenite in quenching and partitioning
steel. Materials Science and Engineering: A. 615:107-115

53
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

[12] Van Dijk, N. H., Butt, A. M., Zhao, L., Sietsma, J., Offerman, S. E., Wright, J.
P., and Van der Zwaag, S. 2005. Thermal stability of retained austenite in TRIP
steels studied by synchrotron X-ray diffraction during cooling. Acta Materialia.
53:5439-5447

[13] Garcia-Mateo, C., Caballero, F. G., Chao, J., Capdevila, C., and De Andres,
C. G. 2009. Mechanical stability of retained austenite during plastic deformation
of super high strength carbide free bainitic steels. J.Mater.Sci. 44:4617-4624

[14] Xiong, X. C., Chen, B., Huang, M. X., Wang, J. F., and Wang, L. 2013. The
effect of morphology on the stability of retained austenite in a quenched and
partitioned steel. Scr.Mater. 68:321-324

[15] Santofimia, M. J., Petrov, R. H., Zhao, L., and Sietsma, J. 2014.
Microstructural analysis of martensite constituents in quenching and
partitioning steels. Mater Charact. 92:91-95

[16] Santofimia, M. J., Zhao, L., and Sietsma, J. 2011. Overview of mechanisms
involved during the quenching and partitioning process in steels. Metallurgical
and Materials Transactions A. 42:3620

[17] De Moor, E., Lacroix, S., Clarke, A. J., Penning, J., and Speer, J. G. 2008.
Effect of retained austenite stabilized via quench and partitioning on the strain
hardening of martensitic steels. Metallurgical and Materials Transactions A.
39:2586

[18] Krauss, G. 2015. Steels: processing, structure, and performance. Asm

International

[19] D. T. Pierce;Coughlin, D. R.;Clarke, K. D.;De Moor, E.;Poplawsky,

J.;Williamson, D. L.;Mazumder, B.;Speer, J. G.;Hood, A., and Clarke, A. J.;(2018).
- Microstructural evolution during quenching and partitioning of 0.2C-1.5Mn-
1.3Si steels with Cr or Ni additions - Acta Materialia. 151:454-469

[20] Santofimia, M. J., Zhao, L., Petrov, R., Kwakernaak, C., Sloof, W. G., and
Sietsma, J. 2011. Microstructural development during the quenching and
partitioning process in a newly designed low-carbon steel. Acta Materialia.
59:6059-6068

[21] Forouzan, F., Vuorinen, E., and MÜcklich, F. 2017. Post weld-treatment of
laser welded AHSS by application of quenching and partitioning technique.
Journal of Materials Science and Engineering: A

[22] Kim, D. H., Speer, J. G., Kim, H. S., and De Cooman, B. C. 2009.
Observation of an isothermal transformation during quenching and
partitioning processing. Metallurgical and Materials Transactions A. 40:2048-
2060

54
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

[23] Yakubtsov, I. A., and Purdy, G. R. 2012. Analyses of transformation kinetics

of carbide-free bainite above and below the athermal martensite-start
temperature. Metallurgical and Materials Transactions A. 43:437-446

[24] Wang, M., Hell, J., and Tasan, C. C. 2017. Martensite size effects on damage
in quenching and partitioning steels. Scr.Mater. 138:1-5

[25] Edmonds, D. V., He, K., Miller, M. K., Rizzo, F. C., Clarke, A., Matlock, D.
K., and Speer, J. G. 2007. Microstructural features of'quenching and
partitioning': a new martensitic steel heat treatment. Pages 4819-4825 in:
Materials science forum. Anonymous . Trans Tech Publ

[26] De Moor, E., Speer, J. G., Matlock, D. K., Kwak, J., and Lee, S. 2011. Effect of
carbon and manganese on the quenching and partitioning response of CMnSi
steels. ISIJ Int. 51:137-144

