See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/305778568

A review on the effect of microstructure, texture and inclusion on Charpy

impact transition behaviour of low- carbon ferritic steels

Article · March 2016

CITATION READS

1 1,650

6 authors, including:

Pranabananda Modak Abhijit Ghosh

Indian Institute of Technology Kharagpur Indian Institute of Science
7 PUBLICATIONS   30 CITATIONS    20 PUBLICATIONS   237 CITATIONS   

SEE PROFILE SEE PROFILE

Nirmalya Rarhi Vinod Kumar

Steel Authority of India Ltd (SAIL)
3 PUBLICATIONS   5 CITATIONS   
112 PUBLICATIONS   323 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

High Strength Steels View project

DMR-254 View project

All content following this page was uploaded by Pranabananda Modak on 03 August 2016.

The user has requested enhancement of the downloaded file.

 A review on the effect of microstructure, texture and
inclusion on Charpy impact transition behaviour of low-
carbon ferritic steels
P. Modak1, A. Ghosh1, N. Rarhi2, Vinod Kumar3, R. Balamuralikrishnan2,
D. Chakrabarti1*
1 Department of Metallurgical and Materials Engineering, Indian Institute of Technology
Kharagpur, 721 302, Kharagpur, West Bengal, India.
2
Defence Metallurgical Research Laboratory, Kanchanbagh, P.O. Hyderabad, Andhra
Pradesh, 500058, India.
3 Research and Development Certre for Iron and Steel (RDCIS), Steel Authority of India

Limited, Ispat Bhawan, Doranda, Ranchi - 834002, Jharkhand, India.

*Corresponding Author: Dr. Debalay Chakrabarti, Department of Metallurgical and Materials

Engg., IIT Kharagpur, 721 302, Kharagpur, West Bengal, E-mail: India. debalay@gmail.com;
Phone: 03222-283282 (Office).

Abstract:

The paper presents an extensive review on the effect of microstructural parameters and
crystallographic texture on Charpy impact toughness and ductile-brittle transition temperature
(DBTT) of low-carbon ferritic steels. Specified level of Charpy impact toughness is an
essential requirement for the steel plates / steel components used in structural, linepipe,
automotive, naval and defence applications. Special emphasis has been given to understand
the effect of ferrite grain size distribution, micro-texture and particles / inclusions on impact
toughness. Combined effect of microstructure, inclusion, and crystallographic texture on
Charpy impact properties has been studied after different finish rolling (935 °C-650 °C) and
normalizing (1250 °C-940 °C) treatments of low-carbon ferritic steels. Finish rolling within
the austenite-ferrite two-phase region leads to the formation of low-angle boundaries, which
strengthen ferrite matrix but are ineffective in restricting the cleavage crack propagation;
thereby deteriorate the upper shelf energy (USE) and increase the ductile to brittle transition
temperature (DBTT). The presence of coarse cuboidal TiN particles (>1µm) also increase the
DBTT, whereas, stringer shaped MnS inclusions deteriorate the USE. In spite of the presence
of large TiN particles, refinement in ‘effective grain size’ of ferrite can improve the impact
toughness. Finish rolling just above the austenite to ferrite transformation start temperature
(820 C) or normalizing of as-rolled plates at low austenitization temperature (940 C)
develop fine strain-free ferrite grains with small ‘effective grain size’, and therefore, can be
recommended for achieving high USE and low DBTT. The severity of delamination on the
fracture surface of Charpy impact tested samples of low-carbon steel has been found to be
dependent on finish rolling temperature and texture.

Keywords: Low-carbon ferritic Steel, Crystallographic texture, Inclusions, Charpy impact

testing, Effective grain size, Fissure, Anisotropy.

1
1. Charpy impact testing of thermo-mechanically treated low-carbon
ferritic steels
Excellent combination of high-strength and high impact toughness is an essential
requirement of steel plates used for linepipe, construction, pressure vessel, naval and defence
applications [1, 2]. The same combination of properties is also required in hot-forged steels,
used for automotive engine and load bearing components, such as, front axle, crankshaft and
connecting rod. Increase in strength leads to the deterioration in impact toughness [1, 3-5].
Thermo-mechanically controlled rolled (TMCR) high-strength low-alloy (HSLA) steels are
used for meeting those requirements. TMCR rolled HSLA steels are more economical and
show better bendability and weldability than high-alloyed and hot-rolled quenched and
tempered (Q&T) steels [1, 2]. Extensive research has been carried out over the past four
decades for improving the strength-impact toughness combination of TMCR rolled HSLA
grades. However, steel industries often find it difficult to meet the desired impact toughness
as set in quality standard for high-strength steel grades.
Through extensive grain refinement TMCR not only improves strength but also
improves the impact toughness (at sub-zero temperatures) of steels [1, 2]. Industrially the
impact toughness of steel and its transition over temperature is determined by the Charpy
impact testing [3-5]. In Charpy testing a notched bar sample is loaded typically at high strain
rates (3-6 m/s) over a range of test temperature. The samples are generally prepared along T-L
orientation (Transverse-Longitudinal) with respect to the rolled plates as shown in Fig. 1 [3-
6].

Fig. 1: Size and orientation of the Charpy impact test specimens with respect to the rolled
plate.

The test results are presented in the form of impact transition curves, either in terms of
fracture energy or fracture surface appearance, Fig. 2. Impact transition curves show the
transition in fracture behaviour over temperature: (i) ductile fracture at higher temperature by
a fibrous mechanism provides high fracture energy and (ii) brittle fracture at lower
temperature by cleavage mechanism provides low fracture energy. Fracture transition
behaviour of steel can be characterised by impact transition temperature (ITT), which can be
defined in various ways as mentioned below, Fig. 2 [3-5]:
 27J-ITT: Temperature corresponding to 27 J impact energy absorbed,
 Fracture appearance transition temperature (FATT): Temperature corresponding to
50 % cleavage and 50 % fibrous fracture surface appearance,
 Ductile-brittle transition temperature (DBTT): Temperature corresponding to the
mean fracture energy between upper shelf energy (USE) and the lower shelf energy
(LSE), Fig. 2.

