Formation Mechanism of Pearlite Colony by Multiple Orientation Relationships between Ferrite and Cementite

Abstract

To better understand the formation mechanism of lamellar pearlite, the orientation relationship (OR) between ferrite and cementite in spheroidized pearlite was analyzed using electron backscattered diffraction (EBSD) over a wide area and with high precision. Lamellar cementite was sufficiently spheroidized and coarsened, and substructures of lamellar pearlite, i.e., block and colony boundaries, were maintained even after a long period of spheroidized annealing. As a result, the cementite spheroidization facilitated the indexing of the cementite orientation in the EBSD analysis. It was consequently discovered that three different ORs—Bagaryatsky, Isaichev, and Pitsch–Petch ORs—were simultaneously established in the pearlite. Each OR deviated from the theoretical one, and the deviation was significantly larger in the Pitsch–Petch OR than in the other two. Also, the ferrite/cementite boundaries satisfying each OR were unevenly distributed, and the transition of ORs tended to correspond to colony boundaries with low-angle misorientation. From these results, the pearlite colony boundary was classified into three types: (1) the colony boundary maintaining an orientation relationship with different cementite variants, (2) the colony boundary characterized by two different ORs with low-angle misorientation, i.e., the transition between the Bagaryatsky and Isaichev ORs, and (3) the colony boundary with high-angle misorientation, i.e., the transition between the Isaichev and Pitsch–Petch ORs. In addition to the multiple variant formation, the coexistence of multiple ORs contributed to the change in lamellar alignment, with a high degree of freedom in the pearlitic transformation.

1. Introduction

Pearlite is the transformed microstructure formed through a eutectoid reaction in Fe–C systems, exhibiting a fine lamellar structure owing to the cooperative growth of ferrite (α) and cementite (θ). In addition to the lamellar structure, pearlite has hierarchical substructures known as pearlite colonies, where the cementite lamellae are aligned in an identical direction.¹⁾ A larger region composed of several colonies in which the ferrite matrix has an identical crystal orientation is defined as a pearlite block and a nodule.¹⁾ It is thought that lamellar cementite has the same crystal orientation in the pearlite block²⁾ and maintains a unique orientation relationship (OR) with respect to the ferrite matrix.¹⁾ The following four ORs have been reported:

Bagaryatsky OR³⁾ (010)_θ // (112)_α, [001]_θ // [110]_α, [100]_θ // [111]_α

Isaichev OR⁴⁾    (011)_θ // (112)_α, [011]_θ // [110]_α, [100]_θ // [111]_α

Pitsch–Petch OR⁵⁾ (010)_θ // (215)_α, [001]_θ 2.6° from [311]_α, [100]_θ 2.6° from [131]_α

Unknown OR⁶⁾   (011)_θ // (215)_α, [011]_θ 2.6° from [311]_α, [100]_θ 2.6° from [131]_α

Figures 1(a) and 1(b) show the unit cells of cementite and ferrite, respectively, where the crystal planes for the parallel-plane relationships are hatched. The iron atomic configurations in these planes are illustrated for the (010) and (011) planes in cementite [Figs. 1(c) and 1(d), respectively] and for the (112) and (215) planes in ferrite [Figs. 1(e) and 1(f), respectively]. Therefore, the comparisons between (c)–(e), (d)–(e), (c)–(f), and (d)–(f) indicate the lattice correspondence on the lamellar interface in the four ORs, respectively. The Bagaryatsky, Isaichev, and Pitsch–Petch ORs have been identified in previous studies.^3,4,5) However, the fourth OR was proposed by Zhou and Shiflet but has never been confirmed.⁶⁾ Thus, it is called the Unknown OR in this study. It is traditionally thought that α/θ ORs in pearlite vary depending on the carbon composition in steel; the Pitsch–Petch OR tends to be established near the eutectoid composition, whereas the Isaichev and Bagaryatsky ORs are observed in steels with hypoeutectoid and hypereutectoid compositions, respectively.³⁾ The reason for the variation in ORs depending on the carbon composition is explained in terms of the differences in the nucleation sites of pearlite accompanied by the precipitation of proeutectoid ferrite and cementite at the austenite grain boundaries. However, the explanation is unpersuasive because it is based on local observation using transmission electron microscopy. Instead, the authors have studied the internal stress in pearlite originating from the misfit between ferrite and cementite.⁷⁾ Based on the measurement results of the macroscopic internal stress state using neutron diffractometry, we proposed the possibility that some ORs can coexist, even in pearlite without proeutectoid ferrite or cementite. Considering that α/θ ORs may have a significant influence on the internal stress state affecting mechanical properties, it is important to clarify the influence of ORs in pearlite. Therefore, it is necessary to analyze ORs over a wide area containing substructures of pearlite while maintaining high spatial and angular resolution. The electron backscattered diffraction (EBSD) technique with scanning electron microscopy (SEM) is therefore appropriate for investigating this issue. However, it is not easy to analyze the crystallographic orientation of cementite in pearlite using EBSD. Cementite has a lower crystal symmetry because of its orthorhombic crystal structure, which makes indexing the EBSD pattern difficult. Furthermore, when an EBSD pattern reflected from cementite is overlapped with that from ferrite with higher crystal symmetry at the α/θ pearlite interface, the overlapped EBSD pattern is preferentially regarded as ferrite.⁸⁾ It has been reported that it is easier to analyze the orientation of cementite particles in spheroidized pearlite by EBSD, even if they are relatively fine.⁹⁾ Therefore, if the orientations of both ferrite and cementite are maintained during the spheroidization process, it is expected that the α/θ ORs in pearlite can be extensively analyzed not with lamellar pearlite but with spheroidized pearlite.