[27] Tsuchiyama, T., Tobata, J., Tao, T., Nakada, N., and Takaki, S. 2012.
Quenching and partitioning treatment of a low-carbon martensitic stainless
steel. Materials Science and Engineering: A. 532:585-592

[28] Li, H. Y., Lu, X. W., Li, W. J., and Jin, X. J. 2010. Microstructure and
mechanical properties of an ultrahigh-strength 40SiMnNiCr steel during the
one-step quenching and partitioning process. Metallurgical and Materials
Transactions A. 41:1284-1300

[29] Yan, S., Liu, X., Liu, W. J., Lan, H., and Wu, H. 2015. Comparison on
mechanical properties and microstructure of a C–Mn–Si steel treated by
quenching and partitioning (Q&P) and quenching and tempering (Q&T)
processes. Materials Science and Engineering: A. 620:58-66

[30] Somani, M. C., Karjalainen, L. P., Porter, D. A., and Misra, R. 2012.
Evaluation of the behaviour and properties of a high-Si steel processed using
direct quenching and partitioning. Pages 2824-2829 in: Materials Science Forum.
Anonymous . Trans Tech Publ

[31] Wu, R., Li, W., Zhou, S., Zhong, Y., Wang, L., and Jin, X. 2014. Effect of
retained austenite on the fracture toughness of quenching and partitioning
(Q&P)-treated sheet steels. Metallurgical and Materials Transactions A. 45:1892-
1902

[32] Bagliani, E. P., Santofimia, M. J., Zhao, L., Sietsma, J., and Anelli, E. 2013.
Microstructure, tensile and toughness properties after quenching and
partitioning treatments of a medium-carbon steel. Materials Science and
Engineering: A. 559:486-495

55
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

[33] Caballero, F. G., Bhadeshia, H., Mawella, K., Jones, D. G., and Brown, P.
2001. Design of novel high strength bainitic steels: Part 1. Materials Science and
Technology. 17:512-516

[34] Caballero, F. G., Bhadeshia, H., Mawella, K., Jones, D. G., and Brown, P.
2001. Design of novel high strength bainitic steels: Part 2. Materials Science and
Technology. 17:517-522

[35] Bhadeshia, Harshad Kumar Dharamshi Hansraj, and Christian, J. W. 1990.

Bainite in steels. Metallurgical transactions A. 21:767-797

[36] Caballero, F. G., and Bhadeshia, H. 2004. Very strong bainite. Current
Opinion in Solid State and Materials Science. 8:251-257

[37] Sourmail, T., Smanio, V., Ziegler, C., Heuer, V., Kuntz, M., Caballero, F. G.,
Garcia-Mateo, C., Cornide, J., Elvira, R., Leiro, A, Vuorinen,E., and Teeri, T.
2013. Novel nanostructured bainitic steel grades to answer the need for high-
performance steel components (Nanobain). Research and Innovation, ed. 2013.
https://publications.europa.eu/en/publication-detail/-/publication/99671fbf-
f40c-4d07-9b6c-01d7dd17ad95/language-en

[38] Bhadeshia, H. 2010. Nanostructured bainite. Pages 3-18 in: Proceedings of

the Royal Society of London A: Mathematical, Physical and Engineering
Sciences. Anonymous . The Royal Society

[39] Garcia-Mateo, C., Sourmail, T., Caballero, F. G., Smanio, V., Kuntz, M.,
Ziegler, C., Leiro, A., Vuorinen, E., Elvira, R., and Teeri, T. 2014.
Nanostructured steel industrialisation: plausible reality. Materials Science and
Technology. 30:1071-1078

[40] Cornide, J., Garcia-Mateo, C., Capdevila, C., and Caballero, F. G. 2013. An
assessment of the contributing factors to the nanoscale structural refinement of
advanced bainitic steels. J.Alloys Compounds. 577:S47

[41] Long, X. Y., Kang, J., Lv, B., and Zhang, F. C. 2014. Carbide-free bainite in
medium carbon steel. Mater Des. 64:237-245

[42] Garcia-Mateo, C., and Caballero, F. G. 2005. Ultra-high-strength bainitic

steels. ISIJ Int. 45:1736-1740

[43] Handbook, A. 1992. Friction, lubrication and wear technology. ASM

International. 18:175-183

[44] Chattopadhyay, R. 2001. Surface wear: analysis, treatment, and prevention.