2
 200 100
Cleavage USE

150 75

Cleavage fracture (%age)

DBTT =

Fracture energy (Joule)

½ (USE+LSE)

100 50

FATT
50 (50 % Fibrous) 25

27
LSE
27 J-ITT Fibrous
0 0
-100 -80 -60 -40 -20 0 20 40
Temperature ( C)

Fig. 2: Typical impact energy transition curve (solid line) and fracture appearance (fibrous-
cleavage) transition curve (dashed line) for ferritic steel. USE represents ‘upper shelf
energy’ and LSE represents ‘lower shelf energy’ for impact fracture. Various ductile
to brittle transition temperatures like 27 J-ITT, FATT and DBTT are indicated on the
figure.

2. Mechanism of ductile and brittle fracture during impact testing

2.1. Temperature dependence of the fracture mechanism
In Body Centred Cubic (BCC) metals, brittle ‘cleavage’ fracture occurs by the trans-
granular de-cohesion of {001} planes [3, 4]. Low temperature, high strain rate and plain strain
conditions are favourable for the brittle fracture. Due to asymmetric stress distribution around
a dislocation, the glide stress, i.e. the Peierls-Nabarro stress is higher in BCC metals
compared to the Face Centred Cubic (FCC) metals [3, 4]. On the other hand BCC metals have
only twelve primary slip systems. Therefore, in BCC metal, the contribution of the glide stress
is higher compared to the dislocation interaction stress on the yield strength. Now, the Peierls-
Nabarro stress is sensitive to the temperature, therefore, the yield strength of BCC metal is
strongly dependent on the temperature.

Fig. 3: Schematic diagram showing the variation of yield stress and fracture stress with
temperature. TGY denotes the general yield temperature.

In the schematic diagram in Fig. 3, the variation in yield stress and fracture stress with
temperature has been shown. The temperature (TGY) at which yield stress intersects with the
fracture stress is known as the general yield temperature [3]. The significance of this
temperature is, brittle fracture occurs below this temperature without yielding. In BCC metals,
due to the high temperature sensitivity of yield strength, cleavage fracture occurs below the
transition temperature. The fracture stress is also a crucial parameter and depending on the

3
fracture mechanism, the cleavage fracture stress varies. Cleavage fracture surface can be
characterised by the presence of river lines (showing the crack paths) and cleavage facets
(section of grains on the fracture surface) on the flat fracture surface. Fibrous fracture surface,
on the other hand, can be characterised by the presence of micro-voids on generally slanting
fracture surface. Typical ductile and brittle fracture surfaces obtained from Charpy impact
testing of HSLA steel at different test temperatures is presented in Fig. 4.

(a) (b)

Fig. 4: Typical features observed on the fracture surfaces of Charpy impact tested low-
carbon ferritic steels: (a) ductile fracture surface showing dimples and (b) brittle
fracture surface showing cleavage facets.

2.2. Ductile fracture mechanism

A large amount of plastic deformation is associated with ductile fracture which causes
high energy consumption during this process. In general, ductile failure occurs by three steps:
(i) void initiation, (ii) void growth and (iii) void coalescence, Fig. 5 [3, 4]. Low strain rate,
high testing temperature and plane stress conditions are favourable for ductile fracture. Void
initiation occurs by either separation of the interfaces of the inclusions like, MnS, second
phase particles or cracking of particles. Depending on the particle size and shape, the void
size and shape varies. The void growth occurs by the plastic deformation of the matrix
surrounding the particle / inclusion. It is the most energy consuming process among the three
stages of ductile failure and higher the void growth, higher the USE. The extent of plastic
deformation depends on the shape, size, distribution of particles, inter-void distance and also
on the deformability of the matrix [7]. After a certain amount of void growth, plastic
instability sets in the ligament between the two adjacent voids, leading to the ligament
necking and void coalescence [7-9]. McClintock [8, 9] suggested that such plastic instabilities
develop in the form of localized shear bands and the condition behind the formation of shear
bands can be expressed by the following equation [7-9]:
1  d   vf  2
  KF     1
2

  d   1  v 
f
……eqn. 1
 d 
 
where, σ is the flow stress in terms of true stress,  d   is the strain-hardening rate,
vf is the volume fraction and  is the aspect ratio of the inclusions, K is a constant and F is the
hole-growth factor. Eqn. 1 indicates that the extent of void growth and hence, the impact
energy absorption capacity of the steel not only depends on the inclusions parameters, i.e.
volume fraction and the aspect ratio of inclusions, but also on the strength and the strain-
hardening ability of the ferrite matrix. Lower initial yield strength and higher strain hardening
ability resists strain localization by the shear band formation. This promotes void growth and
therefore, increases the capacity to absorbed impact energy.

4
 Void initiation Void growth Void coalescence
Fig. 5: Different stages of ductile fracture around elongated inclusions (dark).

2.3. Brittle fracture micro-mechanism

The process of cleavage fracture in steels involves three critical steps as shown in Fig. 6 [10-
14]:
(i) In the first step, micro-crack nucleates at some microstructural inhomogeneities,
such as large inclusions (i.e. TiN), hard phases (pearlite, MA constituents etc),
grain boundary carbides, dislocation pile up and twins.
(ii) Next step is the propagation of this micro-crack across the particle-matrix
interface.
(iii) In the final stage, the micro-crack propagates across the first matrix-matrix
boundary.
The conditions that dominate each of these processes typically vary with microstructure,
temperature, stress and strain distribution ahead of a notch / pre-crack tip [12-14].

TiN

Step-1 Step-2 Step-3

Fig. 6: Schematic diagram showing different steps necessary for particle induced cleavage
fracture: Step-1: nucleation of sharp micro-crack from hard and brittle (TiN) particle,
Step-2: propagation of micro-crack across particle-matrix interface, Step-3:
propagation of micro-crack across matrix-matrix boundaries.

