Experimental and Modelling Research on the Effect of Prior Ferrite on Bainitic Transformation in Medium-Carbon Bainitic Steel

Abstract

In this study, we investigate the impact of prior ferrite on the bainite transformation kinetics and microstructure of medium-carbon steel interrupted by an intercritical annealing (IAA) process. It was found that the incubation time and completion time decreased from 687 s and 6018 s to 20 s and 4680 s, with the volume fraction of ferrite increasing from 9.5% to 28.6%, while the maximum transformation rate increased from 00271 μm/s to 0.0436 μm/s. The ferrite/austenite interface is introduced, and the nucleation sites are increased to accelerate the subsequent bainite transformation due to the formation of prior ferrite. However, there is a competitive relationship between the number and activation energy of bainite nucleation. According to the experimental results and theoretical calculations, the activation energy of the bainite transformation in the medium-carbon bainite steel decreases gradually with an increase in the volume fraction of prior ferrite.

1. Introduction

Carbide-free bainitic steel has been the subject of considerable attention in recent years because of its effective combination of strength, ductility, and toughness [1,2,3,4,5]. The microstructure of carbide-free bainitic steel typically consists of nano-sized bainitic ferrite plates and austenite films [6,7]. The thickness of bainitic ferrite plates is influenced by chemical composition and transformation temperature [8,9,10]. Generally, as the transformation temperature decreases from the bainite start temperature (Bs) to 10 °C above the martensite start temperature (Ms), or the carbon content of bainitic steel increases from low carbon (0.2–0.4 wt.%) to high carbon (0.7–0.9 wt.%), the bainitic plates become finer [6,7,9,10]. Since the grain refinement of bainitic ferrite and the transformation-induced plasticity of the retained austenite contribute to enhanced strength and toughness, bainitic steel exhibits superior mechanical properties [11,12].

At first, the lower bainite transformation temperature for high-carbon bainitic steel leads to a decrease in transformation kinetics [11,13]. Compared to martensitic transformation, carbide-free bainite transformation takes a longer time, particularly in high-carbon, high-silicon, carbide-free steels, where this process can extend over hundreds of hours, which is detrimental to industrial production and, thus, limits its application [11,14]. Subsequently, researchers have explored several strategies to accelerate the kinetics of low-temperature bainite formation, such as reducing the carbon content of steel [15], optimizing the size of prior austenite grain [16], Nb/B microalloying [17,18,19], decreasing the austempering temperature [20,21], ausforming [22], and forming bainite/martensite before bainite transformation [23,24]. Considering the long bainite transformation time and poor weldability of high-carbon steel, the current research on carbide-free bainitic steel focuses on medium-carbon composition. Ausforming at low temperatures has been identified as a condition conducive to bainite nucleation, thereby shortening the incubation time of bainite formation [22,25]. This process also inhibits the growth of bainitic ferrite, which, in turn, slows down the overall transformation rate [26]. The presence of prior martensite in carbide-free bainitic steel can increase the nucleation sites by introducing an additional α/γ interface, accelerating the subsequent bainitic transformation [20,27,28]. Nonetheless, martensite formation can simultaneously increase the carbon content and improve the stability of untransformed austenite, potentially prolonging the transformation finishing time. However, the process of accelerating carbide-free bainitic steel using prior bainite/martensite requires high-precision transformation temperature control, and the industrial production window is narrow. Several studies have shown that one effective strategy to enhance the bainite transformation rate is to introduce a small fraction of prior ferrite before the bainite formation [29,30,31]. However, the mechanism of bainite kinetic acceleration is disputed. K. Y. Zhu et al. [29] proposed that the acceleration of bainitic and martensitic transformation can be attributed to the nucleation site at the α/γ interface and the related concentration gradients of the C/Mn element. Ashwath M. Ravi et al. [32] researched the effect of the prior ferrite and austenite grain boundary on bainite transformation, and the results show that a small volume fraction of prior ferrite (<5%) can increase the rate of bainite transformation by increasing the number density of bainite nucleation sites. They also noted that the transformation rate depends on the bainitic nucleation rate when the volume fraction of prior ferrite is small. K. F. Li et al. [31] investigated the effects of the austenitization temperature (above or below Ac3) on bainite transformation. They suggested that the synergetic effect of the refinement of austenite grains and the percentage of intercritical ferrite can refine the bainite microstructure and decrease the fraction of martensite/retained austenite when the austenitizing temperature is below Ac3. Nevertheless, J. Lu et al. [33] studied the effect of heat treatment, including intercritical annealing, isothermal holding, and final quenching, on bainite transformation kinetics and microstructural evolution. They proposed that the carbon enrichment from prior ferrite or martensite to untransformed austenite decelerates the bainite transformation and prolongs the finishing time.