ASM international

56
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

[45] Devanathan, R., and Clayton, P. 1991. Rolling-sliding wear behavior of

three bainitic steels. Wear. 151:255-267

[46] Bhushan, B. 2000. Modern tribology handbook, two volume set. CRC press

[47] Bakshi, S. D., Leiro, A., Prakash, B., and Bhadeshia, H. 2014. Dry
rolling/sliding wear of nanostructured bainite. Wear. 316:70-78

[48] Sourmail, T., Caballero, F. G., García-Mateo, C., Smanio, V., Ziegler, C.,
Kuntz, M., Elvira, R., Leiro, A., Vuorinen, E., and Teeri, T. 2013. Evaluation of
potential of high Si high C steel nanostructured bainite for wear and fatigue
applications. Materials Science and Technology. 29:1166-1173

[49] Leiro, A., Kankanala, A., Vuorinen, E., and Prakash, B. 2011. Tribological
behaviour of carbide-free bainitic steel under dry rolling/sliding conditions.
Wear. 273:2-8

[50] Leiro, A., Vuorinen, E., Sundin, K., Prakash, B., Sourmail, T., Smanio, V.,
Caballero, F. G., Garcia-Mateo, C., and Elvira, R. 2013. Wear of nano-structured
carbide-free bainitic steels under dry rolling–sliding conditions. Wear. 298:42-47

[51] Speer, J., Matlock, D. K., De Cooman, B. C., and Schroth, J. G. 2003. Carbon
partitioning into austenite after martensite transformation. Acta materialia.
51:2611-2622

[52] Caballero, F. G., Bhadeshia, H., Mawella, K., Jones, D. G., and Brown, P.
2002. Very strong low temperature bainite. Materials science and technology.
18:279-284

[53] Porter, D., Easterling, K., Sherif, M. (2009). Phase Transformations in Metals
and Alloys, Third Edition (Revised Reprint). Boca Raton: CRC Press.
Taylor&Francis.

[54] Lee, S., Matlock, D. K., and Van Tyne, C. J. 2011. An empirical model for
carbon diffusion in austenite incorporating alloying element effects. ISIJ Int.
51:1903-1911

[55] De, A. K., Speer, J. G., and Matlock, D. K. 2003. Color tint-etching for
multiphase steels. Adv Mater Processes. 161:27-30

[56] Vol, A. H. 2004. 9: Metallography and Microstructures. ASM International,

Materials Park, OH

[57] Chen, X., and Vuorinen, E. 2009. In-situ high temperature X-ray studies of
austempering transformation in high silicon cast steel. ISIJ Int. 49:1220-1224

57
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

[58] A. Leiro. (2014). - Microstructure analysis of wear and fatigue in

austempered high-Si steels - Microstructure analysis of wear and fatigue in
austempered high-Si steels. PhD Thesis: http://www.diva-
portal.org/smash/record.jsf?pid=diva2%3A999219&dswid=8283.

[59] Bhadeshia, H. 2012. Steels for bearings. Progress in materials Science.

