2.1 Precipitation in ferritic matrix a. The interphase precipitation
The interface is known to be a topic of interest in the field of solid transformation [21-26]. The term “interphase precipitation” represents the carbide nucleation process repeats periodically during the austenite-to-ferrite transformation, resulting in a sheeted structure[2, 27]. Such carbide dispersion and morphology have been reported to depend on the transformation temperature. One of the proposed classifications is shown in Figure 2-1.
5
Figure 2-1 The variations of carbide morphology with transformation temperatures [28]
The microstructure of the interphase-precipitated carbide was studied by Honeycombe et al. via TEM in different alloyed steels [27, 29], and the ledge mechanism of austenite-to-ferrite transformation is proposed to be operated during transformation, as shown in Figure 2-2. The interphase-precipitated carbide obeys one variant of Baker-Nutting orientation relationships, {0 0 1}carbide || {0 0 1}α and <1 1 0>carbide || <1 0 0>α. The single variant selection of interphase-precipitated carbides is explained to minimize the interfacial energy and to maximize the diffusion efficiency of solute atoms.
Figure 2-2 Schematic diagram showing the interphase-precipitated carbides with planar spacing
6
The process of interphase precipitation (with planar spacing) is briefly described in Figure 2-3. During γ to α transformation, the formation of α increases the carbon content in front of the ferrite/austenite interface. As the carbon concentration reaches a critical value, alloy carbides would start to precipitate on the ferrite/austenite interface.
The formation of carbide lowers the carbon content ahead of interface, increasing the driving force for austenite-to-ferrite transformation. The interface then advances and the above process repeat periodically, leading to dense of carbides arranged in regular spacing.
Figure 2-3 The carbon concentration profile ahead of ferrite/austenite interface during interphase precipitation process [30]
The ledge mechanism of austenite-to-ferrite transformation is viewed as the spirit of interphase precipitation [31-34]. A typical ledge structure of interface contains a broad plane and a mobile step, which are illustrated in Figure 2-4. The overall growth rate of a ledged interface, V, depends on the ledge height, h, and its spacing, b:
s
V V h
= b (2-1)
7
where Vs is the step velocity, h is the ledge height, and b is the spacing of ledges.
Figure 2-4. Schematic illustration showing a unit ledge accompanying with features of interests. The meanings of each term are referred to the text.
Based on the point of view of Honeycombe et al., the broad interface is supposed to be semi-coherent. Ferrite follows {110}α || {111}γ Kurdjumov-Sachs orientation relationships with respect to the parent austenite. On the contrary, the mobile steps are incoherent. Based on the metallurgical considerations, these steps are assumed to be the preferred carbide nucleation sites. However, in the case of interphase precipitation, these steps move too fast during transformation, making the carbide nucleus difficult to nucleate on them. Instead, carbides prefer to precipitate on coherent/semi-coherent interfaces, leading to interphase precipitation.
In addition to the planar interphase precipitation, other mechanisms had been proposed to explain the interphase precipitation with curved spacing. Ricks and Howell proposed the bowing mechanisms (for curved interphase precipitation with irregular spacing) and quasi-ledge mechanisms (for curved interphase precipitation for regular spacing) to illustrate the curved interphase precipitation [35, 36]. The bowing mechanism, as shown in Figure 2-5, requires the carbides to pin the moving interface. It
8
can be expected that the carbide density and the particle spacing in rows determine the resulting sheet spacing. For quasi-ledge mechanism (Figure 2-6), the interface breaks into austenite at the place where the particle spacing is relatively wider. The segment of the advancing interface accumulates sufficient carbon and solute, and then carbide starts to precipitate on its top. The newly formed carbide pins the advancing interface again, the mobile steps then move by sideway. Obviously, the carbide distribution determines which mechanism, bowing and quasi-ledge, is operated during transformation.