The existing research reveals that the prior ferrite process, i.e., the intercritical annealing process, is easier to realize and control in industrial manufacturing compared to bainite transformation acceleration methods. Consequently, this work employs isothermal treatment interrupted by the intercritical annealing (IAA) process to accelerate bainitic transformation in medium-carbon steels. By combining experimental analysis with the establishment of kinetic models, in this work, we examine the interactions between various factors affecting the kinetics of bainite transformation with different volume fractions of prior ferrite. This analysis elucidates the role of prior ferrite in the process of bainite transformation and its impact on the kinetics of bainitic transformation. This can provide theoretical guidance for the industrial production of medium-carbon carbide-free bainitic steel through a prior ferrite process.

2. Experimental Procedure

The steel under investigation was a medium-carbon, silicon-rich alloy with a chemical composition of 0.50 wt. %C, 2.56 wt. %Si, 1.07 wt. %Mn, 2.30 wt. %Cr, 0.0016 wt. %Nb, and 0.0052 wt. %B. Detailed descriptions of the steel preparation technology and the phase transformation temperatures (Ac₁, Ac₃, Ms) can be found in our previous studies [34,35]. Ac₁, Ac₃, and Ms were measured at 774 °C, 864 °C, and 187 °C, respectively.

The austempering experiments were meticulously carried out using a DIL 805A dilatometer. The heat treatment process was similar to Q&P steel treatment, referred to in this study as interruption by the intercritical annealing (IAA) process, as shown in Figure 1. The size of the cylindrical specimen was Ф4 mm × 10 mm. The specimen was rapidly heated in the two-phase zone, that is, at 810 °C, 830 °C, and 850 °C, for 10 min at a rate of 10 °C/s. It was then cooled to 300 °C with a cooling rate of 20 °C/s for enough time for austempering until the bainite transformation was complete, and, thus, cooled to room temperature. The changes in the length of the specimens during isothermal bainitic heat treatment were recorded to characterize the transformation kinetics. The specific heat treatment specimens are referred to as IAA−810, IAA−830, and IAA−850.

The metallographic samples were etched with 2% nital after grinding and polishing. The microstructure was meticulously examined using a ZEISS ULTRA 55 field-emission scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany) at 20 kV.

3. Results

3.1. Experimental Results

3.1.1. Dilatometric Analysis

The isothermal bainite transformation kinetics curves of the experimental steel with different intercritical annealing processes (IAA) are shown in Figure 2, and the specific values are shown in

Table 1. Transformation parameters of bainite transformation during the IAA process. (t_I is incubation time of bainite transformation, t_C is completion time of bainite transformation, and r_max is maximum bainite transformation rate).