57:268-435

58
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Appendix A: An Optimization of Quenching Temperature to

Minimize the Banding Phenomena in Q&P Steels

Introduction
The effect of segregation on conventional steels has been largely analysed [1-4].
Macrosegregation and microsegregation (or interdentritic segregation) denote
compositional variations in the scale of the size of the specimen and the
secondary dendritic arms respectively [5]. Due to the sluggish diffusion of
substitutional atoms in steels the interdentritic chemical segregation cannot be
completely removed by hot work processing [1]. On the contrary, carbon atoms
that are interstitially dissolved in steel have a high diffusion coefficient that
allows the compositional gradients to be homogenised [2]. Nevertheless, as the
presence of other elements can locally modify the carbon activity, the
segregation of substitutional atoms can lead to carbon content variations. For
example, manganese and chromium decrease the carbon activity, and thus Mn
or Cr-rich areas can also result in a local increase of the carbon content. In
contrast, silicon, phosphorous and nickel increase the carbon activity and thus
reject it to other regions [2]. During hot rolling, the micro-segregation is aligned
into bands in the rolling direction generating areas that contain different
amounts of various elements. Consequently, a final microstructure exhibiting
bands of different microstructures (usually referred as banding) is triggered by
an interdentritic segregation aligned during hot rolling [1].

The effect of segregation inducing banding in Q&P process has been recently
investigated by HajyAkbary et. al [6]. In addition, in previous analysis by the
author of the present study, a microstructural banding was detected in 06CV
treated by quenching and partitioning and it was associated to segregation of
both Mn and Cr. In the following sections, an optimization of the quenching
temperature in order to avoid the banding phenomenon is proposed.

Materials and Methods

Decreasing the quenching temperature and generating a higher volume of
initial martensite in Mn-Cr rich and poor areas could reduce the banding and
gradient in mechanical properties due to the lower amount of fresh martensite
transforming during final cooling.

Thus, four different heat treatments were performed in order to evaluate the
local hardness (HV0.2) in the bands and the influences of the quenching
temperature on the difference in hardness (∆HV) between bands. The heat
treatment starts with austenitization at 890 °C for 60 minutes, followed by
quenching to 190 °C (QT190), 165 °C (QT165), 100°C (QT100) and 25 °C (QT25)

59
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

while the nominal martensite’s start temperature (Ms) is 223 °C. Subsequently,
samples are partitioned at 280 °C for 900 seconds and quenched to room
temperature (Fig. 1).

Fig. 1. Schematic illustration of experimental procedure.

Results
The optical microscopy analysis shows a significant microstructural banding on
0.6 %C-1.6 %Si-1.25 %Mn-1.75 %Cr-0.15 %Mo-0.12 %V (wt%) at low
magnification in samples quenched until 190 ºC and etched using LePera
reagent (Fig. 2). It is possible to distinguish differences between bright and dark
regions. The latter evidences the presence of a higher amount of initial
martensite that is etched as blue, whilst either retained austenite or fresh
martensite predominates in white regions.

60
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 2. Banding structure detected via OM. Etched with LePera.

Performing a line-scan by EDS in a direction that is transversal to the banding

(Fig. 3 blue line) segregation of both chromium and manganese was detected
(Fig. 4). None of the other elements of alloy has shown a tendency. The content
of Mn and Cr decreases constantly in the regions associated with higher
amount of initial martensite, while the peaks in the contents are located in the
regions exhibiting the presence of fresh martensite.

Fig. 3. Microstructural banding observed by SEM of sample QT190-PT250-Pt600. The blue line
shows the location where the line-scan was performed.

It is possible to observe that while the nominal composition of the steel has a
1.25 % and 1.75 % for Mn and Cr respectively, the analysis reports a variation
from 1 % up to 1.75 % in Mn and between 1.5-2.1 % for Cr content (Fig. 4). On

61
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

the other hand, a carbon alteration is expected due to the effect of Mn and Cr in
the carbon activity.

Fig. 4. Chromium and manganese compositional variations in the direction perpendicular to the
banding. The content of Mn and Cr decreases constantly in regions associated with higher
amount of initial martensite.

As these elements of alloy decrease the martensite start temperature, a bands

formation are associated with a local difference in Ms between Mn-Cr rich and
poor regions. Thus, the following mechanism was proposed (Fig. 5).