Figure 2-5 The bowing mechanism for the curved interphase-precipitated carbides with irregular spacing [35]
Figure 2-6 Schematic illustration of quasi-ledge mechanism of curved interphase precipitated carbides with regular spacing [35]
9
b. The carbide fiber growth
In addition to interphase precipitation, another carbide aggregate: carbide fiber, is usually found in Fe-Mo-C, Fe-Cr-C, and Fe-V-C steels as well, but the strengthening mechanism of fibrous carbide in ferrite is less discussed [37]. The observed features and proposed formation conditions for fibrous carbide are summarized in Table 2-1. Carbide fibers are known to be straight and branchless, generally growing in a single direction normal to the transformation interface. The occurrence of fibrous carbide depends on transformation temperature, chemical composition, and the nature of interfaces. The related literatures have been reviewed and the important concepts and ideas are summarized as follows:
(1) The carbide fiber in the Fe-V-C Steel
Generally, the addition of vanadium is widely used to strengthen steels by interphase precipitation hardening due to the temperature for precipitation is approximately close to austenite to ferrite transformation [29, 38]. The occurrence of alloy fiber in Fe-V-C steels is not greatly reported. Edmonds studied the VC fibers by varying cooling rate and by changing the content of Mn [3]. He found that as the amount of Mn was elevated from zero to 1.6 wt%, the density of VC fibers was increased of 50%.
The Mn has been reported to have strong effects on slowing down the advancement of the transformation interface [39]. Edmond’s works indicate that the interface mobility is an important role in carbide morphology determination. He suggested that a higher transformation temperature would be favored for the development of fibrous carbide and pointed out that the selection of habit plane is a
10
function of the transformation temperature. For interphase-precipitated carbide, its habit plane is expected to be as parallel to the interface because that minimizes the interfacial energy and maximizes the diffusion of solute atoms [28]. However, the habit plane of VC does not to be as parallel to the transformation interface at the high transformation temperature because the solute now requires sufficient energy [28]; the formation of fibrous carbide becomes possible. These results indicate that the change in carbide morphology requires changing the transformation temperature.
VC carbide is f.c.c structure. The Baker-Nutting orientation relationships [40, 41]
are commonly found for these carbides with respect to ferrite:
(
1 0 0) (
VC || 1 0 0)
α[
0 0 1] [
VC || 0 1 1]
αThe structure of VC fiber is identical as that of interphase precipitated VC carbide.
Berry had claimed that the VC fibers exhibited Kurdjumov-Sachs orientation relationships with respect to ferrite [42]:
(
1 1 1) (
VC || 1 1 0)
α1 1 0 || 1 1 1
VC α
The difference in orientation relationships implies the nucleation and growth of carbides follow different process. If the VC fibers follow Kurdjumov-Sachs orientation relationships with respect to ferrite, the nuclei of fibers are formed at the side of
11
austenite at the austenite/ferrite interface, then growing in austenite, instead of ferrite.
Otherwise, if Baker-Nutting orientation relationships are followed by ferrite and carbide, the development of carbides is related to ferrite, not austenite. The orientation relationships of carbide with respect to austenite reveal important information of where nucleation and growth of carbide occur. The detailed results and discussions will be presented in the next chapter
12
Reference Basis of proposition Features for carbide fiber The clarified features
Transition in morphology without changing matrix orientation
Plane normal to the interface (incoherent) becomes dominant for precipitation by decreasing temperature
Curved interface is required
The carbide aggregates are controlled by changing transformation temperatures.
VC fiber still follows Baker-Nutting orientation relationships with respect to ferrite
Coherency of interface would not be the factor controlling carbide morphology. originate from prior austenite grain boundary
The transition of carbide morphology does not have to be associated with changing transformation temperature Bee and
Campbell [44, 45]
Carbide fiber is in the side of ferrite without Kurdjurmov-Sachs orientation relationship and interphase precipitation in the other side.
Carbide fiber originates from prior austenite grain boundary
Berry and Honeycombe [42]
Higher transformation is favored for carbide fiber
VC fiber and ferrite matrix follows Kurdjurmov-Sachs orientation relationships
Edmonds [3]
The ledge mechanism was not emphasized
Higher transformation temperature is favored because the habit plane does not have to be parallel to the interface.
Higher addition of Mn promotes the formation of carbide fiber.
Carbide fiber can develop inside the ferrite grain.
Change in carbide morphology has to be associated with changing transformation temperatures
Table 2-1 The observed features and proposed mechanisms for the development of fibrous carbide
13
(2) The carbide fiber in the Fe-Cr-C steel
The carbide fibers in isothermally transformed Fe-Cr-C steels were studied by Bee [44] and Campbell [45] et al. The interphase-precipitated carbides and the fibrous carbides were both observed among these works. From their TEM evidence, they pointed out that interphase precipitation was related to a stepped interface, but the carbide fiber was in associated with curved interfaces (Figure 2-7). They then concluded that the formation of the interphase-precipitated carbide or the fibrous carbide would be controlled by the coherency of interface.