Specimen	tI (s)	tC (s)	rmax (μm/s)
IAA−810	20	4680	0.0436
IAA−830	103	5695	0.0302
IAA−850	6048	6018	0.0271

. The volume fractions of formed bainite at 3% and 98% refer to the incubation time and the completion time of bainite transformation, respectively [36]. The transformation–time curves can be obtained by derivation of the dilatation–time curves of the specimens. The incubation time and completion time of bainite transformation for the experimental steel at 810 °C are 20 s and 4680 s. When the intercritical annealing temperature rises to 830 °C, the incubation time and completion time of bainite transformation are extended to 103 s and 5695 s, respectively. After the intercritical annealing temperature increases further to 850 °C, the incubation time and completion time of bainite transformation are 687 s and 6018 s, respectively. In addition, the times to reach the maximum bainite transformation rate for the IAA−810, IAA−830, and IAA−850 specimens are 474 s, 1177 s, and 1898 s, respectively. The maximum transformation rates are 0.0436 μm/s, 0.0302 μm/s, and 0.0271 μm/s, respectively.

As the intercritical annealing temperature increases, the incubation time and completion time also increase, as shown in Figure 3. These results demonstrate that when the intercritical annealing temperature increases, the volume fraction of prior ferrite in the intercritical zone of the experimental steel decreases. The volume fraction of prior ferrite was counted based on around five SEM photographs via Image Pro Plus 6.0 software. The microstructural morphology of the three specimens is shown and compared in Figure 4. The bainite transformation kinetics of the specimen austempered at 300 °C are reported in the literature [37]. When the volume fraction of prior ferrite in the intercritical zone increases from 0 to 9.6%, the incubation time (completion time) of the bainite transformation decreases from 1364 s (5.3 h) to 687 s (1.7 h). When the prior ferrite content of the experimental specimen is increased further to 28.6%, the incubation time and completion time of the specimen are only 20 s and 1.3 h, respectively. These results are similar to those in reference [38]. With an increase in prior ferrite content to 28.6% in the experimental steel, the maximum bainite transformation rate increases to 0.0436 μm/s. This can be attributed to the introduction of prior ferrite in the prior austenite grain, and the initial interface of bainite transformation changes from the austenite/austenite interface to the sum of the austenite/austenite interface and the ferrite/austenite interface. Therefore, as the intercritical annealing temperature decreases, there is a greater number of ferrite/austenite interfaces formed in the intercritical zone—that is, a greater number of bainite nucleation sites—thus increasing the bainite transition rate [32]. In addition, another consequence of the increment in bainite nucleation sites is that the maximum bainite transformation rate occurs earlier, as shown in Figure 2b and Table 1. However, with a low intercritical annealing temperature, there are more prior ferrites and the untransformed austenite formed before the isothermal bainite is finer, which inhibits the growth of bainite sheaves, in turn [16]. Therefore, there are more nucleation sites in the initial stage of bainite transformation, leading to a shorter incubation time and completion time for bainite transformation.

3.1.2. Microstructure Characterizations

The microstructural morphology of the experimental steel interannealed at different temperatures is shown in Figure 4. The composite microstructures of intercritical zone ferrite (F), bainitic ferrite (BF), and retained austenite (RA) were obtained via the intercritical annealing and austempering processes. Prior ferrite with elliptical, flake, or triangular morphology can be clearly observed in the IAA−810 specimen, mostly around bainite sheaves, as shown in Figure 4a. The presence of prior ferrite before bainite transformation can increase the number of nucleation sites for bainite, as shown in Figure 4b. With an increase in the intercritical temperature, the size of prior ferrite in the final microstructure gradually decreases, and the shape gradually changes from oval to flake. Moreover, it is difficult to distinguish the ferrite and bainite ferrite formed before isothermal transformation in the microstructure, as shown in Figure 4c,d. This is because austenite nucleates and grows with an increase in the intercritical annealing temperature, which reduces the volume fraction of prior ferrite. As the volume fraction of prior ferrite decreases, its shape gradually becomes a long strip, as shown in Figure 4e,f. In addition, with an increase in the intercritical annealing temperature, the length of the bainitic ferrite plates increases and the size of the blocky retained austenite increases. The increase in the length of the bainite ferrite plate is due to an increase in the prior austenite grain size, and the increase in the size of the blocky retained austenite is related to a decrease in the nucleation of bainite.