Fig. 5. Schematic illustration emphasizing the effect of segregation on quenching and

partitioning process.

Due to the sluggish diffusion of Mn and Cr atoms, the segregation is not

eliminated during austenitization. In the first quenching, a higher volume of
martensite is transformed in the Mn-Cr poor regions because of its superior Ms.
During partitioning this initial martensite losses carbon and becomes softer, the
final hardness is determined by partitioning conditions (temperature and time).

62
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Additionally, the carbon partitioning to the surrounding austenite is insufficient

to stabilize all the austenite and thus, during final cooling more fresh martensite
is formed in Mn-Cr rich areas because of its higher amount of austenite to
transform.

Since it is not possible to obtain a reliable quantification of carbon concentration

by using EDS, its expected segregation profile has been modelled by
Thermocalc-Dictra using TCEFE9 and MOBFE3 database. Two austenite phases
were considering, one containing 1% of Mn and 1.5 % of Cr, and the other with
compositions of 1.75 % and 2.1 % for Mn and Cr respectively. The other
elements were considered equal to the nominal composition. A treatment at
high temperature (890 ºC-60 min) was simulated in order to allow C diffusion in
austenite. Thus, results show that carbon content varies from 0,57 % up to
0,615 % from Mn depleted regions toward enriched regions, which leads to a
local difference in Ms up to 40 ºC.

Considering this local difference in Ms, the fraction of initial martensite as a

function of the quenching temperature can be calculated (Fig. 6). It is clearly
seen in green that while decreasing the quenching temperature, there is also a
reduction in the differences of initial martensite fraction between the two
regions.

Fig. 6. Initial martensite fraction in the depleted and enriched regions, and the difference
between them (∆fm), as a function of the quenching temperature.

63
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Table 1 shows the hardness in the depleted and enriched regions, together with
the difference in hardness between the two bands for each quenching
temperature applied. It is possible to observe that a reduction in the quenching
temperature leads to a decrease in ∆HV (Table 1-Fig. 7).

Table 8 Hardness of Mn-Cr enriched and depleted region for each quenching temperature.
HV0.2 of Mn-Cr HV0.2 of Mn-Cr
QT (°C) HV0.2
enriched region depleted region
190 845 703 142
165 845 774 71
100 742 726 16
25 720 747 27

160
140
120
100
80
HV

60
40
20
0
-20
-40
20 40 60 80 100 120 140 160 180 200 220
Quenching Temperature (ºC)

Fig. 7 Difference in hardness between depleted and enriched regions as a function of the
quenching temperature.

Discussion
In Table 1 and Fig. 6 is possible to elucidate that the enriched region represents
the hardest band in the material for the highest quenching temperatures, even
though it has a lower amount of initial martensite generated during the first
quenching. That means that a considerable higher amount of austenite remains
in the Mn-Cr enriched bands. The martensite transformed becomes softer
during partitioning, and the final hardness is determined by partitioning
conditions. During the final cooling, a higher fraction of fresh martensite is
formed in Mn-Cr rich regions due to the superior amount of austenite to
transform and because this austenite is less stable since the carbon available to

64
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

partition from martensite is lower and it is distributed in larger pools (because

less martensite was transformed during the first quenching).

The inversion of the hardness that can be seen in Fig. 7. (negative ∆HV) is
related to the smaller chance to obtain fresh martensite during the last cooling
when the QT is decreased. When that happens, a higher amount of initial
martensite is transformed during the first quenching; it implies a lower amount
of austenite in the microstructure and a higher carbon content stabilizing this
austenite during the partitioning (the higher carbon content available is linked
with the increase in the initial martensite after the first quenching). Because of
that, when the possibility to transform fresh martensite during the final cooling
is minimized, the Mn-Cr enriched region retain a higher amount of austenite
(because it is subjected to a lower undercooling and it has a higher amount of
austenite to stabilize). Thus, the Mn-Cr enriched region starts to represent the
softest band in the microstructure. In this particular composition and levels of
segregation, this inversion in hardness takes place between 100 ºC and 25 ºC but
it should be evaluated for each particular situation

