Metallurgical and Materials Transactions A

, Volume 43, Issue 6, pp 2043–2055

Texture and Microstructural Evolution in Pearlitic Steel During Triaxial Compression

Authors

  • Pankaj Kumar
    • Department of Materials EngineeringIndian Institute of Science
  • Nilesh P. Gurao
    • Department of Materials EngineeringIndian Institute of Science
  • Arunansu Haldar
    • Tata Steel, Research and Development Section
    • Department of Materials EngineeringIndian Institute of Science
Article

DOI: 10.1007/s11661-011-1043-y

Cite this article as:
Kumar, P., Gurao, N.P., Haldar, A. et al. Metall and Mat Trans A (2012) 43: 2043. doi:10.1007/s11661-011-1043-y

Abstract

This article presents the deformation behavior of high-strength pearlitic steel deformed by triaxial compression to achieve ultra-fine ferrite grain size with fragmented cementite. The consequent evolution of microstructure and texture has been studied using scanning electron microscopy, electron back-scatter diffraction, and X-ray diffraction. The synergistic effect of diffusion and deformation leads to the uniform dissolution of cementite at higher temperature. At lower temperature, significant grain refinement of ferrite phase occurs by deformation and exhibits a characteristic deformation texture. In contrast, the high-temperature deformed sample shows a weaker texture with cube component for the ferrite phase, indicating the occurrence of recrystallization. The different mechanisms responsible for the refinement of ferrite as well as the fragmentation of cementite and their interaction with each other have been analyzed. Viscoplastic self-consistent simulation was employed to understand deformation texture in the ferrite phase during triaxial compression.

1 Introduction

Materials with ultra-fine grains produced by severe plastic deformation (SPD) exhibit improved mechanical and physical properties.[1,2] The characteristic feature of SPD includes retention of the dimension of the work piece even after imparting large strain. Various SPD processes such as equal channel angular extrusion, high-pressure torsion, and mechanical milling are well reported for producing ultra-fine-grained materials. However, complicated dies involved in many of these processes limit their applicability on industrial scale. Two processes have potential for scaling up, namely, accumulative roll bonding (ARB) and multiaxial forging (MAF). ARB can be used to produce flat products, whereas MAF can impart a large amount of strain in large billets, which can be used to obtain any desired shape. This process involves a combination of uniaxial and plane strain compression along the three directions normal to the faces of a cuboidal work piece. A similar but much simpler manifestation of the process is triaxial compression in which the work piece is uniaxially compressed successively on three mutually perpendicular faces. This process is based on normal up-set forging and, therefore, does not involve any die. MAF or triaxial compression has been carried out on a variety of materials[36]; however, most studies are limited to single-phase ductile metals and alloys. Investigations of ferrous materials have been rare.[7,8] One such material with tremendous industrial applications is pearlitic steel.

The microstructure of pearlitic steel consists of alternate lamellae of ferrite (α-Fe) and cementite (Fe3C) phases arranged in colonies with random orientations. It has excellent mechanical properties such as high strength, good wear resistance, high fatigue strength, etc. Cementite is the harder phase that imparts strength to pearlitic steel. It has been reported that coarse lamellar cementite reduces the ductility of steels and hence significantly limit its application. Breaking these cementite lamella causes refinement and, hence, improvement in properties such as increase in toughness, formability, and machinability. Several studies have been focused on the change in morphology of the cementite lamella[810] and also on the evaluation of mechanical properties associated with the resulting microstructure. These investigations suggest that a composite microstructure with tiny cementite particles distributed in ultra-fine ferrite matrix yields a good balance between strength and ductility.

The present study is primarily aimed at examining the effect of triaxial compression on microstructural refinement in pearlitic steel. Different combinations of the strain path and temperature have been attempted and the resulting microstructural features have been examined. Because such large deformations are associated with the development of crystallographic texture, which has a large impact on mechanical properties, it is imperative to study the evolution of texture during such processes. Moreover, the trends in evolution of crystallographic texture can also indicate the dominant deformation mechanism. This study is complemented by crystal plasticity simulations for texture development using the viscoplastic self-consistent (VPSC) model.

