Research Article

Materials Science and Technology

1–11
Microstructural analysis of cold-drawn © The Author(s) 2024
Article reuse guidelines:
medium-carbon low-alloy martensite: sagepub.com/journals-permissions
DOI: 10.1177/02670836241257073
effect of deformation and tempering journals.sagepub.com/home/mst

temperature

María C. Marinelli1,2, Fernando D. Carazo2,3,

Virginia H. Aranda3 and Raúl Bolmaro1,2

Abstract
Microstructure of cold-drawn (≈ 0.9 effective deformation) medium-carbon steel containing Cr and Mo, was investigated
in samples tempered at 923 K (tempering at high temperature [HTT]) and 723 K (tempering at intermediate temperature
[ITT]) using transmission electron microscopy, X-ray diffraction and microhardness. The study revealed distinct features
in lath martensite. HTT samples showed reduced dislocation density and larger precipitates compared to ITT samples.
After cold drawing, ITT samples exhibited lamellar dislocation cells and cell blocks, while HTT specimens primarily
showed cell block structures. Analysis of precipitates revealed changes in orientation after drawing, as well as alterations
in both inter- and intragranular precipitate density. ITT samples also displayed higher post-tempering hardness. This ana-
lysis provides quantitative insights into the microstructural evolution and mechanical behaviour after tempering and cold-
drawing, which can help optimise processes for achieving ultrafine-grained structures.

Keywords
ultrafine-grain steel, lath martensite, sub-structures, intra and inter-lath carbides, cold-drawing

Submitted: 27 July 2023; received: 4 January 2024; accepted: 29 April 2024

Introduction uniform carbide precipitation.14–16 The development of an

ultrafine microstructure is attributed to the fine martensite
While many hardening mechanisms often trade strength for
starting microstructure and the effect of plastic deformation
reduced toughness and ductility, grain refinement stands out when the material undergoes shaping processes like cold-
as a method widely employed for enhancing both attri- rolling, causing elevated dislocation density. The small
butes.1 Ultrafine-grained (UFG) steels exhibit exceptional
and closely spaced martensitic structures are more respon-
mechanical properties, boasting elevated strength, com- sive to the deformation forces, leading to greater changes
mendable fracture toughness and a low ductile-to-brittle in their arrangement and size, promoting enhanced frag-
transition temperature. Currently, two predominant techni-
mentation of the existing microstructure into smaller
ques for UFG steel production exist: severe plastic deform- grains and contributing to the formation of an UFG micro-
ation (SPD)2–7 and advanced thermo-mechanical structure.9,10,17 Another significant factor influencing the
processing (ATP).8–12 SPD involves subjecting materials
to significant plastic strains, either at room or elevated tem-
peratures, primarily within the warm deformation regime. 1
Instituto de Física Rosario, Universidad Nacional de Rosario, Rosario,
Conversely, ATP employs alternative strategies to generate Argentina
UFG steels, utilising relatively low accumulated deform- 2
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET),
ation, making it appropriate for industrial applications.13 Argentina
3
In recent years, several thermo-mechanical processes have Instituto de Mecánica Aplicada IMA, Facultad de Ingeniería, Universidad
Nacional de San Juan, San Juan, Argentina
been developed for the production of UFG steels.13 In par-
ticular, many researchers have focused on processes charac- Corresponding author:
María C. Marinelli, Instituto de Física Rosario, Universidad Nacional de
terised by cold deformation of martensitic structures, which Rosario, Bv. 27 de febrero 210 bis, 2000 Rosario, Argentina; Consejo
involves cold-rolling of a martensitic microstructure fol- Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina.
lowed by annealing to achieve ultrafine ferrite grains and Email: marinelli@ifir-conicet.gov.ar
2 Materials Science and Technology 0(0)

degree of microstructure refinement is the carbon content.