In addition, it can be seen that decreasing the quenching temperature leads to a

reduction in the ∆HV between bands (Fig. 7). The reason relies on the
exponential effect of temperature in the martensitic transformation, which
means that a drop in the quenching temperature sharply reduces the difference
in fraction of martensite generated during first quenching (Fig. 6). Additionally,
comparing the difference between the hardness of the bands (∆HV) with the
difference in the fraction of martensite (∆fm) shows a linear relationship (Fig. 8).
As the carbon content is the preponderant parameter to define the hardness in
steels and also the substitutional elements of alloys may affect, it is expected
that the slope of this line depends on the chemical composition of the steel and
degree of segregation, which could affect the volume fraction of different
phases.

65
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 8. Relation between difference in fraction of martensite (∆fm) and hardness (∆HV) between
bands.
Linear relation showed in Fig. 8 illustrates that, if the levels of segregation in
bands are identified, the fraction of initial martensite generated in the two
bands after the first quenching is meaningful information that can be used to
optimize the QT in order to avoid the formation of bands with difference in
mechanical properties. In conclusion, when ∆fm is ~0.1, ∆HV is almost
insignificant.

Conclusions
The effect of micro-segregation of substitutional elements on the microstructure
and mechanical properties after Q&P has been investigated. The banding
phenomenon has been described through the existence of two different regions
with a depletion and enrichment in these substitutional elements of alloys (Mn
and Cr), which in turn implies two regions with a local difference in M s
temperature. The prevention of banded microstructure results of high interest
because it leads to a non-homogeneity and anisotropy of mechanical properties.
The following conclusions can be drawn:

a) Difference in Ms temperature between the bands is extended up to 40 ºC

in this case.

b) The difference in hardness (∆HV) between the regions decreases when

the quenching temperature is reduced. It occurs due to the decrease in
the difference fraction of initial martensite between regions (∆fm)
transformed after quenching (Fig. 6). ∆HV starts to be negligible when
∆fm is~0.1. The relation between ∆HV and ∆fm is linear, and the slope is
expected to depend on the chemical composition and the level of
segregation present in the material.

66
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

c) The hardness in the Mn-Cr enriched region is higher than the hardness
in the depleted region until a certain quenching temperature, where the
inversion in hardness comes about. At high quenching temperatures the
microstructure in the Mn-Cr enriched region consists of a considerable
amount of fresh martensite, but when the quenching temperature is
reduced, it is possible to retain more austenite in this region and it
becomes the softest region.

d) Although a reduction in the quenching temperature leads to a decrease

in the banding phenomenon, the optimum quenching temperature is
represented by considering the balance between the banding formation
and the possibility to retain a considerable fraction of austenite to
improve the ductility. Thus, when bands with different level of elements
of alloy are present, the quenching temperature should not be high to
prevent the banding formation. But, on the other hand, it must be high
enough to obtain a significant amount of retained austenite after the heat
treatment. For this reason, the optimum quenching temperature may be
evaluated for each particular chemical composition and level of
segregation.

References

[1] Krauss, G. 2003. Solidification, segregation, and banding in carbon and alloy
steels. Metallurgical and Materials Transactions B. 34:781-792

[2] Krauss, G. 2015. Steels: processing, structure, and performance. ASM

International. Second Edition

[3] Penha, R. N., Vatavuk, J., Couto, A. A., de Lima Pereira, Silvio Andrè, de
Sousa, S. A., and de CF Canale, L. 2015. Effect of chemical banding on the local
hardenability in AISI 4340 steel bar. Eng.Failure Anal. 53:59-68

[4] Verhoeven, J. D. 2000. A review of microsegregation induced banding

phenomena in steels. Journal of materials engineering and performance. 9:286-
296

[5] Porter, D., Easterling, K., Sherif, M. (2009). Phase Transformations in Metals
and Alloys, Third Edition (Revised Reprint). Boca Raton: CRC Press.
Taylor&Francis.