Several models were proposed to describe the mechanism of cleavage fracture from
the relation between cleavage origination stress (f) and the critical flaw size [15-24].
Considering those models and the experimental studies on cleavage fracture, and depending
on the critical step for fracture: either the propagation of a grain sized (D) micro-crack to the

5
neighbouring grain (changing its orientation) or the propagation of a particle sized (t’) micro-
crack to the neighbouring matrix [13, 14, 18, 19, 21, 22]; the fracture stress can be
summarised in the following form:
1 1
 
f  D 2
or  f  t  2
.….eqn. 2

The effect of carbide size on fracture can also be related indirectly to the grain size as
Curry and Knott [13] observed that the largest carbide size in the microstructure can be
proportional to the ferrite grain size in annealed or normalised mild steels. Recent studies on
microalloyed steels indicate that the presence of a large TiN particle in a large grain makes
the steel more prone to fracture [11, 25, 26].
Therefore, the combination of large second phase particle along with large grain size
is detrimental for fracture. All the above conditions can be represented by a normalised
Griffith-Orowan type equation [11-16, 24-26], that considers the plastic deformation at the
crack-tip before crack propagation, in the following form:
1
 4 E p  2
f  ….eqn. 3
  1   .x 
2 
where, x is the length of the micro-crack that causes failure, E is the Young’s modulus,  is
the Poisson’s ratio. For grain sized micro-crack: x = D and p = mm (can vary between 50
J/m2 to 300 J/m2 [11-14, 24-28]) i.e. the effective surface energy for matrix-martix interface.
For particle size micro-crack: x = t  and p = pm (usually in the range 7-20 J/m2 [3, 13, 24,
27]), i.e. that effective surface energy for particle-matrix interface.

2.4. Empirical equations for the prediction of impact transition temperatures

The effect of microstructural parameters on the impact transition temperatures of low-
carbon ferritic steels can be quantitatively explained in view of the established empirical
equations. Impact transition temperature (ITT) is known to be proportional to the inverse
square root of grain size (D-1/2) following the Cottrell-Petch relation [20, 24, 25]:
1

ITT  A  B.D 2
…..eqn. 4
where, ITT can be 27J-ITT, FATT or DBTT and A and B are constants.
Various researchers have developed quantitative relationships by linear (or other) regression
analysis to predict different Charpy impact transition temperatures (ITT), such as 50%-ITT
(FATT) and 27J-ITT for steels considering the influence of factors such as grain size (D),
carbide thickness (t), pearlite content (%-pearlite), precipitation hardening effects of alloying
elements (Y), cleanliness and embrittlement effects.
1) The Gladman-Pickering regression equation [1, 29] for the 50%-ITT (in C), which is
valid for metallographic two-dimensional (2-D) grain size, D (in mm), in the composition
range 0-0.20 C, 0-0.013 N and more than 0.2 Si is:
50%  ITT  19  44wt % Si   700wt % N f 2  2.2% pearlite  11.5D 
1 1

2

….eqn. 5
where, Nf is the free nitrogen content and %pearlite is the pearlite content (up to 30 %
pearlite) in ferrite-pearlite steel.
2) Mintz et al. [30, 31] developed regression relationships between 27J-ITT or 50%-ITT and
microstructural parameters such as, grain diameter, D (in mm), pearlite content and grain
boundary carbide thickness, t  (in m), for microalloyed ferrite-pearlite steels. The
relationships are:

6
  
1 1


27J  ITT C  173t  8.3D   0.37Y   42

2 2
0
.…eqn. 6

50%  ITT  C   112t

1 1


 13.7D   15% pearlite

 0.43Y   20
2 2 0.33
0

…..eqn. 7
where, Y is the precipitation hardening component in MPa, which is calculated from the
differences between the measured yield strength (y in MPa) at room temperature and the
non-precipitation hardened yield strength (np_ys) (in MPa) determined from the following
equation:
1

 np _ ys  43.1wt % Mn   83( wt % Si)  15.4( D ) 2
 1540( wt % N f )  105
…..eqn. 8
The range of composition for which Mintz’s relations apply are: 0.11-0.20 C, 0.63-1.56 Mn,
0.02-0.49 Si, 0.003-0.021 N (total), Nb up to 0.071, V up to 0.20, Ti up to 0.16 and Al (in
solution) up to 0.12. In Nb-microalloyed steels, Bhattacharjee et al. [32, 33] observed the
crack initiation from NbC particles, instead of grain boundary carbides, and hence, considered
NbC particle size as t’ for ITT prediction. The above mentioned equations, however, did not
consider the effect of two very important parameters i.e. (i) the ferrite grain size distribution
and (ii) the effect of crystallographic texture on impact toughness and transition temperature.
Those aspects have been discussed in subsequent sections in detail.

3. Effect of processing and microstructural parameters on the Charpy

impact toughness
Commercial TMCR microalloyed steel grades show higher Charpy energy (usually
more than 100 J at –40 C) compared to plain-C-steels or normalised microalloyed steels
(usually less than 30 J at –40 C) [2]. Addition of different microalloying elements such as,
Ti, Nb and V improves the strength of low-carbon ferritic steel by their contribution towards
grain refinement and precipitation hardening. Fine-scale Nb and V precipitates form during
rolling or final cooling, whereas, undissolved TiN particles control the austenite grain size
during reheating and welding treatment [1]. However, depending on the concentration of Ti
and N in steel, coarse and cuboidal TiN particles of several micrometer in size can form,
which act as the potential sites for cleavage crack initiation and thereby, deteriorates the low
temperature impact toughness [1, 10-12, 27]. During steel making process, if sulphur level is
not properly controlled then large MnS inclusions can also form during solidification. Soft
MnS inclusions elongate during the subsequent hot-rolling process, which deteriorate the
ductility and impact toughness of the steel and cause anisotropy in properties [7].
It is also difficult to completely remove the harmful inclusions, such as, coarse TiN
and MnS from the steel. Hence, the challenge is to enhance the impact toughness of the high-
strength steel (especially at low temperature), in spite of the presence of non-metallic
inclusions. This can only be done through proper control over the microstructural parameters
of the steel. The beneficial effect of ferrite grain size on both strength and toughness of steel
is well known [1-5]. The ferrite grain sizes in low-carbon steel, however, not necessarily
remain uniform and a large grain size variation has also been noticed in TMCR rolled steels
[32-35], Fig. 7. Understanding the effect of such grain size variation on Charpy impact
toughness and the impact transition temperature is extremely important for commercial
processing of high strength steels with high impact toughness. Besides that, the role of
crystallographic texture on the Charpy impact transition behaviour of ferritic steels is also not
well understood. The following sections provide a review on the effect of grain size variation
and crystallographic texture on the impact transition phenomenon in low-carbon ferritic
steels.