Studies indicate that with an increase in carbon content,
there is a decrease in the width of martensitic blocks and
packages18 and, similar to grain boundaries, lath boundaries
act effectively to strength enhancement.11 Nevertheless,
most studies on utilising martensite as a primary microcon-
stituent to produce UFG steel have mainly concentrated on
plain low-carbon steel sheets (up to ≈ 0.20 wt.%
C).9,10,16,17,19–22 Notably, comprehensive studies on the
grain refinement of medium-carbon steels with an initial
martensitic microstructure are limited in the current litera-
ture,11,23,24 warranting detailed exploration in future
investigations.
In a recent work, Aranda et al.25 adapted the method pro-
posed by Tsuji et al. in low-carbon steel sheets10,14 to obtain
an UFG structure in a cylindrical bar of medium-carbon
(0.40 wt.%C) low-alloy steel. The process involved auste- Figure 1. Schematic diagram of the decoupled thermo-
nisation followed by quenching in oil to obtain a full lath mechanical process applied to the bars. Austenitising tempera-
martensite structure and then tempering at 723 K for 1 h. ture: 1148 K, tempering temperature: 723 K25 and 923 K.
Subsequently, the bars underwent cold drawing up to an
effective deformation of ≈0.9. Building upon this prior equivalent strain of ≈ 0.9) in a four-stage drawing
work, this article aims to further explore the morphological sequence. The area reductions were achieved successively,
features of lath martensite obtained through tempering at reducing the diameter from 12.6 to 8.7 mm.
high and intermediate temperatures (HTT and ITT, respect-
ively). Additionally, it seeks to analyse the microstructural
changes induced by cold drawing. Therefore, the objective Specimen preparation for optical microscopy, XRD
of this article is to analyse deformation microstructures and and TEM analysis
structural parameters using transmission electron micros-
copy (TEM) and X-ray diffraction (XRD). The study also The microstructure analysis of both tempered undeformed
encompasses an evaluation of mechanical properties and cold-drawn specimens, which underwent a 51.8% accu-
through microhardness assessments to track the work- mulated area reduction, was carried out using optical
hardening effect during the cold drawing process. microscopy (OM), TEM and XRD. For OM and XRD ana-
lysis, samples were ground to 2000 grit paper and then
etched with 2% Nital. TEM was used to observe internal
Material and experimental procedure dislocation structures and precipitates, and thin foils were
prepared from slices, cut in the longitudinal direction of
Material the bar, coinciding with the drawing direction (DD).
The material under investigation in this article is a ferritic- Discs of 3 mm in diameter were cut using a hole punch
perlitic AISI 4140 steel with the chemical composition of and subsequently thinned and electro-polished in a
Fe-0.40C-0.78Mn-0.25Si-0.85Cr-0.16Mo (wt.%). This twin-jet polishing unit (Struers TenuPol-5) using a 10%
steel has a remarkable combination of strength, toughness perchloric acid and 90% ethanol mixture at −15°C and
and hardenability, making it a popular choice for use in 20.5 V. The dislocation density was calculated by XRD
structural and mechanical components. The material was using a Philips X pertPro-MPD diffractometer with
provided in the form of a cylindrical bar with a diameter Cu-Kα radiation at 40 kV and 15 mA, Panalytical Ltd,
of 12.6 mm and a length of 400 mm. Malvern, UK. Diffractograms were measured with a scan
step size of 0.02° and scan rate of 0.01°/s. Data were ana-
lysed by a Rietveld model as implemented in Maud soft-
Thermo-mechanical processes ware.26 Peak broadening is calculated by Langford
Figure 1 illustrates the thermo-mechanical process used in model27 and domain sizes and dislocation densities by
this investigation. The heat treatment involved austenisa- W-S equations.28
tion at 1148 K for 30 min, followed by quenching in oil
to obtain a full lath martensite structure. The resulting
material was then tempered at 923 K (HTT) and 723 K Microhardness test
(ITT) for 60 min. The tempered lath martensitic bars were To assess the work hardening at each stage of cold
then cold-drawn using an adapted laboratory-scale drawing, microhardness tests were conducted on
drawing equipment installed in a tensile test machine samples tempered at HTT and ITT, in accordance with
(Instron 4206). To reduce friction between the bar and the ASTM E384-17. The microhardness tests were performed
die, a calcium-type powder lubricant was used. The speci- using a SHIMADZU MHV-2 microdurometer, applying a
mens were drawn up to a 51.8% area reduction (an load of 1.96N for 15 s.
Marinelli et al. 3

Results and discussion tempering is employed to enhance the material’s toughness.

Elevating the tempering temperature induces various micro-
Undeformed state microstructure structural changes, with the most significant associated with
The typical martensitic microstructure comprises a multi- the precipitation and growth of carbides, alleviating the
scale structure formed during quenching, where prior aus- supersaturation of carbon in the BCT iron. Thus, carbon dif-
tenite grains are divided into several packets composed of fuses from interstitial sites in the BCT martensite, and the
parallel blocks with the same habit plane. Each block is crystal structure reverts to BCC ferrite with dispersions of
further divided into sub-micrometer martensite laths with carbide particles. With increasing tempering temperature,
similar crystallographic orientation.11,29 Figure 2(a) shows stable and coarser mixtures of ferrite and carbides are
the optical micrograph of the as-quenched specimen, in formed from as-quenched martensitic microstructures.30
which typical morphology of lath martensite is observed. Figure 2(b) and (c) show the microstructure of tempered
During quenching, the original austenite structure under- martensite at ITT and HTT, respectively, and it can be
goes transformation through diffusion-less and complex observed that the lath martensite structure is preserved
shear mechanisms into small laths of tetragonally distorted even after tempering at relatively high temperatures, such
martensite (BCT). As BCT martensite is inherently brittle, as 923 K. Furthermore, TEM images in Figures 3 and 4

Figure 2. Optical micrographs of undeformed specimens (a) as-quenched, (b) tempered at 723 K and (c) tempered at 923 K, etched
with 2% nital.