[6] HajyAkbary, F., Sietsma, J., Petrov, R. H., Miyamoto, G., Furuhara, T., and
Santofimia, M. J. 2017. A quantitative investigation of the effect of Mn

67
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

segregation on microstructural properties of quenching and partitioning steels.

Scr.Mater. 137:27-30

Appendix B – Problems with Tensile Test Specimens

Several complications came about in the preparation of tensile test specimens.
First, a sub-size sample with a gage length of 15 mm and ɸ6 mm was designed
and heat treated to be tested in Gleebe 3800. After heat treatment, samples were
sent to an external workshop located in Skellefteå in order to perform the
grinding and remove the decarburized layer. It took 3 weeks to receive samples
back, which means that is a time-consuming stage. Furthermore, during
experiment specimens broke down from thread (Fig. 1).

Fig. 1. Sub-size specimen of 6mm diameter tested in Gleebe 3800 at room temperature. Fracture
took place along thread.

Considering the hardness, the tensile strength expected should be ≈2500 MPa.
Then, since it is possible to record the applied force in Gleebe 3800, the
calculation of the stress in the fractured surface was executed and the minimum
stress concentration factor in the thread was found dividing 2500 MPa by the
latter mentioned stress. Thus, a minimum stress concentration factor of 5 was
considered to redesign the dimensions of the gage length. Previously machined
samples were sent again to LTU workshop to decrease the diameter down to
3,4 mm. After that, the heat treatment was performed and the posterior
grinding was also carried out in Skellefteå. It is important to mention that each
grinding stage required 3 weeks, while the machining at LTU workshop and
heat treatment took 2,5 weeks in total. The aspect of samples as-machined, after
heat treatment and after grinding can be observed in Fig.2.

68
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

Fig. 2. Sample as-machined, after heat treatment and after grinding.

Although with the new dimensions samples broke down in the gage length, it is
possible to observe that the fracture comes about close to the initiation of the
gage length in almost all the specimens, where a grinding defect could be
perceived (Fig. 3). This grinding defect could have also caused a stress
concentration and affect the measure of UTS and ductility.

Fig. 3 Examples of fracture initiation in a pre-existing grinding defect in sample QT25-PT400-

Pt30 (left) and QT165-PT400-Pt60 (right)

69
Investigation of the Effect of Different “Q&P” Parameters on the Mechanical Properties of AHSS
Borasi, Luciano

The extensometer available and utilized does not represent the best option to
measure strain at room temperature, due to the fact that it is designed to
operate at high temperature. Fig. 4a-b show the changes in the extensometer
along the time of the experiment and the stress-strain plot respectively. The
circle emphasized a slip in the extensometer that happened in some
experiments and thus, the extensometer does not record any alteration in length
during this period (Fig. 4a). For this reason, the stress-strain plot has a sharp
increase in stress without the respectively variation in strain (Fig 4b). In
addition, the glass fragment of the extensometer could not withstand the elastic
energy released after the fracture of the sample and some of them brown down.

Fig. 4. The deformation recorded in the extensometer during time (left) and the respectively
stress-strain curve. An sliding in the extensometer can be detected. Sample QT165-PT400-Pt60.

In conclusion, values of strain were obtained after the test, measuring the
alteration in distance between two lines. However, since the deformation is
extremely low in some cases (≈1 %) and hence the change in distance is also
low, the values of strain are not totally reliable. The manufacturing of a new
shape for the samples that can be tested in a universal testing machine is highly
recommended as a future work. Regarding the design of the specimens, shapes
utilized for cast iron or other brittle material could be considered.

70