7
 Typical processing schedule used in Thermo-mechanical controlled rolling (TMCR) is
shown in Fig. 8. In order to achieve the desired mechanical properties, TMCR treatment can
be applied using different finish rolling temperatures [1]. If the desired impact toughness is
not achieved even after controlled rolling, steel plates are also subjected to normalizing
treatment [36]. For improving the strength-toughness combination of existing steel grades, the
effect of finish rolling temperature and normalizing temperature on microstructure, texture
and impact toughness is necessary to understand.

(a) (b)

Fig. 7: Bimodal ferrite grain structures developed in TMCR rolled Nb-microalloyed steel
plates for structural application (coarse grain regions are arrowed).

1200ºC

1150ºC
Temperature →

Roughing

950ºC
TNR

Ar3 Finishing

Ar1
Rolling Air-cooling Normalizing

time →
Fig. 8: Typical thermo-mechanical processing schedule used for industrial processing of
HSLA steels.

4. Effect of ferrite grain structure and meso-texture on impact

toughness

For a duplex grain structure comprising of coarse polygonal ferrite grains ( 17-30
m) and fine acicular ferrite grains ( 0-16 m) in TMCR-line pipe steel (0.5 C, 0.06 Nb,
0.011 S) Shehata and Boyd [37] have reported that 27J-ITT correlates with a measure of the
average polygonal ferrite grain size (the coarse grains covering top 25 % of the grain-size
distribution), and not with the mean grain size (considering both coarse-polygonal and fine-
acicular grains). The average polygonal grain sizes were 3-4 times larger than the
corresponding mean grain sizes. Bhattacharjee et al. [32, 33] have also reported that the
prediction of ITT, using eqn. 5 - eqn. 8, on the basis of average metallographic grain size has
not been successful for steels with duplex ferrite grain structure and mesotexture.

8
 Bouyne et al. [38] and Kim et al. [39] have shown that the grain size matches facet
size when groups of grains with low-angle boundaries are considered to be a single grain, the
‘effective grain’, ignoring boundaries with angles less than a threshold value of 15-18. Such
groups of grains (with low-angle boundaries) can develop during TMCR by a local texture
effect, known as ‘mesotexture’, and that can affect the fracture properties by influencing the
crystallographic unit of fracture. Transformation of a deformed austenite grain into multiple
ferrite grains may result in low-grain boundary misorientation angles between the neighboring
ferrite-grains, owing to the specific orientation relationship between austenite and ferrite,
which is termed as ‘meso-texture’ [32-35, 38, 39]. Low-angle boundaries are ineffective in
resisting cleavage crack propagation and hence, a single cleavage facet can be comprised of
multiple ferrite grains [32-35]. Bhattacharjee et al. [32, 33] selected a 12 misorientation
angle as the threshold to calculate the ‘effective grain size’ in TMCR microalloyed steel with
bimodal ferrite grain structure (Fig. 7), based on their observation by orientation imaging
microscopy, OIM, on the cleavage facet that a single cleavage facet can comprise of more
than one grain with up to 12 misorientation (due to mesotexture).
In general, mixed or bimodal ferrite grain structure is undesired from toughness point
of view of thicker steel plates (say, > 10 mm) for structural, linepipe and naval applications.
This is because coarse grains can not only act as the potential sites of cleavage crack
initiation, but also allow the crack to propagate over larger distances without any deviation
[35, 40]. Therefore, increase in size and frequency of coarse grains in bimodal grain structure
hampers the impact toughness. According to eqn. 3, larger the size of a grain, lower the stress
required for its cracking and hence, the grain becomes more potent crack initiator. Increase in
the frequency of coarse grains, on the other hand, increases the probability of finding those
grains within the active zone for fracture (in front of the notch-root), where principal stress
reaches a high value. This again increases the chance of cleavage crack initiation and
propagation by the catastrophic fracture of those grains [35, 40]. In order to understand the
effect of grain size distribution on toughness, Chakrabarti et al. [40] carried out blunt notch-
bend tests (at -160 C) to determine the cleavage fracture stress on three sets of samples of
low-C steels having different ferrite grain structures: (i) uniformly-fine grain structure, (ii)
uniformly-coarse grain structure and (iii) mixed or bimodal grain structure comprised of both
fine and coarse grains. The study not only showed the detrimental effect of bimodal grain
structure on toughness, but also indicated that the increase in grain size distribution (i.e.
bimodality) can increase the scatter in the measured toughness parameter [40].
Recent studies have showed that the development of bimodal grain structures (i.e.
mixed coarse- and fine-grain structure) can be a useful concept for improving the tensile
ductility of ultra-fine grains metals and alloys [41-48]. In general, ultra-fine grain structures
exhibit extraordinarily high strength. However, such microstructures also show limited room
temperature ductility, which can be attributed to the lack of strain-hardening during plastic
deformation [41-48]. Bimodal grain structures, on the other hand, shows excellent
combination of strength and ductility owing to the high strain-hardening capacity of the
coarse-grains and strengthening ability of the ultra-fine grains [41, 42]. However, the concept
of generating bimodal grain structure may be useful for thin-strip application (say, for auto-
body or deep-drawn components), where plane stress condition prevails and toughness is not
the prime concern. This concept may not be suitable for thick-plate applications (say,
linepipe, pressure vessel, naval applications) where high toughness is required. For the
prediction of DBTT in ferritic steels having bimodal grain structures, weighted average grain
size or even the largest grain size have been found to be useful, rather than numerical average
grain size, as the coarse grain sizes impose grater influence on toughness, than the fine grain
sizes [22, 23].