A diagram of microstructure growth under the IAA process is shown in Figure 5. The specimens contained prior ferrite (α) and untransformed austenite (γ₀) before isothermal bainite transformation, as shown in Figure 5a. During the isothermal process, bainite ferrite can nucleate and grow in the α/γ and γ/γ interfaces. The final microstructure of the experimental steel is composed of prior ferrite (α), bainite sheaves (B), and retained austenite (γ₁), as shown in Figure 5b.

3.2. Modelling and Results

3.2.1. Bainite Transformation Model

The microstructure of the experimental steel does not comprise full austenite grain before the bainite transformation, but is a mixture of prior ferrite and the untransformed austenite region. The transformation of bainite in the experimental steel subjected to the IAA process occurred in two distinct stages. The first stage comprised intercritical annealing, where prior ferrite nucleates at the prior austenite grain boundaries, propagating through autocatalytic transformation. In the second stage, BF nucleated and grew at the prior austenite grain boundaries, ferrite/austenite interfaces, and bainite ferrite/austenite interfaces. Therefore, the transformation rate of bainite under the IAA process (df/dt) can be expressed as follows [38]:

$$\frac{df}{dt}={\left(\frac{df}{dt}\right)}_G+{\left(\frac{df}{dt}\right)}_A+{\left(\frac{df}{dt}\right)}_F$$

(1)

where (df/dt)_G, (df/dt)_A, and (df/dt)_F are grain boundary nucleation, autocatalytic nucleation, and the nucleation rate of BF in the ferrite/austenite interface, respectively.

The introduction of prior ferrite before the bainite transformation increases the nucleation sites for the following isothermal bainite. At the same time, the introduction of ferrite affects the volume fraction of the available retained austenite, which has an important effect on (df/dt)_G and (df/dt)_A. In the calculation of the model, it is assumed that the influence of ferrite in the intercritical region on the grain boundary nucleation rate and autocatalytic nucleation rate is similar to that of formed bainite. Therefore, the dynamic model of the IAA process for the experimental steel is expressed as follows:

$$\frac{df}{dt}=\frac{kT}{h}\frac{Z\delta }{d}m\left(T_h-T\right)\left(1-f-f_{\mathrm{F}}\right)\frac{T_0-T}{T_{0\overline{x}}-T}\left[1+\left(f+f_b\right)e^{\frac{\Delta Q^{*}}{RT}}\right]\left(e^{\frac{-Q_G^{*}}{RT}}\right)$$

(2)

where k is the Boltzmann constant (1.38 × 10⁻²³ J·K⁻¹), h is the Planck constant (6.626 × 10⁻³⁴ J·s), Z is the geometrical factor (6), δ is the effective thickness of the austenite grain boundary (1 nm), d is the prior austenite grain size, m is the kinetics parameter, giving an account of the relationship between the martensite nucleation number and the degree of undercooling in K⁻¹, f is the bainite volume fraction, R is the molar gas constant (8.314 J·mol⁻¹·K⁻¹), and $Q_G^{*}$ and $Q_A^{*}$ are the activation energy of bainite nucleation at the grain boundary and autocatalytic nucleation, respectively. ΔQ^* = $Q_G^{*}-Q_A^{*}$. T_h is the critical temperature below which bainite nucleation occurs [39], T₀ is the critical temperature below which bainite growth can occur according to the displacive theory of bainite formation [39], and $T_{0\overline{x}}$ is the T₀ temperature at the beginning of the transformation. f_F is the volume fraction of prior ferrite produced in the intercritical annealing process. The parameters (T_h, T₀, $Q_G^{*}$, etc.) vary linearly with an increase in the carbon enrichment of austenite and are described in the literature [40]. Based on the above analysis, the parameters (T_h, T₀, $Q_G^{*}$, etc.) are improved. The improved formulas are expressed as follows:

$$T_h=T_{h\overline{X}}-C_1\frac{\left(f+f_m+f_b\right)\left(\overline{X}-X_b\right)}{1-f}$$