Fig. 1.

Unit cells of (a) cementite and (b) ferrite. Fe atomic configurations of the (c) (010) and (d) (001) planes of cementite and the (e) (112) and (d) (215) planes of ferrite.⁶⁾

In this study, to comprehensively understand the microstructure of pearlite, EBSD was used to analyze α/θ ORs over a wide area and with high precision with spheroidized pearlite. In addition, the formation mechanism of pearlite colonies is discussed in terms of the coexistence of multiple ORs.

2. Experimental Procedure

The material used in this study was a eutectoid steel with the chemical composition Fe–0.83%C–1.46%Mn (mass%); the quench hardenability was enhanced by the addition of Mn. Samples with dimensions of 20^l × 10^w × 10^t mm³ were cut from a hot-rolled steel sheet homogenized at 1523 K for 36 ks. They were solution treated at 1323 K for 1.8 ks and then directly subjected to isothermal holding in a salt bath furnace at 923 K for various periods up to 360 ks, followed by water quenching. The microstructure was observed by optical microscopy (OM) and SEM (JSM-7001F system; JEOL Ltd.). Crystallographic analysis was performed by EBSD using SEM. Phase maps and inverse pole figure maps (IPF maps) were obtained using OIM Data Collection ver. 7.1.0, and the OIM Analysis ver. 7.3.0 (TSL Solutions) for EBSD analysis. The lattice parameters of cementite were set to the default values in the OIM Data Collection (a = 0.5090 nm, b = 0.6748 nm, c = 0.4523 nm). The working distance was fixed at 15.0 mm, the accelerating voltage was 15 kV, and the step size was 0.05 µm for the detailed analysis. EBSD data with confidence index (CI) values higher than 0.1 were applied for the subsequent analysis.

To analyze the crystallographic orientation at the α/θ interface in detail, Euler angles (${\varphi }_1^i,{\mathrm{\Phi }}^i,{\varphi }_2^i$) of each phase (i = α or θ) measured by the OIM Data Collection were used. Euler angles are geometrically defined as three rotation angles that conform a sample coordinate system to an objective crystal coordinate system at each measurement point; the OIM Data Collection features Euler angles according to the Bunge convention. Therefore, the rotation matrix R calculated from the Euler angles is expressed as   

$$\begin{array}{l}{\mathit{R}}_i={\mathit{R}}_{{\varphi }_1^i}\cdot {\mathit{R}}_{{\mathrm{\Phi }}^i}\cdot {\mathit{R}}_{{\varphi }_2^i}=\\ \left(\begin{array}{ccc}cos{\varphi }_1^{i}cos{\varphi }_2^i-sin{\varphi }_1^{i}sin{\varphi }_2^{i}cos{\mathrm{\Phi }}^i& sin{\varphi }_1^{i}cos{\varphi }_2^i+cos{\varphi }_1^{i}sin{\varphi }_2^{i}cos{\mathrm{\Phi }}^i& sin{\varphi }_2^{i}sin{\mathrm{\Phi }}^i\\ -cos{\varphi }_{1}sin{\varphi }_2^i-sin{\varphi }_1^{i}cos{\varphi }_2^{i}cos{\mathrm{\Phi }}^i& -sin{\varphi }_1^{i}sin{\varphi }_2^i+cos{\varphi }_1^{i}cos{\varphi }_2^{i}cos{\mathrm{\Phi }}^i& cos{\varphi }_2^{i}sin{\mathrm{\Phi }}^i\\ sin{\varphi }_1^{i}sin{\mathrm{\Phi }}^i& -cos{\varphi }_1^{i}sin{\mathrm{\Phi }}^i& cos{\mathrm{\Phi }}^i\end{array}\right).\end{array}$$(1)