Figure 3. TEM BF images of the undeformed sample tempered at ITT: (a) blocks of lath martensite with high dislocation density, (b)
intralath and interlath carbides, (c) cell structure within laths highlighted by arrows.
4 Materials Science and Technology 0(0)

Figure 4. TEM BF images of the undeformed sample tempered at HTT: (a) lath martensite along (110) slip plane and cementite
particles, (b) detail of the area enclosed by dashed lines in (a) showing ‘cell-like’ structure, (c) intralath and interlath carbides, (d) detail
of the area enclosed in (c).

are presented to compare the substructure in both tempering Figure 4 illustrates the microstructure of an undeformed
processes. specimen tempered at HTT (923 K). As shown in
Figure 3 presents bright-field TEM images of Figure 4(a), the microstructure consists of lath martensite,
undeformed samples that were tempered at ITT (723 K). with a width of approximately 0.3–0.4 μm, parallel to
As observed in previous work,25 the martensite blocks, {110} planes. This is consistent with previous findings
delimited by dashed lines in Figure 3(a), are subdivided reported by Beladi et al.33 It suggests that most lath bound-
into laths approximately 0.3 μm wide with high disloca- aries within a single martensite block are nearly parallel to a
tion density and carbides inside them. Figure 3(b) specific {110} slip plane. In addition, double spots on the
shows an array of fine carbides within laths (enclosed SAED pattern in Figure 4(a), indicate the presence of inter-
by dotted lines) and film-like carbides at lath/block lath carbides on {110} planes.34 A closer examination of
boundaries (marked by arrows), referred as intralath and Figure 4(a) reveals, enclosed by dotted lines, dislocations
interlath carbides respectively. According to Krauss25,31 arranged in a ‘cell-like’ structure with incomplete cell
and Bhadeshia,32 intralath carbides may originate from walls.35 Figure 4(b) shows a magnified view of the region
the transformation of transition carbides to cementite enclosed by the dotted line in Figure 4(a), where entangled
after tempering, while film-like carbides result from the dislocations are observed along with precipitates. These
decomposition of retained austenite located in lath bound- precipitates hinder the movement of dislocations, causing
aries. Furthermore, dislocation cell structures were them to pile up and form a dislocation cell wall (indicated
observed within some lath martensite as shown in by the red arrow in Figure 4(b)). The partial formation of
Figure 3(c) (indicated by arrows). This suggests that these low-angle boundary substructures suggests that a
during tempering, dislocations are rearranged through a recovery mechanism becomes active during tempering,
process of recovery.19 leading to the rearrangement of dislocations into lower
Marinelli et al. 5

energy configurations.36,37 Figure 4(c) shows a lower oriented nearly parallel to the DD, as shown in Figures
density of dislocations than ITT within the laths and preci- 5(a) and 6(a).
pitates of various sizes and shapes, as shown in detail in In particular, the ITT samples exhibit inhomogeneous
Figure 4(d). Notably, the precipitated cementite particles deformation structures similar to those obtained by cold
are larger at HTT than at ITT. This phenomenon is well- rolling,14,17 such as the lamellar dislocation cell (LDC)
known in the tempering of carbon steels to HTT.32,38 structure, which is primarily composed of lamellar bound-
Recently, it was reported39 that the tempering kinetics are aries elongated along DD, as shown in Figure 5(a). A
controlled by carbon diffusion. As the tempering tempera- rings SAED pattern was observed indicating large misor-
ture rises, the lath martensitic morphology tends to ientation. This point will be discussed later for HTT. In add-
vanish, giving way to a microstructure characterised by ition, the dashed curves in Figure 5(b) indicate the
spheroidal carbides. Moreover, as the tempering time irregularly bent lath (IBL) structure, where a cluster of
increases, the carbide size also increases and hardness laths is bent and oriented in DD. It is noteworthy that
decreases. Therefore, the diffusion of carbon atoms is an certain regions exhibit vanished lath boundaries, forming
important aspect of the tempering treatment. dislocation cell structures, as shown in Figure 5(c).
Figure 5(d) shows a region with two types of structures:
LDCs and dislocation walls forming cells, delimited by a
Deformed state microstructure boundary highlighted by a dotted line. This result has also
Figures 5 and 6 illustrate the microstructure of the ITT been reported by other authors in cold-rolled martens-
and HTT specimens, respectively, after deformation to ite.16,19 They observed that this effect is most pronounced
an average equivalent strain of ≈ 0.9. It is worth noting in areas near high-angle boundaries such as block
that the deformed samples exhibit a lamellar structure boundaries.

Figure 5. TEM micrographs of cold-drawn ITT specimens: (a) LDCs structure in DD, (b) IBL structure, (c) cell blocks, (d) mixed
structure: LDCs and cell blocks.
6 Materials Science and Technology 0(0)

Figure 6. TEM micrograph of cold-drawn HTT specimens: (a) lath martensite in DD with SAED 1 and cell-block structure with SAED
2, (b) high dislocation and precipitate density, (c) IBL structure and twisting into DD of cementite particles, (d) IBLs indicated by arrows
and cell block structures highlighted by dashed lines.