2 Experimental

2.1 Starting Material

The present study was carried out on pearlitic (eutectoid) steel provided by TATA Steel Ltd. (Jamshedpur, India) in the form of hot-rolled plate that exhibited a completely lamellar structure. The detailed composition of the steel is given in Table I.
Table I

Chemical Composition (Weight Percent) of the As-Received Sample

Composition

C

Mn

S

P

Al

Cu

Cr

Wt pct

0.8

0.63

0.1

0.014

0.042

0.005

0.053

2.2 Processing

The specimen for triaxial compression was extracted from the hot-rolled plate in the form of cubes with dimension 10 mm × 10 mm × 10 mm. The samples were uniaxially compressed successively on the three orthogonal planes, X, Y, and Z. The schematic diagram representing the triaxial compression test is shown in Figure 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig1_HTML.gif
Fig. 1

Schematic diagram of triaxial compression (not accurate in dimension)

In triaxial compression, one complete cycle consisted of three consecutive passes. The samples were briefly polished before testing. During the test, a true strain (ε = ln(h0/h)) of 0.46 was imparted at each compression pass. Each test was carried out to one complete cycle, consisting of three consecutive compression passes. A strain rate of 0.001 s−1 was used for the compression experiments. The same strain value imparted in all three directions should theoretically retain the sample dimension at the end of the cycle. However, the sample did bulge during compression, and the sample was lightly polished to make the faces flat before every deformation step. With regard to processing, two different strategies were adopted. In the first scheme, known as Route I, the sample was triaxially compressed at temperature of 773 K (500 °C). In the second scheme, referred to as Route II, the temperature was varied during each pass (i.e., the first pass was carried out at 673 K [400 °C], second pass at 623 K [350 °C], and the third pass at 603 K [330 °C]). The processes parameters for both routes are listed in Table II. The samples were coated with alumina using ethanol for lubrication during the deformation process.
Table II

Processing Parameters Used for the Two Different Tests

Process

ε/Pass

Temp. at First Pass

Temp. at Second Pass

Temp. at Third Pass

Route I

0.46

773 K (500 °C)

773 K (500 °C)

773 K (500 °C)

Route II

0.46

673 K (400 °C)

623 K (350 °C)

603 K (330 °C)

2.3 Characterization of Microstructure and Texture

The deformed samples were machined at midthickness of Z face to get flat surfaces, and a detailed investigation was carried out using X-ray diffraction (XRD), whereas scanning electron microscopy (SEM) and electron back-scatter diffraction (EBSD) were carried out on the mid-thickness of the X face. Each sample was mechanically polished followed by electropolishing and etching to reveal the microstructure. Electropolishing was carried out using 10 pct perchloric acid in methanol, whereas picral (4 pct picric acid and 96 pct ethanol) was used as an etchant. EBSD scans were recorded using a field emission gun SEM (FEG-SEM). XRD studies were carried out to examine the microstructural evolution in terms of crystallite size. Hardness tests were performed using Vickers hardness tester with a load of 3 kg. Five readings were taken, and the mean value was obtained. Bulk texture measurement was performed using X-ray texture goniometer based on Schulz reflection method. The Co Kα radiation was employed for this. Three incomplete pole figures namely, (110), (200), and (211) were measured at the midsection of the Z face of the samples. The orientation distribution functions (ODFs) were calculated from the experimental pole figures using commercially available LaboTex software with the arbitrary discretized cells[11] algorithm.