Because α-ferrite is cubic and θ-cementite is orthorhombic, they have 24 and 8 rotational symmetric bodies, respectively. To analyze the ORs, in this study, the α crystallographic orientation is displayed based on the θ crystal coordinate system owing to the lower crystal symmetry of θ. As shown in Fig. 2, the Euler angles projecting the α crystallographic orientation onto the θ reference coordinate system ${({\varphi }_1^{\alpha },{\mathrm{\Phi }}^{\alpha },{\varphi }_2^{\alpha })}_{\theta \to \alpha }$ were obtained by the product of the matrices, R_α·${\mathit{R}}_{\theta }^{-1}$, where R_α and R_θ are the rotation matrices for α and θ, respectively, with respect to the sample coordinate system. This operation was applied to a data set consisting of two left/right adjoining points belonging the α and θ phases in the EBSD data. By applying the above operation to all data sets (over 45000 in this study), the distribution of the α orientation at the α/θ interface was displayed as a (001)_α pole figure on the θ standard stereographic projection. However, it should be noted here that the above analysis could not identify the crystallographic equivalent orientations, i.e., the variants. Because each band in an EBSD pattern reflected from crystallographically equivalent planes is randomly indexed, the Euler angles of α and θ could not be determined uniquely.

Fig. 2.

Schematic illustration of EBSD data points showing an interphase between the ferrite and cementite phases. The orientations of the neighboring ferrite and cementite points are indicated by each coordinate system together with the sample coordinate system.

The ideal α crystallographic orientation satisfying the Bagaryatsky, Isaichev, Pitsch–Petch, and Unknown ORs with θ were theoretically calculated using the transformation matrix for crystal lattice T_θ→α.   

$${\mathit{T}}_{\theta \to \alpha }\cdot {\mathit{I}}_{\theta }={\mathit{X}}_{\alpha }$$(2)

where I_θ and X_α are the reference orientation of θ and the corresponding orientation of α after the transformation, respectively. T_θ→α was calculated using the parallel-plane and direction relations in each OR.¹⁰⁾ The ideal Euler angles ${({\varphi }_1^{\alpha *},{\mathrm{\Phi }}^{\alpha *},{\varphi }_2^{\alpha *})}_{\theta \to \alpha }$ for α satisfying each OR in the θ reference coordinate system were obtained with T_θ→α and Eq. (1) while taking into account that 2, 4, 8, and 16 equivalent variants exist in the Bagaryatsky, Isaichev, Pitsch–Petch, and Unknown ORs, respectively. Comparing an experimental ${({\varphi }_1^{\alpha },{\mathrm{\Phi }}^{\alpha },{\varphi }_2^{\alpha })}_{\theta \to \alpha }$ with the 30 theoretical types of ${({\varphi }_1^{\alpha *},{\mathrm{\Phi }}^{\alpha *},{\varphi }_2^{\alpha *})}_{\theta \to \alpha }$ according to the misorientation angle between them, Θ, the OR having the lowest Θ was determined to be the optimal one. In addition, the angular deviation for the parallel-plane and direction relations under the OR were separately evaluated as Θ_p and Θ_d, respectively. Because spheroidized pearlite loses optical information on the lamellar alignment of pearlite, the determination of OR was carried out using the misorientation in this study, as mentioned above. However, it is more appropriate for OR identification to investigate the orientation of the lamellar interface by using transmission electron microscopy⁷⁾ because Bagaryatsky and Pitsch–Petch ORs are very similar to Isaichev and Unknown ORs, respectively, with a few degrees of misorientation.