On the other hand, the deformed microstructure of the Figure 6(c). This observation was also reported by Zhang
HTT specimens reveals wider laths (Figure 6(a)) compared et al.40 in cold-drawn pearlitic steel wire.
to the ITT samples and a high density of dislocations and It is important to note that the laths are subdivided, and
precipitates (Figure 6(b)). Figure 6(a) displays a mixed their boundaries partially vanished, as shown in zones
structure of lath martensite and precipitates, both oriented enclosed by dashed lines in Figure 6(d). These results
in the DD. Indexed SAED patterns of the areas labelled suggest that low-angle laths in HTT specimens are reorga-
as SAED1 and SAED2 are also displayed in Figure 6(a). nised into cell structures. Morito et al.19 reported that this
The SAED1 pattern corresponds to a martensite lath phenomenon is associated with dislocation gliding across
oriented in DD while the ring-like pattern observed in adjacent laths, resulting in a cell block structure formed
SAED 2 was obtained from an area of cell blocks with ran- by ordinary cell boundaries and cell block boundaries (indi-
domly distributed precipitates. Lan et al.22 reported this cated by dashed lines in Figure 6(d)). The former has an
phenomenon in cold-rolled low-carbon steel and attributed almost random orientation, and the latter has an orientation
it to cell blocks with large misorientation. However, Zhang nearly planar boundary.37 These extended boundaries
et al.40 reported that the break-up of cementite plates/parti- exhibit diverse characteristics, but a common feature is
cles occurs during drawing in cold-drawn wires. Therefore, that they delineate regions that are further subdivided by
the SAED 2 patterns could be due to both large misorienta- typical cell boundaries.
tion of cell blocks and randomly distributed precipitates.
Furthermore, as observed in the ITT sample, the IBL
structure is present in the high-tempered condition, as indi- Comparison of ITT and HTT specimens
cated by arrows in Figure 6(c)–(d). Specifically, cementite Analysis of the distribution of cementite particles. Table 1 pro-
particles rotated towards the DD are observed in vides a quantitative analysis of cementite particle size and
Marinelli et al. 7

density, before and after drawing, for both ITT and HTT subdivide the structure on a progressively finer scale as
specimens. Intragranular and intergranular cementite parti- deformation increases. In this study, XRD was employed
cles were identified by TEM in both specimens, with the to measure the dislocation density in both initially tempered
latter being larger due to their preferential growth along and deformed samples. The results are presented in Table 2
the boundary plane and acquiring an elongated shape.41 and Figure 7(a), revealing that the dislocation density in the
In this study, ImageJ software42 has been used to quantify undeformed samples decreases as the tempering tempera-
each type of particles. It is noteworthy that coarsening ture increases. The difference in dislocation density
kinetics differ for each type of particle and are influenced between the ITT and HTT samples is 7 × 1014 m−2, which
by different coarsening mechanisms. Nam et al.43 reported corresponds to a relative reduction of 99.8% compared to
that intergranular cementite particle coarsening is signifi- the ITT samples. According to Bhadeshia,32 initial temper-
cantly influenced by boundary diffusion and diffusion ing involves significant dislocation rearrangement resulting
along dislocations, while intragranular particle coarsening from the recovery process, evidenced by the observed cell
is significantly influenced by boundary diffusion and structure in Figure 3(d). In contrast, at higher tempering
matrix diffusion. Table 1 shows that both, intergranular temperatures, the recovery process is replaced by recrystal-
and intragranular particles grow with increasing tempering lisation. While recrystallisation is more easily in low-
temperature. carbon steels, higher carbon contents enhance the Fe3C par-
After the deformation process, a redistribution of car- ticle density, effectively impeding lath boundary move-
bides was observed in both specimens, as evidenced by ment. Consequently, recrystallisation slows, leading to
the presence of concentric rings in the SAED figures gradual lath growth and coarsening of Fe3C particles,
(Figure 5(a) and Figure 6(a)). Table 1 reveals that in showed in Figure 4(d).
both specimens, after tempering, the particle density is On the other hand, the dislocation density increases after
higher at lath boundaries. However, after drawing, some cold drawing in both ITT and HTT samples. However, it is
lath boundaries disappear, causing the cementite particles worth noting that the increase in dislocation density in ITT
that were previously located at boundaries, to become samples is around 16%, while the increase in HTT samples
randomly oriented. This results in a reduction in intergra- is much higher, around 350 times, reaching approximately
nular particle density and an increase in intragranular par- 1014 m−2, which is of the same order of magnitude as the
ticle density for both specimens. It is important to note dislocation density of ITT samples. This phenomenon is
that the total particle density remains practically closely related to the formation and annihilation of disloca-
unchanged after drawing, indicating no fragmentation of tions during plastic deformation, resulting in a final disloca-
cementite particles. tion structure of LDCs and cell blocks in ITT and HTT
samples, respectively, as shown in Figures 5 and 6.
Analysis of the structural parameters. The formation of sub- Furthermore, the samples underwent a four-stage cold
structures during cold-drawn of lath martensite is closely drawing sequence, and the initial structure of lath martens-
linked to the formation and multiplication of dislocations. ite changed at each stage. During low plastic deformation,
These dislocations interact and are stored in the form of dislocation walls are induced across the laths, allowing
low-angle and high-angle dislocation boundaries, which the transition from lath martensite to cell block structure.