9
5. Combined effect of ferrite grain structure, crystallographic texture
and particles / inclusions on Charpy impact properties
In order to investigate the combined effect of ferrite grain structure, crystallographic
texture and particles / inclusions on the Charpy impact property, Ghosh et al. [51-53] studied
the impact transition behaviour of three different grades of low-C steels (0.06 wt.% C). The
first steel (S1) contained high amount of Ti (0.05 wt.%), which formed coarse TiN particles
up to  7 µm in size (cord length). The second steel (S2) contained high amount of S (0.03
wt.%), which formed large MnS inclusions more than 10 µm in length. The third steel (S3)
contained low S and Ti levels and were free from coarse TiN particles and MnS inclusions.
The as-cast steels were controlled-rolled following a schedule similar to Fig. 8, where same
amount of rolling deformation (65%) was applied, but varying the finish rolling
temperatures (FRT) in the range of 935°C to 650°C. Depending on the steel composition and
finish rolling temperature the rolled samples were coded as (S1/S2/S3)FRT(Temperature). For
example, S1FRT935 suggests that S1 steel is finish rolled at 935°C. The Ar3 and Ar1
temperatures of the investigated steels were 780°C and 610°C, respectively. Therefore,
finish rolling above and below 780°C means that the steel was rolled within austenite single-
phase region and austenite + ferrite inter-critical region, respectively. In order to study the
effect of normalizing temperature on impact properties, the inter-critically deformed samples
were normalized at different temperatures (940-1250C). S3 steel samples were cooled from
the normalizing temperatures by different ways: Furnace Cooling (FC), Air Cooling (AC) and
Water Quenching (WQ). The heat-treated samples were coded as (steel)HT(normalizing
temperature)(cooling). For example, S3HT940FC means S3 steel sample, normalized at
940C and furnace cooled from that temperature. The Charpy impact transition curves of as-
rolled and normalized samples of S1, S2 and S3 steels are presented in Fig. 9.

10
Fig. 9: Charpy impact transition curves of the rolled and heat-treated samples from (a) Steel
S1, (b), Steel S2 and (c) Steel S3.

S1FRT935 and S2FRT820 are finish rolled in single phase, austenite, whilst
S1FRT765, S2FRT730 and S2FRT650 are finish rolled in two phase, austenite-ferrite region.
But the total amount of rolling reduction is kept constant in all the cases. The USE of S1 and
S2 samples varied in the range of 60-100J and 35-60J, respectively, Fig. 9. DBTT of the steels
were in the range of +30 to -33 C for S1 and +9 to -84 C, for S2. Therefore, the presence of
large cuboidal TiN particle deteriorates the DBTT of steel S1 by initiating cleavage micro-
crack, Fig. 10a, whereas, large stringer shaped MnS inclusion deteriorates the USE of steel S2
by generating large voids around the inclusions, Fig. 10b. The propagation of cleavage cracks
depends on the ‘effective grain size’ and the strength of the matrix. Finish rolling within
austenite-ferrite two phase region (i.e. FRT below Ar3) leads to the formation of low-angle
boundaries, Fig 11(b, c), which are known to be ineffective in retarding the crack propagation.
An increase in the dislocation density and formation of low-angle boundaries in ferrite also
enhance the strength and reduce the strain-hardening ability of the steel. This can restrict the
growth of micro-voids, thereby reduce USE and promote cleavage crack propagation,
resulting in an increase in DBTT. Finish rolling at 820 C (just above Ar3) or normalizing
treatment at 940 C (just above Ac3) result in the formation of fine, recrystallized ferrite
grains with small ‘effective grain size’ (< 12 µm), Fig. 11(a,d), which resist the cleavage
crack propagation and therefore, improve the low-temperature toughness of steel, Fig. 9. It
has also been found that in spite of the presence of large TiN particles, refinement in
‘effective grain size’ of ferrite can significantly improve the impact toughness, provided the
matrix strength and mesotexture (i.e. the intensity of low-angle boundaries) are restricted to
low values.

Fig. 10:(a) TiN particles (circled) at the cleavage origin in steel S1 and (b) large MnS
inclusion (arrowed) inside the ductile voids in steel S2.

11
Fig. 11:Band contrast images of the investigated samples as obtained from the EBSD
analysis: (a) S2FRT820, (b) S2FRT730, (c) S2FRT650, and (d) S2HT940. High-angle
( 15 misorientation) and low-angle (< 15) boundaries are indicated by thick and
thin black limes, respectively.

The size and fraction of TiN and MnS inclusions were much smaller in S3 steel (due
to low Ti and S contents), than S1 and S2 steels. The Charpy transition curves of S3 samples
in Fig. 9c clearly showed that the absence of coarse inclusions improved the USE and reduced
the DBTT of that steel, compared to the other investigated steels. Furnace cooling from
940 °C provided USE as high as 250 J and reduced the DBTT to as low as -90 C. This can
be attributed to the formation of fine strain-free ferrite grains with small ‘effective grain size’
after low-temperature reheating and furnace cooling of S3 steel. Therefore, development of
uniformly fine, strain-free ferrite grain structure free from non-metallic inclusions is expected
to provide best combination of strength and impact toughness in low-carbon ferritic steel.
However, development of ultra-fine ferrite grain structure and complete removal of inclusions
from steel are industrially challenging and hence, impact toughness of commercial steel
grades cannot be improved beyond a certain limit. From this respect, it is necessary to
understand the effect of crystallographic texture on impact transition behaviour of ferritic
steel.

5.1. Concept of critical grain size for inclusion induced cleavage fracture

In view of the above experimental study, Ghosh et al. [51-53] proposed a model for
the prediction of critical grain size, below which the ferrite grain size has to be refined in
order to nullify the harmful effect of coarse TiN particles. This model is based on eqn. 3,
which is used for the determination of local cleavage fracture stress (f) considering either
grain size or particle size. If particle size controls the fracture and the propagation of the crack
across the particle-matrix (p-m) interface becomes the critical step for fracture then x is
particle size, p = pm and f = pm. On the other hand, if ferrite grain size controls the fracture
and the propagation of the crack across the matrix-matrix (m-m) grain boundary becomes the
critical step for fracture then x is grain size, p = mm and f = mm. Fig. 12 shows the σpm
values for different particle sizes and the variation of σmm as the function of average ferrite
grain size. The proposed model considers a competition between these two values at the local
scale. For a fixed particle size, from the intersection point of the σpm and σmm curve in Fig. 12

12
(indicated by arrows) a critical value of the grain size can be obtained below which, σmm > σpm
and hence, the local fracture stress (σf) will correspond to σmm, i.e. the grain size will control
the crack propagation. Above the critical grain size, σpm > σmm, hence σpm will be the local
fracture stress i.e. particle controlled crack propagation will prevail. In order to calculate σ mm,
the ‘effective grain size’ should be considered in eqn. 3 because low-angle boundaries are
ineffective in arresting a propagating micro-crack. In a similar way, for the calculation of σpm,
the 95th percentile particle size can be used in eqn. 3 as these particles are equal to or larger
than the size observed to initiate cleavage cracks [13], which has also been confirmed
experimentally in the present study. About 4 to 5 TiN particles per mm2 area have a size equal
or larger than the 95th percentile size and hence, approximately 400 to 500 such particles are
expected to be present on the entire cross-section (100 mm2 area) of a Charpy sample. Hence,
there is enough possibility of finding a large particle (95th percentile or higher) in front of the
notch-root (within say 500 m distance equating to an area of 5 mm2 and hence, approx 20-25
TiN particles), where the principal stress reaches the maximum value [54].