3. Results and Discussion

3.1. Coexistence of Multiple α/θ Orientation Relationships in Pearlite

3.1.1. Microstructure Change of Pearlite by Spheroidized Annealing

Figure 3 shows the microstructure of the eutectoid steel isothermally held at 923 K for 180 s. The black and white regions correspond to pearlite and martensite, respectively, indicating the isolated growth of pearlite nodules within the austenite matrix. The pearlite nodules containing some colonies with different lamellar alignment directions grew isotopically. It was confirmed that the nodules grew at a constant rate⁷⁾ and eventually became pearlite blocks approximately 40 μm in diameter upon completion of the pearlitic transformation.

Fig. 3.

Optical image showing a growing pearlite nodule in 0.83%C–1.46%Mn steel isothermally held at 923 K for 180 s. (Online version in color.)

OM and SEM images and IPF maps of ferrite in lamellar and spheroidized pearlite are shown Figs. 4 and 5, respectively, which were isothermally held at 923 K for 0.6 ks and 360 ks, respectively. In the IPF map, low- and high-angle grain boundaries with misorientations of 2°–15° and 15° and higher are indicated by white and black lines, respectively. In the material held for 0.6 ks (Fig. 4), pearlitic transformation was completed, and a fine lamellar structure consisting of ferrite (dark phase) and cementite (white phase) was observed throughout [Figs. 4(a), 4(b)]. The lamellar pearlite had the crystallographic characteristics of pearlite, as previously reported,^{1,2,3,4,5,6,7,8)} where the orientation of the pearlitic ferrite changed continuously and low- and high-angle grain boundaries, which seemed to correspond to colony and block boundaries, respectively, were densely developed [Fig. 4(c)]. After isothermal holding for 360 ks, the cementite lamellae were sufficiently spheroidized [Figs. 5(a), 5(b)], but the substructures developed in the lamellar pearlite were still observed [Fig. 5(c)]. From these results, it was confirmed that the spheroidized annealing did not change the crystallographic features of the pearlite but rather the morphology of the distributed cementite. Therefore, spheroidized pearlite was used for the subsequent detailed analysis.

Fig. 4.

The microstructure of lamellar pearlite in 0.83%C–1.46%Mn steel isothermally held at 923 K for 0.6 ks. (a) OM and (b) SEM images and (c) ferrite IPF map. (Online version in color.)

Fig. 5.

Microstructure of spheroidized pearlite in 0.83%C–1.46%Mn steel isothermally held at 923 K for 360 ks. (a) OM and (b) SEM images and (c) ferrite IPF map. (Online version in color.)

3.1.2. Identification of Crystallographic Orientation of Cementite in Spheroidized Pearlite

Figure 6 shows IPF and phase maps of spheroidized pearlite. To analyze a wide area with higher resolution, 10 adjacent pictures taken at high magnification were combined as an image. As mentioned for Fig. 5, the crystal orientation of the pearlitic ferrite rotated continuously, and there were several blocks surrounded by high-angle grain boundaries in the observation area. However, since the rotation axis of the continuous crystal ration of pearlitic ferrite tends to be different from colony to colony,^8,11) the misorientaiton angle of colony boundaries sometimes becomes to be higher than 15°. Therefore, there are high-angle boundaries with a certain density even in a block. In addition, spheroidized cementite particles were distributed in the blocks relatively uniformly, and the area fraction of the spheroidized cementite was measured to be 0.082. According to thermodynamic calculations, the equilibrium phase fraction of cementite at 923 K is approximately 0.12, and thus cementite in pearlite could be identified at a probability of approximately 70% after spheroidization. Each cementite particle keeps an identical orientation and the misorientaiton deviation was lower than 2°. We confirmed that lamellar cementite (Fig. 4) was difficult to identify, even using EBSD. The orientation distribution of ferrite and cementite in area A, shown as the black rectangle, is shown in Figs. 7(a) and 7(b), respectively, with (001) pole figures. The dotted circles in Fig. 7(a) show the orientation deviation by 10° from the average orientation of the measured ferrite. Although the ferrite orientation was scattered with relatively high orientation deviation owing to the orientation rotation of pearlite, the results indicated an identical orientation. In contrast, there were two distinctly different (001)_θ orientations in cementite [Fig. 7(b)]. Because the crystal structure of cementite is orthorhombic, these two (001)_θ orientations clearly revealed the presence of cementite with different orientations in the area under investigation, and it implies that multiple ORs coexist within a single pearlite block.

Fig. 6.