3 Results

3.1 Starting Material

Figure 2 shows the microstructure of the as-received sample at two magnifications. Figure 2(a) shows the size of the pearlitic colony, whereas Figure 2(b) reveals the fully lamellar pearlite structure consisting of ferrite and cementite lamellas with inter-lamellar spacing ranging from 400 nm to 1 μm. The X-ray diffraction pattern of the as-received sample (Figure 3(a)) shows the ferrite peaks in the starting material. The cementite peaks being weaker in intensity were not visible. The (101) pole figure for the ferrite phase shown in Figure 3(b) indicates an almost random texture.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig2_HTML.jpg
Fig. 2

SEM images of the as-received sample (a) at lower magnification and (b) at higher magnification

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig3_HTML.gif
Fig. 3

(a) XRD pattern and (b) 101 pole figure for the ferrite phase of the as-received sample

3.2 Evolution of Microstructure

Figure 4 displays the microstructure of the material after triaxial compression for the Route I sample that was deformed to one cycle at 773 K (500 °C). The microstructural features indicate complete fragmentation and near globurization of cementite (Figure 4(a)). Figure 4(b), which represents the microstructure from another region of the same sample, shows the fragmentation of cementite. It is evident that thinning of cementite lamellae is followed by fragmentation and partial globularization. The representative SEM micrographs of the sample deformed to one cycle by Route II on three different faces are shown in Figure 5. These micrographs show that the fragmentation of lamellar cementite occurs in some grains but not in all them.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig4_HTML.jpg
Fig. 4

Microstructural features of the material after one cycle of triaxial compression by Route I at 773 K (500 °C)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig5_HTML.jpg
Fig. 5

SEM images of the sample deformed by Route II (at three different temperatures) on three faces

The phase and inverse pole figure (IPF) maps of the deformed samples as obtained from EBSD are shown in Figure 6. The globularization of cementite in Route I deformed sample and fragmentation of cementite in Route II sample is clearly evident. In addition, significant refinement of the ferrite phase is evident from the IPF map of the Route II sample that also shows some recrystallized grains (marked A). The average grain size of the ferrite phase, as obtained from EBSD with a criterion of 5 deg misorientation, was 0.78 μm and 0.36 μm for ferrite phase in Route I and Route II samples, respectively. The misorientation distribution plots for the samples (Figure 7) show a strong peak in the low-angle boundary regime (<15 deg) and a diffuse population of peaks in the high-angle boundary regime (>15 deg). The sample deformed by Route I shows a lower fraction of high-angle grain boundaries (HAGB) and a higher fraction of low-angle grain boundaries (LAGB) compared with the Route II sample (Figures 8 (a) and (b). In addition, the fraction of coincidence site lattice (CSL) boundaries observed in Route II processed samples is almost twice that of the sample processed through Route I. The distribution shows that most CSL boundaries are of ∑3 and ∑13b type with Route I, whereas for Route II, the fraction of ∑3 boundaries is almost twice that of ∑13b boundaries.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig6_HTML.gif
Fig. 6

(a) Phase map of sample deformed by Route I. (b) IPF map of ferrite of sample deformed by Route I. (c) Phase map of sample deformed by Route II. (d) IPF map of ferrite of sample deformed by Route II (for appropriate representation in color, see the web version of this article)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig7_HTML.gif
Fig. 7

Misorientation angle distribution for the ferrite phase of the samples deformed by (a) Route I and (b) Route II

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig8_HTML.gif
Fig. 8

Grain boundary character distribution of sample deformed by (a) Route I and (b) Route II, and CSL boundaries distribution of sample deformed by (c) Route I and (d) Route II

EBSD investigation also revealed the evolution of strong intragranular misorientation and grain fragmentation in the ferrite phase. The scalar and vector measure of intragranular misorientation in terms of grain orientation spread (GOS) and grain average misorientation (GAM) was also determined. The Route I sample showed a higher value of GAM and GOS (2 deg and 4.3 deg, respectively) compared to the Route II sample (1.4 deg and 3.1 deg). The X-ray line-profile analysis indicated an increase in crystallite size with an increase in strain for Route I sample, whereas the Route II sample showed a small crystallite size (Table III).
Table III

Crystallite Size (Å) of the Samples Determined by Line-Broadening Analysis

Process

Pass 1

Pass 2

Pass 3

Route I

774 Å (±91)

894 Å (±55)

1200 Å (±106)

Route II

290 Å (±90)