(3)

$$T_0=T_{0\overline{x}}-C_2\frac{\left(f+f_m+f_b\right)\left(\overline{x}-x_b\right)}{\left(1-f\right)}$$

(4)

$$Q_G^{*}=Q_{G\overline{X}}^{*}+K_{\Gamma }{C}_1\frac{\left(f+f_m+f_b\right)\left(\overline{X}-X_b\right)}{1-f}$$

(5)

where $T_{h\overline{X}}$ is the T_h temperature at the beginning of the transformation, $Q_{G\overline{X}}^{*}$ is the initial activation energy for grain boundary nucleation, C₁ (C₂) is a constant between $T_h$ ($T_0$) and carbon content, $K_{\Gamma }$ is the proportionality constant-related activation energy for bainite nucleation and temperature, and $X_b$ and $\overline{X}$ are the carbon content in bainite and blocky retained austenite.

3.2.2. Modeling Results

Each phase in the microstructure has its own linear thermal expansion coefficient (LTEC), so the proportion of each phase for the experimental steel can be estimated by the expansion rate of the specimens [41]. According to the results of previous studies, the linear expansion coefficient of austenite is 2.26 × 10⁻⁵ K⁻¹ and the linear expansion coefficient of ferrite is 1.45 × 10⁻⁵ K⁻¹ [42]. After the experimental steels were annealed at 810 °C, 830 °C, and 850 °C, the linear expansion coefficients of the specimens were 2.06 × 10⁻⁵ K⁻¹, 2.16 × 10⁻⁵ K⁻¹, and 2.23 × 10⁻⁵ K⁻¹, respectively, as shown in Figure 6. Calculations show that the volume fractions of the prior ferritic (F) of the experimental steel annealed at 810 °C, 830 °C, and 850 °C are 28.6%, 17.4%, and 9.3%, respectively, as shown in

Table 2. Volume fraction of each phase after isothermal annealing at different temperatures (%).

Specimen	Ferrite (F)	Bainitic Ferrite (BF)	Retained Austenite (RA)
IAA−810	28.6	54.4	17.0
IAA−830	17.4	58.4	20.0
IAA−850	9.3	55.9	35.3

. The calculation formula is shown below:

LTEC^A × V^A + LTEC^F × V^F = LTEC

(6)

V^A + V^F = 1

(7)

where LTEC^A and LTEC^F represent the linear expansion coefficients of untransformed austenite and prior ferrite, respectively, V^A and V^F are the volume fractions of untransformed austenite and prior ferrite, and LTEC represents the linear thermal expansion coefficients of the specimens at different intercritical annealed temperatures.

The final microstructure of the specimens manufactured through the IAA process is a mixture of prior ferrite, bainitic ferrite, and retained austenite. Therefore, the volume fraction of bainite can be calculated according to the linear expansion coefficient after the bainite transformation is completed, like in the case of calculating the volume fraction of prior ferrite. The linear expansion coefficients of the IAA−810, IAA−830, and IAA−850 specimens in the final cooling stage are 1.62 × 10⁻⁵ K⁻¹, 1.63 × 10⁻⁵ K⁻¹, and 1.75 × 10⁻⁵ K⁻¹, as shown in Figure 6. The volume fractions of bainitic ferrite for the IAA−810, IAA−830, and IAA−850 specimens are 54.4%, 58.4%, and 55.9%, respectively. The volume fractions of the phase for the different specimens—IAA−810, IAA−830, and IAA−850—are shown in Table 2.

The values of the various parameters in the model are presented in

Table 3. Calculated parameters used in the model of the different intercritical annealed specimens.