(a) IPF map and (b) phase map of spheroidized pearlite isothermally transformed at 923 K for 360 ks. In (b), high- and low-angle grain boundaries are indicated by black and white lines, respectively. (Online version in color.)

Fig. 7.

(001) pole figures showing orientations of (a) ferrite and (b) cementite. Analyzed region corresponds to the black-framed region A in Fig. 6.

3.2. Relation between Hierarchical Substructures and α/θ Orientation Relationships in Pearlite

3.2.1. Comprehensive Analysis for α/θ Orientation Relationship

Figure 8 shows the orientation intensity distribution of ferrite (001)_α on the cementite standard stereographic projection for all α/θ interfaces shown in Fig. 6. The theoretical (001)_α orientation of ferrite with the Bagaryatsky, Isaichev, Pitsch–Petch, and Unkown ORs are indicated by colored circles, and the variant number of each OR is indicated, for example, B1 for variant 1 of the Bagaryatsky OR. Comparing the four theoretical ORs, it was found that ferrite orientations under the Bagaryatsky and Pitsch–Petch ORs were very close to those of the Isaichev and Unknown ORs, respectively, although they had different parallel-plane relations. The pole figure clearly reveals that the ferrite was highly oriented, satisfying the Bagaryatsky, Isaichev, and Pitsch–Petch ORs, whereas it was hardly oriented to the Unknown OR of U2, 4, 7, 8, 10, 12, 15, and 16. That is, it was discovered that the Bagaryatsky, Isaichev, and Pitsch–Petch ORs clearly coexisted, even within a selected area of pearlite, but the Unknown OR did not. Incidentally, the α/θ interface boundaries satisfying P 1, 2, 7, and 8 in Pitsch-Petch OR were hardly detected, probably because they did not exist within the selected area. Additionally, the pole figure implies the presence of another OR, as indicated by the white arrows, although further research is needed to identify the crystal orientations taking lamellar interface orientation into account.

Fig. 8.

Cementite standard stereographic projections showing the distribution of (001)_α of ferrite facing cementite in spheroidized pearlite. Theoretical (001)_α under Bagaryatsky, Isaichev, Pitsch–Petch, and Unknown ORs are plotted as colored circles. (Online version in color.)

3.2.2. Crystal Orientation Deviation of Three Orientation Relationships

Figure 9 shows the relationship between the deviations of misorientations Θ_p and Θ_d, which correspond to the angle deviations for parallel-plane and parallel-direction relations, respectively, for the Bagaryatsky, Isaichev, and Pitsch–Petch ORs under tolerance angle Θ^* = 10°. The Θ_d of the Pitsch–Petch OR in particular, [35 10 11]_α, which was tilted from [311]_α by 2.6°, was used. Θ_d tended to increase with increasing Θ_p under all ORs, which indicates that the orientation deviation could not be characterized as in-plane rotation on the lamellar interface. However, the degree of the orientation deviation was different among the three ORs. In the Bagaryatsky and Isaichev ORs [Figs. 9(a) and 9(b), respectively], the data were concentrated from (Θ_p, Θ_d) = (0, 0) within 2°, indicating a relatively small deviation from the theoretical OR. In contrast, for the Pitsch–Petch OR [Fig. 9(c)], the data were distributed more widely, with (Θ_p, Θ_d) = (0, 3) as the most frequent value. These results suggest that the orientation relations of the Pitsch–Petch OR were not as strict as those of the other two ORs. Although the reason is not clear, it may be because of the higher indexes of the parallel-plane and direction relations of the Pitsch–Petch OR.

Fig. 9.

Relationship between the deviations of plane-normal Θ_p and direction-normal Θ_d in the (a) Bagaryatsky, (b) Isaichev, and (c) Pitsch–Petch ORs. The tolerance of misorientation is 10°.