Figure 2-7 Fe-5Cr-0.2C (wt%) isothermal transformation at 650 oC for 30 min showing the interfaces associated with interphase precipitation (Right-hand side) and alloy fibers (Lef-hand side) [45]
Furthermore, two kinds of carbide fiber were found (M7C3 or M23C6). The examinations showed that the chemistry of carbide depends on the transformation temperature only. As transformation temperature is lowered, the spacing of fibers becomes finer, and the chemistry of carbide changes from M23C6 to M7C3.
14
The crystal structures of M23C6 and M7C3 are f.c.c. and h.c.p., respectively. For M23C6, the orientation relationships of these fibers with respect to ferrite were identified as:
which are known as Kurdjumov-Sachs orientation relationships. For M7C3, which the structure is h.c.p., the orientation relationships with ferrite had been identified as follows:
Bee and Campbell considered the carbide morphology was related to crystallographic of ferrite/austenite interface. As austenite starts to decompose to ferrite, ferrite initially nucleates on prior austenite grain boundaries. In the growing stage, this ferrite nucleus would not possess any orientation relationships with respect to the austenite into which it is growing (for the case of allotriomorphic ferrite). The other side where growth of ferrite does not occur, the ferrite and austenite exhibit Kurdjumov-Sachs orientation relationships. Thus, it is expected that the carbide fibers would form predominantly at the side of austenite where the ferrite is growing into, and interphase precipitation would occur at the other side of ferrite/austenite interface that
15
follows Kurdjumov-Sachs orientation relationships. This argument is consistent with that proposed by Honeycombe et al. It implies that austenite/ferrite interfacial structures could be another factor on determining carbide morphologies but the above suggestions cannot explain the transition in carbide morphologies.
(3) The carbide fiber in the Fe-Mo-C steel
The carbide fibers formed in Fe-Mo-C steels were widely studied by Edmonds and Honeycombe et al [8, 46]. Three morphologies, interphase precipitation, Widmanstätten arrays, and carbide fibers were found in this alloyed system, as demonstrated in Figure 2-8.
The carbide in Fe-Mo-C steels is mainly Mo2C with h.c.p structure. The analysis on the diffraction patterns showed that there are two orientation relationships of carbides with respect to ferrite. The orientation relationships of Mo2C with respect to ferrite are:
However, for the carbides that arranged in Widmanstätten way and interphase-precipitated carbide, the orientation relationships become to be:
( ) ( )
Such orientation relationships were also confirmed by other studies.
Figure 2-8 Mo2C carbides with different morphologies in Fe-3.5Mo-0.22C (wt%) steels after isothermal transformation at 750 oC for 30 min. (a) interphase precipitation, (b) Widmanstätten arrays, and (c) Mo2C fibers [46]
c. The transition in carbide morphology
The change in carbide morphologies is an important issue in alloyed steels because it directly affects performed mechanical properties. The transition in carbide morphologies was investigated by Barbacki and Honeycombe in Mo and V containing steels by simply altering the transformation temperatures [43]. These fibers mainly originated from curved interfaces, and with different spacing as the transformation temperature was changed (Figure 2-9). Even though the carbide morphology and its distribution are controlled by the transformation temperature, the transition in carbide morphology should not be associated with altering the orientation of ferrite matrix.
Compared to Edmonds’ proposal, they thought that the carbide fibers were predominant at lower transformation temperatures. There explanation was based on both transformation kinetics of Mo containing steels and misfits that result from different crystal structures of carbide/ferrite and carbide/austenite. By decreasing the transformation temperature, the solute atoms would mainly diffuse along the incoherent
17
interface. For the ledge mechanism of austenite-to-ferrite transformation, the terrace plane was viewed as a semi-coherent interface for precipitation. Consequently, the carbide would like to precipitate on the mobile steps as the transformation temperature is lowered, and then grow in a direction perpendicular to the transformation front.
However, this feature is not seen in other systems. For Fe-Cr-C and Fe-V-C steels, the fibrous carbides were found to be majority at higher transformation temperatures, instead of lower transformation temperatures.
Figure 2-9 Fe-4Mo-0.2C (wt%) steels showing the spacing and fineness of fibers changed as transformation temperature was varied from 850 oC (left-hand side) to 750 oC (right-hand side) [43]