Table 1. Size and density of cementite particles using image-J software.42

Cementite size dθ (nm) Density of particles (1/m2)

Tempering temperature
(K) Samples Intergranular Intragranular Intergranular Intragranular Total density

723 (ITT) Undeformed 190 ± 60 47 ± 14 4.4 × 1013 3.3 × 1013 7.7 × 1013
Deformed 124 ± 50 55 ± 12 1.5 × 1013 5 × 1013 6.3 × 1013
923 (HTT) Undeformed 260 ± 60 99 ± 20 1.2× 1013 1.0 × 1013 2.2 × 1013
Deformed 242 ± 60 154 ± 40 4.5 × 1012 1.8 × 1013 2.2 × 1013

The quantitative analysis of the distribution of cementite particles has been performed on at least 10 TEM micrographs for each condition. dθ: major
diameter.

Table 2. Dislocation density, spacing between GNBs and IDBs and average boundary spacing Da considered as the equivalent grain size.

Dislocation density Boundary spacing (nm) Da (nm)

Specimen
(m−2) DGNB DIDB

ITT HTT ITT HTT ITT HTT ITT HTT

Undeformed 7.2 × 10 14
1.5 × 10 12
328 ± 130 400 ± 143 173 ± 50 300 ± 68 165 ± 52 258 ± 69
Deformed 8.3 × 1014 5.3 × 1014 250 ± 110 435 ± 159 64 ± 25 174 ± 50 70 ± 27 176 ± 53
8 Materials Science and Technology 0(0)

Figure 7. (a) Variation of dislocation density with tempering temperature (botton line) and after cold-drawing (top line), measured by
XRD peak broadening characterisation, (b) Average boundary spacing before and after cold-drawing.

Finally, with higher deformation, LDC regions are formed. that act as a barrier and exert a certain resistance to disloca-
These deformation states were also observed in tempered tion slip.32,47 This fact is highlighted by dashed lines in
low-carbon steel subjected to cold rolling.16,19 Huang16 Figure 6(d), where the GNBs are delimited by thicker pre-
analysed the transition of lath martensite into a cell block cipitates. In contrast, in the ITT samples, the initial structure
structure and concluded that the driving force for the tran- of lath martensite is mostly reorganised to a finer structure
sition is a decrease in energy per unit length of dislocation as LDCs after final deformation, leading to a reduction in
line, possibly supplemented by a reduction in stored energy. the average GNB spacing. However, it is important to
According to several researchers,16,36,37,44 a key structural note that some GNBs were preserved during cold deform-
element for grain refinement is the cell block composed ation, especially those that were initially in the DD. This
of cell block boundaries or extended boundaries classified could be explained by the carbides acting as a barrier to
as ‘geometrically necessary boundaries (GNBs)’ and ordin- the dislocation glide, and the fact that the carbides remain
ary cell boundaries or interconnecting boundaries classified in the boundary plane parallel to the DD, as reported by
as ‘incidental dislocation boundaries (IDBs)’. The former Zhang et al.40 in cold-drawn pearlitic steel. This sets an
separates regions that deform by different slip system com- important difference with regards to deformation structures
binations, strain amplitudes, and strains, while the latter of low-carbon tempered steels, where the initial lath bound-
results from trapping of glide dislocations.45,46 Therefore, aries disappear, and dislocation cells are recon-
from this point of view, it is more appropriate to discuss structed.9,19,40 Alternatively, it is possible to use
the boundary spacing of both GNBs and IDBs instead of stereological analysis to calculate an equivalent grain size
lath width as a measure of grain size. The method employed in structures subdivided by GNBs and IDBs.16,44,45
to measure the boundary spacing of GNBs and IDBs in According to Hughes et al.,45 the magnitude of the
TEM images follows the approach proposed by Hughes average boundary spacing, Da, is considered as the equiva-
et al.,45 where the spacing between the boundaries is mea- lent grain size and is expressed as 2/Sv, where Sv is the total
sured in longitudinal sections containing the DD. In the area per unit volume given by the relationship:
present paper, 100 measurements of each class of boundary
were taken using ImageJ software, from a total of 10 TEM
images and the average value is presented in Table 2. The 1 π
Sv = + (1)
standard deviation is considered as a measure of error in DGNB 2DIDB
the boundary spacing measurements, which corresponds
to 35–40%, indicating the inhomogeneity of initial as well
as final structures. The results shown in Table 2 indicate a where DGNB and DIDB are the average GNB and IDB
reduction in both GNB and IDB spacing in the deformed spacing, respectively, indicated in Table 2. Da data are
ITT samples. However, in the HTT samples, only the represented in Figure 7(b) and Table 2 which show that
IDB spacing is reduced, while no significant changes in the average spacing decreases after cold drawing in both
the GNB spacing are observed. This indicates that the ITT and HTT specimens. In particular, Da decreased a
average GNB spacing is practically conserved after cold 42% in ITT samples reaching a value of 70 nm, whereas
drawing in HTT medium carbon steels, due to the high in HTT samples it decreases 68% reaching a value of
density of coarse cementite particles at the boundaries 176 nm.
Marinelli et al. 9