3.3 Hardness Measurements

For the Route I sample, after the first pass, a slight increase was observed in the hardness from 290 VHN to 300 VHN (Figure 9). During the subsequent passes, hardness decreased with an increase in strain. However, the Route II sample showed a hardness of 380 VHN, which is higher than the initial material as well as the Route I samples. However, this is lower than that of cold-drawn wires of pearlitic steel subjected to a similar amount of deformation.[12]
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig9_HTML.gif
Fig. 9

Variation in the hardness value for the samples deformed by different routes

3.4 Texture Evolution

Figure 12 shows the recalculated (101) pole figures of the samples using X-ray diffraction. For easy comprehension, the texture evolution is depicted using the standard rolling coordinate system described by {hkl}\( \left\langle {\text{uvw}} \right\rangle \), where {hkl} is the normal direction (ND) plane and \( \left\langle {\text{uvw}} \right\rangle \) is the compression direction (CD) and is equivalent to the rolling direction (RD). Certain weak orientations are observed in the processed samples (Figures 10 and 11). The overall texture is weak. The major components and their corresponding volume fractions for the samples deformed by Routes I and II are given in Tables IV and V, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig10_HTML.gif
Fig. 10

Experimental (101) pole figure for the ferrite phase of samples deformed by (a) Route I, (b) Route II, and (c) representative orientations

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig11_HTML.gif
Fig. 11

ODFs for the ferrite phase of samples deformed by (a) Route I and (b) Route II

Table IV

Major Texture Components Observed in Route I

Component Designation

{hkl} \( \left\langle {\text{uvw}} \right\rangle \)

Component Description/Name

Volume Fraction

A

\( \left\{ {001} \right\}\left\langle {6\bar{1}0} \right\rangle \)

9.5 deg about \( \left\langle {00\bar{1}} \right\rangle \) with \( \left\{ {001} \right\}\left\langle {100} \right\rangle \)

2.10

B

\( \left\{ {\bar{1}\bar{1}\bar{1}} \right\}\left\langle {\overline{1} 10} \right\rangle \)

TC

2.12

C

\( \left\{ {110} \right\}\left\langle {1\bar{1}2} \right\rangle \)

Brass

1.94

D

\( \left\{ {001} \right\}\left\langle {100} \right\rangle \)

Cube

1.91

Table V

Major Texture Components Observed in Route II

Component Designation

{hkl} \( \left\langle {\text{uvw}} \right\rangle \)

Component Description/Name

Volume Fraction

D

\( \left\{ {001} \right\}\left\langle {100} \right\rangle \)

Cube

8.66

C

\( \left\{ {110} \right\}\left\langle {1\bar{1}2} \right\rangle \)

9.5 deg about \( \left\langle {\overline{1} 10} \right\rangle \) with \( \left\{ {110} \right\}\left\langle {\overline{1} 10} \right\rangle \)

8.10

A

\( \left\{ {001} \right\}\left\langle {6\bar{1}0} \right\rangle \)

9.5 deg about \( \left\langle {00\bar{1}} \right\rangle \) with \( \left\{ {001} \right\}\left\langle {100} \right\rangle \)

5.95

It is observed that \( \left\{ {001} \right\}\left\langle {100} \right\rangle \) and \( \left\{ {001} \right\}\left\langle {6\bar{1}0} \right\rangle \) components evolved as major texture components with \( \left\{ {110} \right\}\left\langle {1\bar{1}2} \right\rangle \) in both cases. A key figure depicting the ideal orientations is given in Figure 10(c).

3.5 Texture Simulations

The viscoplastic self consistent code VPSC 7b[13,14] has been used to simulate deformation texture evolution. The model is based on an interaction equation that couples the stress–strain rate in the grain with the overall stress–strain rate of the bulk (homogenous effective medium). The algorithm enables one to choose the initial texture, active slip system/s, rate sensitivity, hardening parameter, etc. Once the slip system activity is obtained, grain rotation can be estimated, which leads to texture evolution.