Fig. 12: Variation of local cleavage fracture stress with ferrite grain size and TiN particle size.

According to the present hypothesis a cleavage crack initiated from a coarse TiN
particle (situated within a ferrite grain), may be resisted by the ferrite grain boundary provided
the ferrite grain size is smaller than a critical value. On the other hand, the ferrite grain
boundary will not be effective in resisting the cleavage crack propagation, if the ferrite grain
size is larger than the critical size. Therefore, the present hypothesis provides a theoretical
basis to the experimental finding that coarse TiN particle situated within a coarse ferrite grain
promotes cleavage cracking [10-12, 26]. Considering the 95th percentile TiN size, the critical
grain size calculated from Fig. 12 (36 m) is smaller than the effective grain size of HT1150
sample, whilst it is larger than the effective grain sizes of FRT765, FRT935 and HT940
samples. Hence, cleavage fracture in HT1150 sample is expected to be particle controlled,
however for the other samples it is expected to be ferrite grain controlled. This is contrary to
the experimental observations, where TiN initiated failure was observed for all samples.
However, it is possible that TiN particles may be observed to initiate failure, but if the critical
grain size is similar to the actual grain size the effect on toughness may be small. It is also
known that mm is not independent of temperature and has been reported to increase
significantly (up to as high as up 500 J/m2) with an increase in test temperature [3, 11, 26, 27],
which may increase the Dcrit value.

13
6. Effect of microstructure and crystallographic texture on
delamination behaviour in impact tested samples.
During the Charpy impact testing of low-C ferritic steel at an intermediate test
temperature range (0°C to -80°C), a special type of crack is often found, which propagates
through the transverse plane with respect to the main fracture plane [55–58], Fig. 13. This
type of crack on the fracture surface is commonly known as fissure, or splitting or
delamination [55–58]. It has been reported that the formation of fissures can be advantageous
as it reduces the ductile to brittle transition temperature [57,59]. However, in general, fissures
are considered to be detrimental as these reduce the impact energy absorption capacity of
Charpy specimens machined along T-L (Transverse-Longitudinal) orientation with respect to
the rolled plates [55-60].

Fig. 13: Presence of fissures (arrowed) on the fracture surface of Charpy impact tested
specimen of inter-critically rolled low-C steel.

Finding the root cause behind the fissure formation during the Charpy impact testing
has been an interesting and debatable issue since many years [55,60–65]. Mintz et al. [66]
mentioned that the early stage of fissure formation is always intergranular, primarily initiated
from inclusion/carbide particles situated at the grain boundaries. Presence of elements like S
and P in steel at high level can develop coarse sulfide or phosphide inclusions (Such as MnS
and FeP) that can weaken the grain boundaries and promote intergranular cracking [55,70].
Microstructural banding is also reported to be responsible for fissure formation as the strain
incompatibility between the adjacent hard and soft phases can generate fissures along the
interface [55,56,67]. Fissures can also be developed by the decohesion at grain boundaries of
the deformed ferrite grains having elongated structure along the rolling direction [57,66] and
having high dislocation density [67, 68]. Interaction of dislocations with grain boundaries
weakens the grain boundary, which acts as the favorable path for fissure crack propagation
[69].
Compared to the studies on microstructural constituents, fewer studies relating
crystallographic texture with fissure formation are available. Bourell et al. [58] mentioned that
the presence of cube texture (ND ║<001>) can be the primary reason behind the formation of
fissure. They have assumed that the weakness of the transverse ‘fissure plane’ arises from the
preferential alignment of {001} planes of the crystals parallel to the rolling plane. Inoue et al.
[71] attributed the fissure formation to the presence of α-fiber texture (RD║<110>) developed
during the warm caliber rolling of ferritic steel. In contradiction several studies
[55,56,68,69,71] reported that crystallographic texture is not responsible for fissure formation.
Mostly fissures have been found to form during impact testing of steels which are finish rolled
near the lower critical temperature (Ar3) range, having elongated and heavily dislocated ferrite
grains and preferential cube texture [63,72]. This makes it difficult to identify the primary
cause behind fissure formation.

14
 In order to understand the role of crystallographic texture on splitting or delamination
behavior (or fissure formation) in Charpy impact tested samples, Ghosh et al. [53] performed
combined SEM and EBSD analyses on the samples of low-C steel (free from coarse particles /
inclusions), finish rolled at different temperatures (820 °C - 650 °C). Through thickness
texture band composed of cube (ND ║ <001>) and gamma (ND ║ <111>) fibre orientations
has been found to develop during the inter-critical rolling. Fissure crack is found to propagate
in intergranular fashion between those two textures bands on the main fracture plane owing to
the strain incompatibility, Fig 14a. In order to understand the crack deflection at the grain
boundary, different combinations of neighbouring crystal orientations have been considered
and possible tilt and twist angles at the grain boundary have been evaluated. ‘Effective grain
size’ was estimated from the angle between {001} planes of neighbouring crystals, which is
used to determine the cleavage fracture stress on different planes of the sample. The severity
of splitting was found to be directly related to the difference in cleavage fracture stresses
between the ‘main fracture plane and ‘fissure plane’. Clustering of ferrite grains having cube
texture, Fig. 14(b-d), promoted the fissure crack propagation in transgranular fashion across
the transverse ‘fissure plane’, by increasing the ‘effective grain size’ and decreasing the
cleavage fracture stress on that plane. Finally, a phenomenological model has been proposed
to represent the mechanism of splitting.

Fig. 14: Inverse pole figure (IPF) map of (a) the region surrounding a hairline fissure crack on
RD-ND plane (parallel to main fracture plane) (b) Clustering of cube oriented grains
on RD-TD plane, indicated by dotted circle, (c) Colour legend of IPF, (d) Schematic
showing the clustering of cube oriented grain on fissure plane.

