Effect of Intercritical Deformation on Microstructure and Mechanical Properties of Quenching and Partitioning Low Carbon Multiphase High-Strength Steel

Abstract

Low carbon multiphase high strength steel is widely used in the automobile industry. In this work, the effect of intercritical deformation on the partitioning of alloying elements and the evolution of microstructure, as well as the effect of retained austenite stability on mechanical properties, were studied in a low carbon steel. The results demonstrate that the intercritical deformation enhances the C, Mn partition from ferrite to austenite during annealing at 770 ℃, and the volume fraction of the retained austenite increased from 8.8% to 12.3%. The DIQ&PB sample shows good balance between strength (1226.5 MPa) and ductility (24.4%), whose product of strength and elongation reached a larger value of 29926.6 MPa·% due to the intercritical deformation. This research provides theoretical guidance for the process design of automobile high-strength steels, considering the integration between rolling and heat cycles.

1. Introduction

Retained austenite played an important role on the mechanical properties of the multiphase steel. In the deformation process, the transformation of retained austenite into martensite leads to the TRIP effect, which can delay necking and improve the strength of steel [1,2,3]. Therefore, how to improve the stability and content of retained austenite has become a hot topic of research. In the 1980s, researchers have found that some alloy elements, such as Mn, can diffuse from ferrite to austenite in the intercritical region, which can stabilize austenite [4,5,6,7,8]. Thus, it is very important to adopt a reasonable means to promote the partitioning of Mn elements in the intercritical annealing.

Based on the quenching and partitioning process (Q&P), Liu et al. [9] used a novel hot stamping and quenching-partitioning process to obtain fine-grained austenite. Because the finer austenite grains provide more nucleation sites for the bainite, the refinement of prior austenite grains can increase the rate of bainitic transformation [10,11,12]. Wang [13] analyzed the effects of deformation and phase transformation on the martensitic transformation mechanism, microstructure and mechanical properties, and found that a certain amount of plastic strain can increase the retained austenite content and improve the stability of retained austenite. Meanwhile, hot deformation promotes dislocation multiplication and fine grain strengthening, which reduces the bainite transformation driving force and results in finer multi-phase organization after the quenching–partitioning process [14,15]. Tian et al. [16] investigated the effect of the intercritical deformation temperature on the retained austenite, and found that when the intercritical deformation temperature is higher than 800 °C, the enrichment of the carbon element in austenite is promoted, and the content of retained austenite is increased. A previous study has demonstrated that the stabilization of retained austenite shown by an increase in volume fraction and its carbon content, as a result of deformation in the intercritical area [17]. However, the effect of intercritical region deformation on the partitioning of alloying elements is rarely reported. Therefore, based on the theoretical knowledge of the relationship between elemental partitioning and intercritical deformation, a new deformation heat treatment process that is intercritical deformation–partial austenitizing-quenching and partitioning within the bainitic region (DIQ&PB) is proposed.

In this paper, we study the influence of intercritical deformation on the partitioning of alloying elements and the evolution of the bainite structure and discuss the effect of retained austenite stability on mechanical properties and work hardening behavior of low carbon bainite multiphase steel.

2. Materials and Methods

A low carbon steel with chemical compositions of 0.18C-1.54Si-2.12Mn-0.40Cu-0.31Ni (wt %) was melted in a ZG-50 vacuum induction furnace to obtain ingots. The billet was heated to 1200 °C and held for 3 h, then hot rolled to a thickness of 15 mm. The critical phase transformation temperatures Ac₁, Ac₃, Ms and Mf are determined to be 725 °C, 891 °C, 242 °C and 391 °C (see Figure 1), respectively. Ac₁ is the transformation temperature from ferrite to austenite during heating, and Ac₃ is the complete austenitizing temperature. Cylindrical specimens with a diameter of 10 mm and height of 100 mm were cut from the rolled steel, subjected to different thermal-processes. (a) Intercritical hold-quenching (IQ): heated to the intercritical temperature of 770 °C at 5 °C/s and soaked for 1800 s, then quenched to room temperature by water. (b) Intercritical deformation hold-quenching (DIQ): 15% true strain (small deformation) at a compressive strain rate of 1 s⁻¹ at 770 °C + IQ. (c) Intercritical annealing (partial austenitizing)–quenching and partitioning within the bainitic region (IQ&PB): heated to the intercritical temperature of 770 °C at 5 °C/s and soaked for 1800 s, subsequently cooled to 400 °C at 5 °C/s soaking for 180 s, then quenched to room temperature by water. (d) DIQ&PB: in advance of the IQ&PB process, the sample was firstly deformed immediately by a 15% compressive deformation with a strain rate of 1 s⁻¹ at 770 °C. In different thermal processes, considering the stability of equipment operation and material properties under different heating/cooling rates, the rate is selected as 5 °C/s based on Reference [18].