3.2.3. Distribution of Different α/θ ORs

A map of the α/θ phase interface boundaries classified into the three ORs is shown in Fig. 10. The data correspond to those used for Fig. 6, and the tolerance angle for classification Θ^* was set to 10°. The interface boundaries satisfying the Bagaryatsky, Isaichev, and Pitsch–Petch ORs are colored red, green, and blue, respectively. The fraction of the length for each OR is summarized in Table 1. As previously mentioned, the three ORs coexisted in this area, and the fraction was from high to low in the order of Pitsch–Petch, Isaichev, and Bagaryatsky ORs. It should be noted here that a certain fraction of the Bagaryatsky OR was present, although it is, in principle, difficult to distinguish between the Bagaryatsky OR and the Isaichev OR owing to their very similar orientations and few variants. Indeed, Zhou and Shiflet⁶⁾ reported the presence of both Bagaryatsky and Isaichev ORs in pearlite using transmission electron microscopy. The current experimental results suggest the coexistence of multiple ORs in pearlite, regardless of the bulk carbon composition, and that the OR may change during pearlitic transformation. Looking at the distribution of the α/θ interface satisfying respective ORs, it was found to be uneven, and the transition of ORs tended to correspond to the colony boundaries with low-angle misorientation. For instance, the transition between (1) Bagaryatsky and Isaichev ORs and between (2) Isaichev and Pitsch–Petch ORs are indicated by arrows I and II, respectively. The former transition was accompanied by the crystal rotation of cementite with low-angle misorientation, whereas the latter was accompanied by crystal rotation with high-angle misorientation. Thus, these are referred to as Type 1 and Type 2 colony boundaries, respectively.

Fig. 10.

α/θ phase interface boundary map classified into Bagaryatsky, Isaichev, and Pitsch–Petch orientation relationships. (Online version in color.)

Table 1. Fraction of α/θ interface belonging to the three ORs in spheroidized pearlite. The tolerance in each orientation relationship was set to 10°.

Orientation Relationships	Fraction (%) (Datasets) Within 10°
Bagaryatsky	4.6 (1264)
Isaichev	27.3 (7530)
Pitsch–Petch	68.1 (18791)

3.3. Formation Mechanism of Pearlite Colony with Multiple α/θ ORs

The reason for the variation in lamellar alignment for pearlite growth, i.e., the formation of pearlite colonies, is still under debate. Hull and Mehl¹²⁾ suggested that pearlite nodules grow by the repeated nucleation and growth of cementite, assuming that the variation of lamellar alignment in a nodule is caused by the nucleation of cementite lamellae with individual lamellar directions. However, according to OM images showing the three-dimensional morphology of cementite by the serial sectioning technique, Hillert²⁾ pointed out that ferrite and cementite were bound together with the same OR at the boundary of two adjacent colonies and proposed the formation mechanism of colonies by cementite branching. The branching mechanism seems to be accepted so far, but it cannot fully explain the variation in the lamellar alignment. In general, a lamellar interface between pearlitic ferrite and cementite corresponds to the parallel plane of each OR.⁶⁾ Thus, it is expected that the lamellar alignment does not change randomly when ferrite and cementite maintain a unique crystallographic orientation satisfying a certain OR. Zhou and Shiflet⁴⁾ considered the macroscopic curvature of the α/θ lamellar interface while taking into account the introduction of misfit dislocations and structural ledges at the lamellar interface. However, such curvature of the lamellar interface does not lead to discontinuous changes in lamellar alignment, such as colony boundaries, but continuous ones. Furuhara et al.¹³⁾ reported that very fine cementite particles with different variants were distributed within a single block in degenerate pearlite formed in an Fe–0.38%C binary alloy. Furthermore, they recently investigated the interface boundary precipitation of vanadium carbide particles in an Fe–0.10%C–0.43%V–1.49%Mn–0.05%S alloy and found that the VC variant changed discontinuously at the colony boundary introduced during the growth of nodular ferrite.¹⁴⁾ Given that the degenerate pearlite and the interface boundary precipitation are regarded as eutectoid transformation products with noncooperative growth of ferrite and carbide, cementite lamellae with different variants may form in lamellar pearlitic transformation. In fact, different cementite variants were confirmed (see Fig. 11). Therefore, there were three types of colony boundaries in the pearlite as follows:

Type 0: cementite lamellae with two different variants that maintained an OR.

Type 1: cementite lamellae with almost the same orientation and different ORs (e.g., transition between Bagaryatsky and Isaichev ORs)

Type 2: cementite lamellae different not only in crystal orientation but also in ORs (e.g., transition between Isaichev and Pitsch–Petch ORs)

Fig. 11.

(001) pole figure of two cementite particles and surrounding ferrite matrix. The cementite particles are indicated by arrows I and II in Fig. 10 and have a Pitsch–Petch OR with the ferrite matrix.