7. Conclusions
The paper presents an extensive review on the effect of microstructural parameters and
crystallographic texture on Charpy impact toughness and ductile-brittle transition temperature
(DBTT) of low-carbon ferritic steels. Specified level of Charpy impact toughness is an
essential requirement for the steel plates / steel components used in structural, linepipe,
automotive, naval and defence applications. The major conclusions derived from this review
work are as follows:

15
  In case of low-carbon, microalloyed -TMCR steel existing regression
equations suggested by Gladman-Pickering [1, 29] and Mintz et al. [30, 31]
can satisfactorily predict the impact transition temperature from Charpy
impact test. The equation takes into account the effect of carbide size, grain
size, pearlite content and steel composition on impact transition temperatures.
 Beside the above mentioned factors mesotexture can influence impact
tranbsition temperatures. Severe mesotexture increases effective grain size and
raises the ITT, showing poor toughness.
 The ferrite grain size distribution can influence the scatter in Charpy impact
toughness results. Severe grain size bimodality (i.e. mixed coarse and fine
grain structure) is detrimental for toughness as the coarse grains help in
initiating and propagating the cleavage crack, Grain size bimodality can also
increase the scatter in impact toughness.
 Detrimental effect of TiN particles and MnS inclusions on impact toughness
and ductile brittle transition temperature (DBTT) can be encountered by
refining the ‘effective grain size’ and restricting the intensity of low-angle
boundaries.
 Finish rolling at 820 C (just above Ar3) or normalizing at 940 C (just
above Ac3) develop fine, recrystallized ferrite grains, which improve the low-
temperature toughness of steel.
 Strain incompatibility arises due to alternate arrangement of different texture
components (cube and gamma-fibre texture) can result in fissure formation
during Charpy impact testing.

Acknowledgements:
Corresponding author (DC) acknowledges the financial support from Department of Science
and Technology (DST) and the experimental facilities developed under SGIRG funding from
IIT Kharagpur. Sincere thanks to RDCIS, SAIL, Ranchi, and DMRL, Hyderabad for the
provision of research material. The help received from Dr. B. K. Jha and Dr. Ramen Dutta
from RDCIS, SAIL, deserves a special mention.

References:
[1] T. Gladman, The Physical Metallurgy of Microalloyed Steels, Book 615, The Institute
of Materials, London, (1997).
[2] G. Tither, The development and applications of niobium-containing HSLA steels, Proc.
2nd Int. Conf. On ‘HSLA steels: processing, properties and applications’, (ed. G. Tither
et al.), Baijing, China, TMS, Warrendale, (1992), pp. 61-80.
[3] J.F. Knott, Fundamentals of Fracture Mechanics, Butterworths, London, (1973).
[4] R.W. Hertzberg, Deformation and Fracture Mechanics of Engineering Materials, J.
Wiley & Sons, New York, (1995).
[5] G.E. Dieter, Mechanical Metallurgy, McGraw-Hill, London, (1988).
[6] ASTM E23-0.5, Standard Test Methods for Notched Bar Impact Testing of Metallic
Materials, Annual Book of ASTM Standards, vol. 03.01, ASTM International,
Pennsylvania, USA, (2005).
[7] D.K. Biswas, M. Venkatraman, C. S. Narendranath, U. K. Chatterjee, Metall. Trans. A,
23, (1992), 1479–1492.
[8] F. A. McClintock, Int. J. Fract. Mech., 40, (1968), 101–130.
[9] F. A. McClintock, Ductility, ASM, Metals Park, OH, 3, (1968), 255–277.

16
[10] D.P. Fairchild, D.G. Howden, W.A.T. Clark, Metall. Mater. Trans. A, 31, (2000), 641–
652.
[11] M.A. Linaza, J.L. Romero, J.M. Rodríguez-Ibabe, J.J. Urcola, Scr. Metall. Mater., 32,
(1995), 395–400.
[12] A. Ray, M. Tech. Thesis, Cleavage Initiation in Steel: Competition between Large
Grains and Large Particles, Indian Institute of Technology Kharagpur, (2011).
[13] D.A. Curry, J.F. Knott, Metal Science, 12, (1978), 511-514.
[14] C. Yan, J.H. Chen, J. Sun, Z. Wang, Metallurgical Transactions, 24A, (1993), 1381–
1389.
[15] A. A. Griffith, The Phenomena of Rupture and Flow in Solids, Philosophical
Transactions of the Royal Society of London, Series A, 221, (1921), 163-198.
[16] E. Orowan, Notch brittleness and strength of metals, Transactions of Inst. of Engineers
and Shipbuilders in Scotland, 80, (1946), 165-196.
[17] C. Zener, The micromechanism of fracture, Transactions ASM, 40 A, (1948), 3-31.
[18] A.N. Stroh, The formation of cracks as a result of plastic flow, Proceedings of Royal
Society of London, 223A, (1954), 404-414.
[19] E. Smith E, The nucleation and growth of cleavage microcracks in mild steel, Proc.
Conf. Physical Basis of Yield and Fracture, Inst. of Physics and Physical Society,
Oxford, (1966), 36-45.
[20] A.H. Cottrell, Theory of brittle fracture in steel and similar metals, Transactions AIME,
212, (1958), 192-203.
[21] D. Hull, Twining and fracture of single crystals of 3 % silicon iron, Acta Metallurgica,
8, (1960), 11-18.
[22] T.C. Lindley, G. Oates, C.E. Richards, Acta Metallurgica, 18, (1970), 1127-1136.
[23] Z. Liu, Iron Steel Sinica, 17, (1982), 12-20.
[24] P. Bowen, S.G. Druce, J.F. Knott, Acta Metallurgica, 34, (1986), 1121-1131.
[25] N. J. Petch, Acta Metallurgica, 34, (1986), 1387-1393.
[26] A. Echeverria, J.M. Rodriguez-Ibabe, Materials Science and Engineering A, 346,
(2003), 149-158.
[27] J.I. San Martin, J.M. Rodriguez-Ibabe, Scripta Materialia, 40, (1999), 459-464.
[28] J.H. Chen, G.Z. Wang, C. Yan, H. Ma, L. Zhu, International Journal of Fracture, 83,
(1997), 139-157.
[29] F.B. Pickering, T. Gladman T, Investigation into some factors which control strength of
carbon steels, Iron and Steel Institute -- Special Report No. 81, (1963), 10-25.
[30] B. Mintz, E. Maina, W.B. Morrison, Mater. Sci. Technol., 23, (2007), 347.
[31] B. Mintz, W.B. Morrison, Mater. Sci. Technol., 4, (1988), 719.
[32] D. Bhattacharjee, C.L. Davis, J.F. Knott, Predictebility of Charpy impact toughness in
thermomechanically control rolled (TMCR) microalloyed steels, Ironmaking and
Steelmaking, 30, (2003), 249-255.
[33] D. Bhattacharjee, J.F. Knott, C.L. Davis, Metallurgical and Materials Transactions A,
35A, (2004), 121-130.
[34] S.J. Wu, C.L. Davis, Materials Science and Engineering A, 387-389, (2004), 456-460.
[35] Debalay Chakrabarti, PhD thesis, the University of Birmingham, UK, (2007).
[36] B.K. Show, R. Veerababu, R. Balamuralikrishnan, G. Malakondaiah, Mater. Sci. Eng.
A, 527, (2010), 1595.
[37] M.T. Shehata, J.D. Boyd, Quantitative correlations between toughness and
microstructure for commercial line pipe steel, Proc. Conf. On ‘Advances in the
Physical Metallurgy and Applications of Steels’, Liverpool, Metals Soc (Book 284),
London, (1982), 229-236.
[38] E. Bouyne, H.M. Flower, T.C. Lindley, A. Pineau, Scripta Mater., 39, (1998), 295-300.
[39] M.-C. Kim, Y.J. Oh, J.H. Hong, Scripta Mater., 43, (2000), 205-11.