The above thermal-mechanical trials are performed on a Gleeble-3500 thermal simulator, with the sample shape shown in Figure 2. Three tensile samples were tested to obtain the mechanical properties. The tensile sample was tested on the AGX-100kN machine at room temperature at extension rate of 2 mm/min, and the contact extensometer was used to calculate the engineering strain. The microstructure was characterized by using a Hitachi S-4800 field emission scanning electron microscope (SEM) equipped with an electron backscattered diffraction (EBSD) system, which was used to distinguish the retained austenite. In order to study the effect of deformation on elemental diffusion behavior, specimens were studied by a JXA-8230 electron probe micro analyzer (EPMA) equipped with an energy dispersive X-ray spectrum (EDS) system at an operating voltage of 20 kV, current of 2 × 10⁻⁸ A and a step size of 5 nm. The dislocation and retained austenite morphology were characterized by a JEM-2010 transmission electron microscope (TEM). The volume fraction of the retained austenite was measured using a D/max2500PC-X ray diffractometer with Cu Kα radiation on a voltage of 40 kV, a current of 15 mA, a step of 0.02°, and a rate of 1°/min. In order to reduce the error, three samples were used for each study, and five fields of view were used for the measurement of dislocation density and microstructure, and then the typical structure was selected.

The C content in retained austenite was calculated using the Equation (1) according to Ref. [19], where the Cγ is the carbon content in austenite; α_γ is the lattice parameter of the austenite determined using Equation (2). λ is the wavelength of the radiation given as 1.5406 Å, (h, k, l) are the Miller indices of the FCC austenitic crystal, and θ is the Bragg angle. The carbon content in the three retained austenite peaks ((200), (220), (311)) was calculated quantitatively, and took the carbon content in the highest retained austenite peak as the carbon content in the retained austenite.

$${\mathrm{C}}_{\mathsf{\gamma }}=\left({\mathsf{\alpha }}_{\mathsf{\gamma }}-3.547\right)/0.046$$

(1)

$${\alpha }_{\gamma }=\frac{\mathsf{\lambda }\sqrt{{\mathrm{h}}^2+{\mathrm{k}}^2+{\mathrm{l}}^2}}{2\sin \mathsf{\theta }}$$

(2)

3. Results

3.1. Dislocation Multiplication and C, Mn Partitioning Behavior

The typical dislocation morphology of IQ and DIQ samples is shown in Figure 3, where M represents martensite and F represents ferrite. As shown in Figure 3, the dislocation density of DIQ sample (Figure 3b) is greater than that of IQ sample (Figure 3a). To obtain the dislocation density, five fields are selected with the secant method. The dislocation densities are quantitatively measured using the classic formula [20,21,22]:

$$\rho =\frac{1}{\mathrm{t}}(\frac{\sum {\mathrm{n}}_{\mathrm{V}}}{\sum {\mathrm{L}}_{\mathrm{V}}}+\frac{\sum {\mathrm{n}}_{\mathrm{h}}}{\sum {\mathrm{L}}_{\mathrm{h}}})$$

(3)

where ρ (m⁻²) is the dislocation density; n_V, n_h is the number of intersected vertical, horizontal grid lines within the dislocation network, respectively; L_V, L_h is the length of vertical, horizontal grid lines, respectively; due to the limitation of the experimental equipment, the foil thickness t is selected as 200 nm according to the reference (the foil thickness is supposed to be in the range of 180~220 nm) [23]. The intercritical compressive deformation increases the ρ value from 0.281 × 10¹⁴ m⁻² (IQ) to 1.154 × 10¹⁴ m⁻² (DIQ). It is not reasonable to obtain dislocation densities just by evaluation of multiple TEM fields of view, as in no case all dislocations are visible at the same time. The global dislocation density should be determined by X-ray diffraction according to the literature [23]. Therefore, the dislocation density obtained in this study can only reflect the influence of deformation on dislocation, and cannot accurately and quantitatively characterize the global dislocation density.

The EPMA images of C and Mn distribution in IQ and DIQ samples are shown in Figure 4. The microstructure after quenching is composed of martensite and ferrite (Figure 4a,b), and as shown in Figure 4c–f, C and Mn elements are enriched in the martensite region. Moreover, it can be found from Figure 4 that the segregation degree of C and Mn elements in the DIQ sample is higher than that in the IQ sample. After deformation of the sample, the dislocation density increases, and the vacancy and interstitial atoms increase, which leads to the improvement of the diffusion channels of elements, and further promotes the diffusion of C and Mn atoms. Therefore, the partitioning efficiency of C and Mn from ferrite to austenite is higher than that of the IQ sample.