Specimen	${\mathit{T}}_{\mathit{h}\overline{\mathit{X}}}$/°C	C1/K	${\mathit{T}}_{0\overline{\mathit{x}}}$/°C	C2/K	KΓ/J·mol−1·K−1	m/K−1	d/μm
IAA−810	455.0	137.2	485.8	164.5	77.8	0.0134	11.2
IAA−830	400.0	83.3	491.0	127.3	88.9	0.0116	12.7
IAA−850	409.0	85.2	499.5	163.4	96.8	0.0123	14.7

. The parameters $T_{h\overline{X}}$, $T_{0\overline{x}}$, C₁, and C₂ were calculated using the MUCG 83 thermodynamic calculation software. The values of K_Γ and m were determined using the empirical formula detailed in the literature [43,44,45]. Additionally, the grain size of the prior austenite was simulated utilizing JMatPro 6.1 software.

Figure 7 shows a comparison between bainite transformation kinetics and model fitting under the IAA process. Figure 7a,c,e show a fitting comparison diagram of f−t under different IAA processes. It can be seen from the figure that the f−t experimental data under the IAA process are in good agreement with the fitting results, indicating that the fitting parameter results are credible. However, it is worth noting that the fitting results at the second corner of the IAA−850 transformation curve differ greatly from the experimental results, which indicates that the fitting autocatalytic activation energy is greater than the actual autocatalytic activation energy. Figure 7b,d,f show a fitting comparison of the transformation rates. It can be seen from the figure that the fitting result of the bainite transformation rate has the same trend as the experimental result. However, as can be seen from Figure 7b, the fitting rate in the later stage of the fitting result of the transition rate is significantly lower than the experimental result. This indicates that the autocatalytic activation energy is too large in the late transformation period of the IAA process.

The initial grain boundary activation energy ($Q_{G\overline{X}}^{*}$) of bainite transformation under the IAA process is fitted using Formula (2). Under different intercritical annealed temperatures, $Q_{G\overline{X}}^{*}$ and the autocatalytic initial activation energy ($Q_{A\overline{X}}^{*}$) are as shown in Figure 8. It can be seen in Figure 8 that the initial activation energy of the grain boundary and the autocatalytic activation energy decrease linearly with an increase in the volume fraction of prior ferrite. The introduction of prior ferrite in the prior austenite grain leads to the existence of the ferrite/austenite interface in the microstructure of the specimen, and this interface energy is much lower than that of the austenite/austenite interface [46], which is similar to the autocatalytic nucleation of bainite. Therefore, with an increase in prior ferrite content, the initial autocatalytic activation energy decreases. The height difference between the black square and the red dot in Figure 8 is the difference between the initial grain boundary activation energy and the initial autocatalytic activation energy, which ranges from 22.5 kJ/mol to 28.8 kJ/mol.

The influence of the ferrite/austenite interface on the bainite transformation rate in the IAA process can be calculated using Formula (1). The overall transformation rate of bainite in the IAA process is shown in Figure 7a,c,e. During the isothermal bainite transformation process, (df/dt)_G and (df/dt)_A were calculated to research the influence of the ferrite/austenite interface on the bainite transformation, as shown in Figure 9. As can be seen in Figure 9, the transformation rate of bainite at the ferrite/austenite interface gradually decreases with a decrease in the volume fraction of prior ferrite (the increase in intercritical annealing temperature). These results can be explained by the increase in the activation energy and the decrease in the number of bainite nuclei with an increase in the intercritical annealing temperature.

On the basis of the above discussion, (df/dt)_F is expressed by a formula containing (df/dt)_A, as shown in Formula (8):

$${\left(\frac{df}{dt}\right)}_F={\left(\frac{df}{dt}\right)}_A\left(a{f}^b\right)$$

(8)

where ($a{f}^b$) indicates the nucleation density of bainite at the ferrite/austenite interface and a and b are fitting constants, as shown in

Table 4. Fitting values in Formula (8) under IAA process.