An aggregate of 2000 random oriented grains were given as input to the simulation code. The effect of hard second-phase cementite was not modeled as it lacks plasticity at the operating temperatures in the present investigation. Simulations were carried out only for the Route II sample as the Route I sample shows a weak cube texture indicative of recrystallization. The {110}\( \left\langle {111} \right\rangle \) slip system was used to model the texture evolution for Route II. The inverse rate sensitivity exponent was taken as 20 (strain rate sensitivity ~0.05) for the activated slip system. In the first pass of compression, the velocity gradient used to model the deformation texture was L11 = 0.5, L22 = 0.5, and L33 = –1; in the second pass, the velocity gradient was L11 = 0.5, L22 = –1, and L33 = 0.5; and in the third pass, the velocity gradient was L11 = –1, L22 = 0.5, and L33 = 0.5. The results of simulation are presented in Figure 12. Reasonable agreement was found between the experimental and the simulated texture of Route II sample. However, simulations could not account for the central contour observed in the experiment. This finding could be attributed to various factors like the operation of a multiple slip system or the presence of cementite or recrystallization.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig12_HTML.gif
Fig. 12

(101) pole figures of (a) simulation and (b) experimental measurements for the sample deformed by Route II

4 Discussion

The results of the present investigation clearly indicate that the second strategy (Route II) comprising a gradual reduction in temperature during the course of deformation leads to more grain refinement. Various aspects of the evolution of microstructure and texture during the processes associated with the two routes and their relationship with improved strength are discussed.

4.1 Cementite Dissolution and Grain Refinement in Ferrite

The dissolution of cementite lamella in the ferrite matrix in Route I can be explained based on the carbon-dislocation interaction during plastic deformation. The moving dislocations crossing the cementite may drag the carbon atom from them into the ferrite.[15,16] The second explanation is based on the thermodynamic instability, which could be the driving force for the dissolution.[1724] The interfacial energy might play an important role in the process. An increase in the interfacial energy is likely to increase the molar free energy of the cementite and, hence, increase the carbon content in the ferrite phase according to the Gibb–Thompson effect.[25] The interfacial area of the cementite lamellae changes differently in the individual pearlitic colonies depending on the stress distribution. It has been shown that the amount of dissolved cementite increases with an increase in the interfacial area.[1618,26] During the course of deformation, geometrical necessary dislocations (GNDs) are generated at the ferrite–cementite interface to maintain strain compatibility. These dislocations result in high interfacial stress leading to perturbation (kinks) at the cementite lamellae. During deformation at 773 K (500 °C) in Route I, numerous dislocation loops form in the cementite phase[27] near the perturbation, consequently increasing the interfacial energy locally. Cementite is, thus, locally destabilized to maintain the local equilibrium. As a result, a carbon concentration gradient appears in the ferrite phase as the equilibrium carbon solubility in ferrite increases near the local increased interfacial area according to the Gibbs–Thompson effect. If the mobility of carbon is high, then the diffusion of carbon takes place from the interface to the ferrite core (downhill diffusion) and leads to the dissolution of the cementite. This process favors the elimination of the interfacial dislocations,[17,18,2022] thus leading to a decrease in the energy of the system. According to the following Gibbs–Thomson equation, which relates the chemical potential and radius of curvature, the chemical potential of carbon in the vicinity of the lamella having small radius of curvature will be higher compared with that of a larger curvature (see Eqs. [1] and [2]).
$$ X_{\text{eq}}^{\alpha r} = X_{\text{eq}}^{\alpha \infty } \exp \left( {\frac{{2\gamma V_{\text{m}} }}{{r{\text{R}}T}}} \right) $$
(1)
where T is the temperature, γ is the surface energy, and Vm is the molar volume.In generalized form, the equation is expressed as follows[28]:
$$ \frac{{2\gamma v_{a}^{\beta } t}}{rkT} = (1 - X_{p} )\ln \left( {\frac{{1 - X_{\text{eq}}^{r} }}{{1 - X_{\text{eq}}^{\infty } }}} \right) + X_{p} \ln \left( {\frac{{X_{\text{eq}}^{r} }}{{X_{\text{eq}}^{\infty } }}} \right) $$
(2)
Where \( v_{a}^{\beta } \) is the average atomic volume and Xp is composition of β phase.
The diffusion of carbon atoms occurs from regions of lower curvature through dislocations, interface, etc., resulting in the thinning of the lamellae locally (Figure 10). It is reported[24,29] that the formation of excess vacancies during deformation promotes carbon diffusion especially near lamellae perturbation (kinks). Hence, the potential site for the thinning of cementite lamellae is the lower curvature region. During deformation, the increased potential site accelerates the diffusion process and causes faster dissolution of the lamella. Because substructures in ferrite are within the high diffusion rate zone in the lamellae, thinning of cementite mainly occurs near the dislocation substructure in ferrite. The schematic view of cementite dissolution is shown in Figure 13.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig13_HTML.gif
Fig. 13