Figure 8. (a) Microhardness of ITT (top line) and HTT (bottom line), (b) engineering drawing stress as a function of accumulated area
reduction for specimen ITT25 (top line) and HTT (bottom line).

Mechanical properties drawing, surface quality is typically improved, which can

lead to a higher friction coefficient in the first stage as com-
Figure 8 illustrates the relationship between engineering
pared to the subsequent passes.
drawing stress, microhardness and accumulated area reduc-
tion (%AR) for ITT and HTT samples. The results demon-
strate that work hardening tends to increase with the strain
beyond 21% reduction, and is closely associated with an Conclusions
increase in dislocation density, and the microstructural The morphological and crystallographic characteristics of
changes from cell blocks to reconstructed LDC struc- lath martensite tempered at two different temperature
ture.16,19 While previous studies have shown that the hard- ranges, ITT and HTT, and the effect of cold drawing on
ness of cold-rolled low-carbon steels remains unchanged at the microstructure and mechanical properties were ana-
low strains (<20%AR),16 our current findings show that the lysed. The main results are summarised as follows:
hardness of both ITT and HTT samples in this cold-drawn
medium-carbon steel increases during the initial deform- • After tempering ITT and HTT samples show a lath mar-
ation stage. This can be attributed to the pinning of disloca- tensite structure parallel to {110} planes, with a distribu-
tions in cementite precipitates, and the barrier effect of tion of intralath and interlath cementite particles whose
interlath cementite, which reduces dislocation glide. coarsening kinetics differ for each type of particles and
Consequently, dislocations in laths are stored as threading are influenced by different coarsening mechanisms.
dislocations, dislocation tangles, and cell boundaries with Particularly, HTT samples show a significant reduction
low to medium misorientation angles. Notably, the initial in dislocation density within the laths and larger precipi-
increase in hardness for HTT samples is 20%, whereas tates than ITT samples.
for ITT samples, it is 11%. This difference may be due to • In the 51.8% AR cold-drawn specimens, a lamellar struc-
the more significant increase in dislocation density ture oriented nearly parallel to the DD was observed for
observed in the HTT samples, as shown in Figure 7(a). ITT and HTT samples. In particular, the ITT samples
However, it is worth noting that the hardness remains exhibit a dislocation structure of LDCs and cell blocks,
nearly unchanged between 13.8% and 21%AR, and a while cell block structure is mainly observed in HTT spe-
slight decrease in drawing stress by about 30 MPa is cimens. This dislocation structure is closely related to the
observed, particularly in the ITT samples. This behaviour increase of dislocation density after cold drawing. It was
could be attributed to the fact that the nominal drawing shown that the increase of dislocation density in HTT
stress is influenced not only by the material’s yield strength, samples is much higher, reaching approximately the
but also by other factors such as the true drawing strain. The same order of magnitude of ITT samples.
higher the true drawing strain, the greater the force required. • The martensite blocks composed of laths perpendicular
For the second stage (21%AR), the true drawing strain is to DD are resistant to deformation during drawing and
the lowest, which explains the reduction of the nominal they bend, developing a characteristic structure designed
drawing stress. Additional factors that can affect the as IBLs.
drawing stress include material-die friction and die angle. • The analysis of particle density revealed that after tem-
It is possible that material-die friction also played a role pering, the particles are mainly found at lath boundaries
in the observed variation in drawing stress. During cold in both specimens. However, after drawing, some lath
10 Materials Science and Technology 0(0)