	IAA−810	IAA−830	IAA−850
a	9.03	7.78	2.53
b	−0.061	−0.039	−0.025

.

Considering the increase in the volume fraction of bainite in the IAA process, the evolution of the nucleation rate of bainitic ferrite at the austenite grain boundary, bainite ferrite/austenite interface, and prior ferrite/austenite interface is shown in Figure 10. The bainitic ferrite plate (BF) nucleates at grain boundaries and grows into the grain. With an increase in the volume fraction of bainite (the extension of isothermal time), BF nucleated rapidly at the ferrite/austenite interface and dominated the total nucleation rate. This is due to the K-S relationship between the ferrite interface and austenite grain [39], and BF nucleates easily at the ferrite/austenite interface. The autocatalytic nucleation rate also increases with an increase in the volume fraction of bainitic ferrite. With an increase in the bainite transformation fraction, the volume fraction of untransformed austenite decreases. In other words, the volume fraction of austenite became the limiting factor of the BF nucleation rate. The nucleation rate of BF decreases at the ferrite/austenite and bainite ferrite/austenite interfaces. With an increase in the volume fraction of bainite, the grain boundary used for the nucleation site is depleted, so the BF nucleation rate at the grain boundary decreases linearly with an increase in the bainite fraction. The nucleation rate of bainite is always controlled by the nucleation rate at the ferrite/austenite interface, except at the beginning of the bainite transformation. This is the reason why the introduction of prior ferrite before bainite transformation greatly reduces the completion time of bainite transformation and increases the rate of bainite transformation.

According to the analysis of the kinetic transformation model, the transformation process of bainite under the IAA process can be described, as shown in Figure 11. First, carbon atoms diffuse and form a carbon-poor zone at the austenite grain boundary where the bainitic ferrite plate nucleates and grows, as shown in Figure 11a. With a prolonged isothermal time, the bainitic ferrite plate nucleates and grows at the α/γ interface and gradually occupies the dominant position of the nucleation rate. At the same time, the rate of bainitic ferrite plates nucleated at the formed bainite sheave tip also increases, as shown in Figure 11b. As the isothermal time continues to extend and the volume fraction of untransformed austenite decreases, the nucleation rate of the bainitic ferrite plate at all interfaces, including the austenite grain boundaries, the bainite sheave/austenite interface, and ferrite/austenite interface, decreases until the bainite transformation is completed, as shown in Figure 11.

4. Conclusions

Prior ferrite and austenite with different morphologies and volume fractions were prepared via the intercritical annealing process of medium-carbon, high-silicon steel. The effects of the ferrite/austenite interface and the volume fraction of prior ferrite on bainite transformation kinetics were investigated.

(1)

The lower the intercritical annealing temperature, the higher the volume fraction of prior ferrite formed before isothermal bainite transformation. The incubation time decreased from 687 s to 20 s and the completion time also decreased from 6018 s to 4680 s, while the bainite maximum transformation rate increased from 0.0271 μm/s to 0.0436 μm/s with an increase in the volume fraction of prior ferrite.

(2)

The introduction of prior ferrite can increase the number of nucleation sites, improve the bainite transformation rate at the beginning of the isothermal period, and accelerate the whole bainite transformation. The bainite sheaves nucleate not only at prior austenite grain boundaries, but also at the ferrite/austenite interface. The acceleration mechanism of prior ferrite on the bainite transformation is reflected in a shortened incubation time rather than the growth of bainite sheaves.

(3)

The experimental data and theoretical calculation results show that the bainite transformation activation energy decreases with an increase in the volume fraction of prior ferrite in medium-carbon bainite steel. Based on the experimental results, the transformation rate formula for the ultra-fine bainite transformation under the intercritical annealing process is also obtained, which is ${\left(\frac{df}{dt}\right)}_F={\left(\frac{df}{dt}\right)}_A\left(a{f}^b\right)$.