3.2. Microstructure and Retained Austenite

Figure 5a,b shows the microstructure of IQ&PB and DIQ&PB, respectively, which has similar multi-phase structure, including ferrite (F), bainite (B), martensite/austenite (M/A) and retained austenite (RA). The microstructure is obviously refined after the intercritical deformation, the size of the bainite lath decreases from 16 μm for IQ&PB (Figure 5a) to 8 μm for DIQ&PB (Figure 5b) and the morphology of DIQ&PB is dense and uniform. The TEM image of DIQ&PB (Figure 6) shows the cross-distribution of ferrite, bainite and martensite, and as shown in Figure 6b, the martensite exists in twinning-like morphology, which is defined as twin martensite.

Typical XRD patterns of IQ&PB and DIQ&PB samples are shown in Figure 7. Figure 7 shows that both samples contain diffraction peaks of residual austenite, and the peak of the DIQ&PB sample is higher than that of the IQ&PB sample. The content of the retained austenite of IQ&PB and DIQ&PB specimens is calculated to be 8.8% and 12.3%, respectively. In addition, the C content in retained austenite is determined to be 1.0% and 1.3%, respectively. This is due to the fact that the enhancing Mn amount in austenite reduce the carbon chemical potential in the austenite and cause the diffusion of C-atoms from the Mn-depleted area [24]. In addition, the diffusion of carbon atoms is accelerated because the intercritical deformation increases the dislocation density, and along the very boundaries of structural elements, diffusion is often faster [25,26]. In summary, the Mn segregation during annealing at 770 °C promotes the secondary C diffusion to untransformed austenite in bainitic, which is more pronounced in the DIQ&PB sample than the IQ&PB sample. In consequence, a larger amount of retained austenite is obtained in the DIQ&PB sample.

EBSD associated with TEM characterizations are performed to identify the distributions of retained austenite. As shown by the arrow in Figure 8, retained austenite has two different morphologies: film-like and blocky. The film-like retained austenite is observed between bainitic laths (see Figure 9b). The blocky retained austenite mainly exists on the original austenitic grain boundary (Figure 9e), whose shape stays relatively irregular. Besides, the retained austenite in the DIQ&PB specimen shows a refined size and more intensive distribution, due to the introduced intercritical deformation.

3.3. Tensile Behavior

Figure 10 shows the stress–strain curves of different processes. As shown in Figure 10, the strength and elongation of the DIQ&PB sample are higher than that of the IQ&PB sample. After the intercritical deformation, the ultimate tensile strength is increased from 1084.1 ± 2.7 to 1226.5 ± 4.4 MPa, and the elongation is increased from 20.9% ± 1.8% to 24.4% ± 1.2%, resulting in the significant increase in the production of strength and elongation from 22657.7 to 29926.6 MPa·%. The grain refinement and twin martensite induced by intercritical deformation are essential for strength improvement. The intercritical deformation increases the number of dislocations, and then promotes the diffusion of carbon atoms, which increases the carbon content in the retained austenite to 1.3%, resulting in a higher amount of retained austenite after deformation. Therefore, the larger content of retained austenite (12.3%) ensures the superior plasticity compared to the IQ&PB specimen (8.8%). In addition, the hardening mechanism is mainly the dislocation proliferation in ferrite at the initial stage of deformation, and TRIP (transformation-induced plasticity) effect occurs when the strain reaches a certain value. The previous study has demonstrated that the TRIP effect is closely related to the retained austenite in the obtained multi-phase microstructure, and the prolonged TRIP domains is beneficial to resisting plastic deformation and delay the necking occurrence [27]. Studies have demonstrated that the mechanical stability of RA is related to its chemical composition, shape and distribution. In addition, TRIP effect can occur in blocky-like RA and film-like RA [28,29]. The growth mode of blocky-like RA is different from that of film-like RA. Generally, the blocky-like RA is larger in size and its alloy composition is lower than that of the film-like RA. Therefore, blocky-like RA has lower mechanical stability than film-like RA and is more likely to incur TRIP effect [30]. The soft retained austenite can relax the local stress concentration during deformation through its transformation to martensite, and the fresh transformed martensite can passivate the crack tips and provide extra second-phase interfaces for ductile dimple formations.

4. Conclusions

(1)

The intercritical compressive deformation promote the diffusion of C and Mn elements from α-phase to γ-phase. After compression deformation, the size of the bainite lath decreases from 16 μm to 8 μm and the morphology is dense and uniform. Block twin martensite distributes in the position of the adjacent ferrite.

(2)

The blocky and film-like retained austenite can be observed in IQ&PB and DIQ&PB samples. The film-like austenite is located between the bainite lath, the blocky austenite mainly distributes at ferrite/bainite interface. After compressive deformation treatment, the size of the blocky-retained austenite is obviously refined.

(3)

After the intercritical deformation, the ultimate tensile strength and elongation are improved, and the product of strength and elongation reaches 29,926.6 MPa·%.