A schematic diagram of the fragmentation of cementite (a) in a colony, (b) of single lamella

However, for the sample deformed by Route II, the morphology of the cementite particles is different, with features that indicate a cleavage fracture of the cementite phase. The carbon atoms redistribute near the interface because the diffusion of carbon in ferrite is low at lower temperatures. Because of the low temperature of deformation for the Route II sample, neither diffusion nor dislocation annihilation are dominant mechanisms and a higher fraction of GND is expected. In addition, the trapping of dislocations leads to the formation of statistically stored dislocations (SSDs) in the ferrite phase. The density of SSDs depends on the plastic strain in the material, whereas the GNDs depend on the plastic strain gradient in the material.[30] Thus, as the strain increases, both SSDs and GNDs increase at the interface, which leads to a highly stressed interface. Hence, the brittle fractures of cementite in the steel sample takes place simply by shear,[31] however, globularization does not occur because of the lack of carbon diffusion.

A FEG electron probe microanalyzer (EPMA) with a probe diameter of 40 nm and interaction volume of ~1 μm (at an operating voltage of 15 kV) with a layered-dispersed detector was employed to study the distribution of carbon. The line scan running from left to right at the center of the micrographs (Figure 14) obtained from EPMA shows a saw-tooth profile for carbon concentration for the initial sample. The presence of a saw-tooth profile instead of a square-wave profile can be attributed to the higher interaction volume of the electron probe. The maxima corresponds to the concentration of carbon in cementite phase, whereas the minima corresponds to the carbon concentration in the ferrite phase. The initial sample has a regular pattern in carbon distribution. However, after deformation, deviation from the regular pattern occurs, which clearly indicates redistribution of carbon in the ferrite phase. The ratio of minimum to maximum carbon content is 0.38 for the Route I sample, whereas a value of 0.32 is obtained for the Route II sample. This clearly indicates that diffusion was more dominant in the former and led to the enrichment of ferrite with carbon, thereby reducing the ratio of minimum to maximum carbon content. This clearly indicates that a higher fraction of cementite dissolved in the ferrite phase for the Route I sample compared with the Route II sample.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig14_HTML.gif
Fig. 14

Line profile analysis of carbon concentration for (a) initial, (b) Route I, and (c) Route II by EPMA

4.2 Texture Evolution

The presence of stronger cube and rotated cube textures in the Route II sample relative to the Route I sample can be attributed to pronounced recrystallization in the former. However, the increased hardness (Figure 11) and reduced grain size as well as crystallite size for the Route II sample suggests that these components cannot form because of recrystallization. Recrystallization can contribute to the weakening of texture in the ferrite phase as observed in the sample deformed by Route I. This can be attributed to the nucleation of random orientations of ferrite phase near the cementite particles. A reasonable match between the experimental and the simulated texture indicates that the formation of a strong texture in the Route II deformed sample is the result of the complicated strain path change associated with the process.