17
 [40] D. Chakrabarti, M. Strangwood, C. L. Davis, Metallurgical and Materials Transactions
A, 40A, (2009), 780-795.
[41] S. Patra, Sk. Md. Hasan, N. Narasaiah, D. Chakrabarti, Materials Sci. and Engg. A,
538, (2012), 145– 155.
[42] R. Song, D. Ponge, D. Raabe, J.G. Speer, and D.K. Matlock, Mater. Sci. Eng. A, 441,
(2006), 1–17.
[43] J.-E. Jin, Y.-S. Jung, and Y.-K. Lee, Mater. Sci. Eng. A, 449–451, (2007), 786–789.
[44] M.C. Zhao, F. Yin, T. Hanamura, K. Nagai, A. Atrens, Scripta Mater., 57, (2007), 857–
860.
[45] H. Azizi-Alizamini, M. Militzer, W.J. Poole, Scripta Mater., 57, (2007), 1065–68.
[46] A. Ohmori, S. Torizuka, K. Nagai, ISIJ Int., 44, (2004), 1063–71.
[47] R. Song, D. Ponge, D. Raabe, Scripta Mater., 52, (2005), 1075–80.
[48] A. Karmakar, A. Karani, S. Patra, D. Chakrabarti, Metallurgical and Materials
Transactions A, 44A, (2013), 2041-2050.
[49] H. Qiu, R. Ito, K. Hiraoka, Mat. Sci and Engg. A, 435-436, (2006), 648-652.
[50] T. Hanamura, F. Yin, K. Nagai, ISIJ Int., 44, (2004), 610-617.
[51] A. Ghosh, A. Ray, D. Chakrabarti, C.L. Davis, Materials Science and Engg. A, 561,
(2013), 126-135.
[52] A. Ghosh, S. Sahoo, M. Ghosh, R. N. Ghosh, D. Chakrabarti, Materials Science and
Engineering A, 613, (2014), 37-47.
[53] A. Ghosh, Ph.D. thesis, Effect of Microstructure and Crystallographic Texture on
Impact Toughness in Low Carbon Ferritic Steel, I.I.T. Kharagpur, December, (2015).
[54] G. T. Hahn, Metall. Mater. Trans. A, 15, (1984), 947–959.
[55] E.A. Almond, Metall. Trans., 1, (1970), 2038.
[56] H. Herø, J. Evensen, J.D. Embury, Canadian Metall. Quarterly, 14, (1975), 117.
[57] B.L. Bramfitt, A.R. Marder, Metall. Trans. A, 8, (1977), 1263.
[58] D.L. Bourell, Metall. Trans. A, 14, (1983), 2487.
[59] B. Mintz, E.M. Maina, W.B. Morrison, Mater. Sci. Technol., 24, (2008), 177.
[60] M. Joo, D. Suh, J. Bae, H.K.D.H. Bhadeshia, Mater. Sci. and Engg. A, 546, (2012),
314.
[61] Y. Kimura, T. Inoue, F. Yin, K. Tsuzaki, Science, 320, (2008), 1057.
[62] R. Schofield, G. Rowntree, N.V. Sarma, R.T. Weiner, Met. Technol., 1, (1974), 325.
[63] B. Mintz, W.B. Morrison, Mater. Sci. Technol., 23, (2007), 1346.
[64] R. Song, D. Ponge, D. Raabe, R. Kaspar, Acta Mater., 53, (2005), 845.
[65] M. Yang, Y.J. Chao, X. Li, J. Tan, J. Mater. Sci. Eng. A, 497, (2008), 451.
[66] B. Mintz, E. Maina, W.B. Morrison, Mater. Sci. Technol., 23, (2007), 347.
[67] M. Yang, Y.J. Chao, X. Li, D. Immel, J. Tan, Mater. Sci. Eng. A, 497, (2008), 462.
[68] R. Punch, M. Strangwood, C. Davis, Metall. Mater. Trans. A, 43, (2012), 4622.
[69] B. Mintz, W.B. Morrison, Mater. Sci. Technol., 4, (1988), 719.
[70] T. Inoue, F. Yin, Y. Kimura, K. Tsuzaki, S. Ochiai, Metall. Mater. Trans. A, 41,
(2009), 341.
[71] S.Y. Shin, S. Hong, J-H. Bae, K. Kim, S. Lee, Metall. Mater. Trans. A, 40, (2009),
2333.
[72] R.K. Ray, J.J. Jonas, Int. Mater. Rev., 35, (1990), 1.

18

View publication stats