boundaries disappear, causing the cementite particles, angular pressing: microstructure and tensile properties. Acta
previously located at boundaries, to become randomly Mater 2005; 53: 3125–3134.
oriented as evidenced by the presence of concentric 6. Halfa H. Recent trends in producing ultrafine grained steels.
rings in the SAED patterns. This results in a reduction JMMCE 2014; 02: 428–469.
in intergranular particle density and an increase in intra- 7. Arjomandi S, Jafarian HR and Park N. A substantial improve-
ment of tensile properties through microstructure engineering
granular particle density for both specimens. It is import-
in a high Ni-0.2C TRIP steel under severe plastic deformation.
ant to note that the total particle density remains Mater Charact 2019; 155: 109826.
practically unchanged after drawing, indicating no frag- 8. Okitsu Y, Takata N and Tsuji N. A new route to fabricate
mentation of particles. ultrafine-grained structures in carbon steels without severe
• It was observed that, after tempering, hardness of ITT plastic deformation. Scr Mater 2009; 60: 76–79.
samples was higher than HTT ones. In contrast to the 9. Ueji R, Tsuji N, Minamino Y, et al. Ultragrain refinement of
findings reported for cold-rolled low-carbon steels, the plain low carbon steel by cold-rolling and annealing of mar-
hardness increases during the initial strain stage for tensite. Acta Mater 2002; 50: 4177–4189.
both ITT and HTT samples. This can be attributed to 10. Tsuji N, Ueji R, Minamino Y, et al. A new and simple process
the pinning of dislocations in cementite precipitates, to obtain nano-structured bulk low-carbon steel with superior
and the barrier effect of interlath cementite, which mechanical property. Scr Mater 2002; 46: 305–310.
11. Sun C, Fu P, Liu H, et al. The effect of lath martensite micro-
reduces dislocation glide. On the other hand, beyond
structures on the strength of Medium-carbon low-alloy steel.
21% AR, work hardening tends to increase with the
Crystals 2020; 10: 232.
strain due to dislocation density increase at each step. 12. Zhao L, Park N, Tian Y, et al. Novel thermomechanical pro-
cessing methods for achieving ultragrain refinement of low-
Acknowledgements carbon steel without heavy plastic deformation. Mater Res
Secretary of postgraduate and international relations of the Lett 2017; 5: 61–68.
UNSJ. CONICET. Patricio Pérez and Luis Valdés from the 13. Song R, Ponge D, Raabe D, et al. Overview of processing,
microstructure and mechanical properties of ultrafine
Mechanical and Metallurgical Engineering Department at
grained bcc steels. Mater Sci Eng A 2006; 441: 1–17.
the Pontificia Universidad Católica de Chile for their valu- 14. Tsuji N. Advanced thermomechanical processing of steels. In:
able help in carrying out the bar drawing tests. Encyclopedia of materials: metals and alloys. Amsterdam,
Netherlands: Elsevier, 2022, pp.223–234.
Declaration of conflicting interests 15. Ghassemali E, Kermanpur A and Najafizadeh A.
The authors declared no potential conflicts of interest with Microstructural evolution in a low carbon steel during cold
rolling and subsequent annealing. J Nanosci Nanotechnol
respect to the research, authorship, and/or publication of
2010; 10: 6177–6181.
this article.
16. Huang X, Morito S, Hansen N, et al. Ultrafine structure and
high strength in cold-rolled martensite. Metall Mater Trans
Funding A 2012; 43: 3517–3531.
The authors disclosed receipt of the following financial 17. Ueji R, Tsuji N, Minamino Y, et al. Effect of rolling reduction
support for the research, authorship, and/or publication of on ultrafine grained structure and mechanical properties of
low-carbon steel thermomechanically processed from mar-
this article: This work was supported by the Universidad
tensite starting structure. Sci Technol Adv Mater 2004; 5:
Nacional de San Juan, Consejo Nacional de
153–162.
Investigaciones Científicas y Técnicas, 18. Maki T, Tsuzaki K and Tamura I. The morphology of micro-
structure composed of lath martensites in steels. ISIJ Int 1980;
ORCID iD 20: 207–214.
Virginia H. Aranda https://orcid.org/0000-0001-8939- 19. Morito S, Ohba T, Das AK, et al. Effect of solution carbon and
5560 retained austenite films on the development of deformation
structures of low-carbon lath martensite. ISIJ Int 2013; 53:
2226–2232.
References 20. Tokizane M, Matsumura N, Tsuzaki K, et al. Recrystallization
1. Weng Y, Dong H and Gan Y. Advanced steels. Berlin, and formation of austenite in deformed lath martensitic struc-
Heidelberg: Springer Berlin Heidelberg, 2011. ture of low carbon steels. Metall Trans A 1982; 13: 1379–1388.
2. Alexandrov I and Valiev R. Developing of SPD processing 21. Swarr T and Krauss G. The effect of structure on the deform-
and enhanced properties in bulk nanostructured metals. Scr ation of as-quenched and tempered martensite in an Fe-0.2 pct
Mater 2001; 44: 1605–1608. C alloy. Metall Trans A 1976; 7: 41–48.
3. Azushima A and Aoki K. Mechanical properties of ultrafine 22. Lan HF, Liu WJ and Liu XH. Ultrafine ferrite grains produced
grained steel produced by repetitive cold Side extrusion. by tempering cold-rolled martensite in low carbon and micro-
CIRP Annals 2002; 51: 227–230. alloyed steels. ISIJ Int 2007; 47: 1652–1657.
4. Saito Y, Utsunomiya H, Tsuji N, et al. Novel ultra-high strain- 23. Hayakawa M, Matsuoka S and Furuya Y. Atomic force
ing process for bulk materials—development of the accumu- microscopy of local plastic deformation for tempered mar-
lative roll-bonding (ARB) process. Acta Mater 1999; 47: tensite. In: Gdoutos EE (eds) Fracture of nano and engineer-
579–583. ing materials and structures. 2006, pp.233–234.
5. Son YI, Lee YK, Park K-T, et al. Ultrafine grained ferrite– 24. Hayakawa M, Matsuoka S and Furuya Y. Nanoscopic meas-
martensite dual phase steels fabricated via equal channel urement of local plastic deformation for a tempered
Marinelli et al. 11

martensitic steel by atomic force microscopy. Mater Lett http://www.seas.ucla.edu/matrix/papers/1985/1988-amod_