As schematically illustrated in Fig. 12, these colony boundaries enable discontinuous direction changes of the lamellar interface. The Type 1 colony boundary may have been the colony boundary that was incorrectly considered to have formed by the blanching mechanism by Hillert.²⁾ Thus, it can be concluded that pearlitic transformation undergoes repeated nucleation of cementite at the colony boundaries, leading to the variation of lamellar alignment. The formation mechanism of pearlite colonies by multiple ORs indicates a relatively high degree of freedom in the selection of the α/θ OR. In other words, the most important issue for pearlitic transformation is the production of a lamellar structure in various alignment directions, which might originate from the development of anisotropic internal stresses generated by α/θ misfit.¹⁵⁾ In addition, the difference in OR might affect the kinetics of cementite spheroidization owing to the difference in interfacial and/or elastic strain energies, which results in the heterogeneous spheroidization in pearlite.

Fig. 12.

Schematic illustration of the formation mechanism of the pearlite colony. The alignment of the lamellar interface is discontinuously changed by the colony boundaries of Types 0, 1, and 2. (Online version in color.)

4. Conclusions

To investigate the orientation relationship (OR) between ferrite and cementite in pearlite over a wide area and with high precision, EBSD analysis was applied to spheroidized pearlite in Fe–0.83C–1.46Mn (in wt.%). The formation mechanism of pearlite colonies was discussed based on the analysis results. The results obtained are as follows:

(1) Three different ORs were simultaneously established in pearlite: Bagaryatsky, Isaichev, and Pitsch–Petch ORs. However, the OR (011)_θ // (215)_α, [011]_θ 2.6° from [311]_α, and [100]_θ 2.6° from [131] _α was not observed.

(2) The Pitsch–Petch OR had a larger angular deviation from the theoretical orientation than the other two ORs.

(3) The cementite satisfying each OR with the ferrite matrix was unevenly distributed, and the distribution tended to correspond to a pearlite colony. That is, the lamellar cementite discontinuously changed not only the variant but also the OR at the colony boundaries during pearlitic transformation, which led to changes in the α/θ lamellar interface and lamellar structure in various directions.

References

• 1)   T.  Takahashi,  M.  Nagumo and  Y.  Asano: J. Jpn. Inst. Met., 42 (1978), 716 (in Japanese).
• 2)   M.  Hillert: Decomposition of Austenite by Diffusional Processes, ed. by V. F. Zackay and H. I. Aaronson, Interscience, New York, (1962), 197.
• 3)   M. A.  Mangan and  G. J.  Shiflet: Metall. Mater. Trans. A, 30 (1999), 2767.
• 4)   D. S.  Zhou and  G. J.  Shiflet: Metall. Trans. A, 22 (1991), 1349.
• 5)   M.-X.  Zhang and  P. M.  Kelly: Mater. Charact., 60 (2009), 545.
• 6)   D. S.  Zhou and  G. J.  Shiflet: Metall. Trans. A, 23 (1992), 1259.
• 7)   Y.  Amemiya,  N.  Nakada,  S.  Morooka,  M.  Kosaka and  M.  Kato: Tetsu-to-Hagané, 105 (2019), 314 (in Japanese).
• 8)   N.  Nakada,  N.  Koga,  T.  Tsuchiyama and  S.  Takaki: Scr. Mater., 61 (2009), 133.
• 9)   Y.  Wang,  Y.  Tomota,  T.  Ohmura,  W.  Gong,  S.  Harjo and  M.  Tanaka: Acta Mater., 196 (2020), 565.
• 10)   H.  Kitahara,  R.  Ueji,  N.  Tsuji and  Y.  Minamino: Acta Mater., 54 (2006), 1279.
• 11)   N.  Koga,  N.  Nakada,  T.  Tsuchiyama,  S.  Takaki,  M.  Ojima and  Y.  Adachi: Scr. Mater., 67 (2012), 400.
• 12)   F. C.  Hull and  R. F.  Mehl: Trans. ASM, 30 (1942), 381.
• 13)   T.  Furuhara,  T.  Moritani,  K.  Sakamoto and  T.  Maki: Mater. Sci. Forum, 539–543 (2007), 4832.
• 14)   Z.  Wang,  Y.  Zhang,  G.  Miyamoto and  T.  Furuhara: Scr. Mater., 198 (2021), 113823.
• 15)   N.  Nakada and  M.  Kato: ISIJ Int., 56 (2016), 1866.