4.3 Evolution of Grain Boundary Character Distribution in the Ferrite Phase

A detailed insight of the GBCD for the two routes is obtained by plotting cumulative misorientation distribution plots (Figure 15(a)) for the two samples. The deviations from the McKenzie distribution suggest the presence of strong texture in the samples.[32] To study the distribution of dislocation in ferrite during the deformation, the distribution of kernel average misorientation (KAM)[33] has been calculated (Figure 15(b)). For Route II, the distribution of KAM is split into two regions—a low KAM region and a high KAM region—whereas for the Route I sample, no split is observed. This observation essentially shows that a dislocation arrangement for the Route II sample that is absent in the Route I sample. However, the peak shifted to a high KAM value for Route I, which shows an increased dislocation density. The presence of cementite globules could be the reason for this. To show the dislocation-mediated mechanism, the misorientation plot normalized to the scaling factor θ/θav is plotted (Figure 15(c)). The misorientation angles normalized to their average value plotted against the frequency distribution of the grains does not coincide. This result supports the theory that the morphology of cementite affects the deformation substructure evolution in ferrite.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-011-1043-y/MediaObjects/11661_2011_1043_Fig15_HTML.gif
Fig. 15

(a) Cumulative misorientation distribution, (b) KAM distribution, and (c) misorientation plot normalized to the scaling factor θ/θav for the ferrite phase of the samples deformed by Routes I and II

The higher fraction of CSL boundaries in the Route II deformed sample can be attributed to the stronger texture. Vitek et al.[34] have shown that the formation of Σ13 boundaries in base-centered cubic have one of the lowest energy values and therefore occur more readily. The formation of numerous Σ13b boundaries with a coincidence angle of 67.38 deg about \( \left\langle {111} \right\rangle \) may be explained by the presence of the relatively strong texture in the Route II sample.

5 Conclusions

The present investigation was aimed at examining the extent of microstructural refinement and subsequent development of crystallographic texture resulting from triaxial compression of pearlitic steel. The analysis of experimental and simulation results led to the following conclusions:
  1. 1.

    Triaxial compression produced fragmented and globularized cementite in a ultra-fine-grain size matrix of ferrite phase in pearlitic steel. A grain size as fine as ~0.36 μm could be achieved for the ferrite phase through a suitable combination of process parameters.

     
  2. 2.

    The cementite phase underwent fragmentation and partial globularization during deformation. The local thermodynamic instability of cementite lamellae has been accounted for the dissolution of cementite. At high temperature, the uniform fragmentation and globularization of cementite is ascribed to the synergistic effect of diffusion and deformation, whereas the nonuniform fragmentation of cementite at a lower temperature is attributed to deformation-dominated process.

     
  3. 3.

    Strain-induced cementite dissolution occurs during plastic deformation of pearlitic steels. It has been shown that cementite decomposition is caused by the carbon depletion from the cementite followed by lamellae thinning, fragmentation, and finally, globularization. This process leads to carbon enrichment of the ferrite phase.

     
  4. 4.

    Simulations indicate that the development of predominant \( \left\{ {001} \right\}\left\langle {100} \right\rangle \) and \( \left\{ {001} \right\}\left\langle {6\bar{1}0} \right\rangle \) components can be explained by a change in strain path during the deformation of ferrite. This finding clearly indicates that the synergistic effect of the slip-dominated process and diffusion can be exploited to achieve fragmented cementite in ultra-fine ferrite phase matrix.

     

Acknowledgments

The authors duly acknowledge TATA Steel Ltd., India for the financial grant and for providing the material to carry out the experiments. Thanks are owed to Dr. D. Bhattacharjee and Prof. R.K. Ray for their constant encouragement at various stages of this project. We thank Mr. S. Sasidhara for help in carrying out the triaxial compression tests using DARTEC. We acknowledge Department of Science and Technology, Govt. of India for the microscopy facility at the Institute of Nanoscience Initiative and the central X-ray facility at Indian Institute of Science, Bangalore. We are grateful to Professor I. Samajdar for providing access to the national facility for texture and OIM (DST-IRPHA) at I.I.T. Bombay.

Copyright information

© The Minerals, Metals & Materials Society and ASM International 2012