2003; 57: 3037–3042. ghon-review_exp_obs_theo_mod_dislo_cell_subgrn.pdf.
25. Aranda VH, Carazo FD, García LN, et al. Effect of cold 36. Lu K and Hansen N. Structural refinement and deformation
deformation in quench and tempered medium-carbon steel mechanisms in nanostructured metals. Scr Mater 2009; 60:
bars microstructure. Mater Sci Technol 2022; 38: 230–236. 1033–1038.
26. Lutterotti L. Maud: a rietveld analysis program designed for 37. Hansen N and Jensen DJ. Development of microstructure in
the internet and experiment integration. Acta Crystallogr A FCC metals during cold work. Philos Trans Royal Society
Found Crystallogr 2000; 56: s54–s54. London. Series A: Mathematical, Physical and Eng Sci
27. de Keijser TH, Langford JI, Mittemeijer EJ, et al. Use of the 1999; 357: 1447–1469.
voigt function in a single-line method for the analysis of 38. Wu YX, Sun WW, Styles MJ, et al. Cementite coarsening
X-ray diffraction line broadening. J Appl Crystallogr 1982; during the tempering of Fe-C-Mn martensite. Acta Mater
15: 308–314. 2018; 159: 209–224.
28. Williamson GK and Smallman RE III. Dislocation densities in 39. Mosayebi A, Soleimani M, Mirzadeh H, et al. Tempering
some annealed and cold-worked metals from measurements kinetics and corrosion resistance of quenched and tempered
on the X-ray debye-scherrer spectrum. Phil Mag 1956; 1: AISI 4130 medium carbon steel. Mater Corros 2021; 72:
34–46. 1808–1812.
29. Morito S, Huang X, Furuhara T, et al. The morphology and 40. Zhang X, Godfrey A, Hansen N, et al. Evolution of cementite
crystallography of lath martensite in alloy steels. Acta Mater morphology in pearlitic steel wire during wet wire drawing.
2006; 54: 5323–5331. Mater Charact 2010; 61: 65–72.
30. Krauss G. Strengthening mechanisms in steels. In: 41. Furuhara T, Kobayashi K and Maki T. Control of cementite
Encyclopedia of materials: science and technology. precipitation in lath martensite by rapid heating and temper-
Amsterdam: Elsevier, 2001, pp.8870–8881. ing. ISIJ Int 2004; 44: 1937–1944.
31. Krauss G. Tempering of lath martensite in low and Medium 42. Schneider CA, Rasband WS and Eliceiri KW. NIH Image to
carbon steels: assessment and challenges. Steel Research Int ImageJ: 25 years of image analysis. Nat Methods 2012; 9:
2017; 88: 1700038. 671–675.
32. Bhadeshia H and Honeycombe R. Tempering of martensite. 43. Nam WJ and Bae CM. Coarsening behavior of cementite par-
In: Steels: microstructure and properties. Amsterdam: ticles at a subcritical temperature in a medium carbon. Scr
Elsevier, 2017, pp.237–270. Mater 1999; 41: 313–318.
33. Beladi H, Rohrer GS, Rollett AD, et al. The distribution of 44. Godfrey A and Hughes D. Scaling of the spacing of deform-
intervariant crystallographic planes in a lath martensite ation induced dislocation boundaries. Acta Mater 2000; 48:
using five macroscopic parameters. Acta Mater 2014; 63: 1897–1905.
86–98. 45. Hughes D and Hansen N. Microstructure and strength of
34. Liu T, Li T and Liu X. TEM And electron diffraction analysis nickel at large strains. Acta Mater 2000; 48: 2985–3004.
of ω-Fe to cementite transformation in quenched and tem- 46. Hughes DA. Microstructural evolution in a non-cell forming
pered high carbon steels. AIP Adv 2019: 9. DOI: 10.1063/1. metal: al(mg. Acta Metall Mater 1993; 41: 1421–1430.
5091847 47. Zhao X, Shen Y, Qiu L, et al. Effects of intercritical
35. Amodeo RJ and Ghoniem N. A review of experiemtal obser- annealing temperature on mechanical properties of
vations and theoretical models of dislocatios cells and sub- Fe-7.9Mn-0.14Si-0.05Al-0.07C steel. Materials (Basel)
grains. Res Mechanical 1988; 23: 137–160. Available from: 2014; 7: 7891–7906.