Abstract
Microstructure evolution of 0·2C–5Mn steel during the intercritical annealing at different temperatures was studied. It was found that the microstructure gradually evolved from martensite structure into a structure consisted of austenite and ferrite during the intercritical annealing process. The retained austenite volume fraction reached the maximum value of 36·5 after about 10 min annealing at 680°C. It was also found that carbides precipitated during initial annealing stage and gradually dissolved during following annealing process for all annealing temperatures; and interestingly, the fresh martensite was found, indicating that part of the newly forming austenite was unstable and would transform into martensite when quenching. Based on the microstructural analysis and the calculation by Thermo-Calc software, it was proposed that the different evolution behaviours of the microstructure during annealing were not only controlled by the thermal kinetics of the austenite, but also affected significantly by the thermal stability of the austenite.
Keywords
Introduction
With the development of society, the concept of energy saving and ecofriend for vehicles is proposed in recent years. One of the most effective and straightforward way is to reduce the weight of the vehicles. At the same time, it is still very necessary to ensure the safety of the passengers. Therefore, increasingly attention has been paid on the new generation steel which has dual phase structure with very good combination of high strength and high ductility.1,2
As both the transformation induced plasticity (TRIP) steels and twinning induced plasticity (TWIP) steels have excellent ductility, it could be expected that increasing the fraction of the relatively stable austenite would effectively improve the ductility. 3 Existence of the metastable retained austenite has been thought to be beneficial as it can improve the ductility of the steels by TRIP or TWIP effects.4–9 Although the TWIP steels have excellent combination of strength and ductility, the large amount consumption of costly alloying elements and the subsequent troublesome process make it difficult to manufacture cost effective steel sheets for modern vehicles. 10 At the same time, a series of TRIP steels which were relatively cost effective were well developed and widely used in the autoindustry. 11 However, the ductility of those traditionally used TRIP steels with strength of around 1000 MPa could not meet the increasing requirements for the automobile already. Therefore, it is very necessary to find and investigate the new type steels. Recently, the medium manganese steels which have excellent mechanical properties of strength 800–1000 MPa and total elongation 30–40 were reported by Merwin. 12 The new type medium manganese steels with 3–9 wt-Mn which were studied in our previous work13–17 were also reported to be capable of producing substantially improved mechanical properties with total elongation high up to 30–40 and strength high up to 1–1·5 GPa, which were significantly higher than that of the conventional TRIP steels and comparable with that of the TWIP steels. The main heat treatment process designed by us for the medium manganese steels is that the steel will be firstly austenitised at a relatively high temperature (above Ae1) and then followed by quenching and intercritical annealing (between Ae1 and Ae3) to develop ultrafine duplex structure with austenite fraction about 30–40 vol.-. 16
In this paper, the medium manganese steel (0·2C–5Mn) was also processed by the heat treatment mentioned above. The samples were firstly austenitised and then quenched and annealed at different intercritical temperatures with different times. The microstructure evolution during the annealing process was examined by X-ray diffractions (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). It is well known that the formation and development of metastable austenite were significantly affected by temperature and time.
In previous work,11–17 they generally did not consifder both the time and temperature together and usually chose one fixed annealing temperature. The effects of different annealing temperatures with annealing time were not studied. And the variation of austenite fraction, the evolution of microstructure and the partitioning behaviour of manganese and carbon elements during these heat treatment processes are all not clear. Therefore, in this work, the effects of different annealing temperature with annealing time on the 0·2C–5Mn steel were emphatically studied.
Experiment procedures
The steel containing 0·2C–4·70Mn used in the study was prepared by a high frequency induction furnace under the vacuum atmosphere and cast to 50 kg ingots; then the ingots were homogenised at 1250°C for 2 h and forged into rods with diameters of 16 mm between 1200 and 850°C. These forged rods were austenised at 850°C for 0·5 h and then water quenched to room temperature. In the following heat treatments, the rods were cut into small specimens; and those small specimens were intercritically annealed at different temperatures with different times by salt bath and finally water quenched to room temperature again. On the basis of the phase diagram shown in Fig. 1, three heat treatments of interest were selected and the heat treatment process was generally schematised in Fig. 2 and the detailed parameters were given in Table 1. Three temperatures are generally located in the lower, middle and upper part of the α/γ two-phase region.
Phase diagram of designed steels calculated by Thermo-Calc software
Scheme of heat treating process in this study
Heat treatment parameters for study in this paper*
*TAu: austenisation temperature; tAu: austenisation holding time; TAn: annealing temperature; tAn: annealing time.
In this paper, the critical temperatures of Ae1 (630°C) and Ae3 (720°C) and the equilibrium phase diagram of the designed steels shown in Fig. 1 and the equilibrium volume fraction of reversed austenite and its manganese and carbon content were calculated by the commercial thermo-dynamic calculation software Thermo-Calc with its TCFE6 database. Volume fraction of austenite for specimens was estimated by XRD and microstructures of specimens were examined by SEM, and TEM. Manganese contents in ferrite and austenite were measured by energy dispersive spectrometer (TEM–EDS). Samples for XRD and SEM were mechanically grounded, polished and then etched respectively. And samples for TEM were ground to thin film with a thickness about 50 μm, and then electropolished in a twin jet machine in a solution of 5 perchloric acid and 95 alcohol at about −25°C. The XRD with Co Kα radiation was adopted to measure the volume fraction of retained austenite. The calculations were performed based on the integrated intensities of (2 0 0) α, (2 1 1) α, (2 0 0) γ, (2 2 0) γ and (3 1 1) γ diffraction peaks. And for each diffraction peak, there was a volume fraction of retained austenite Vi which was calculated from the integrated intensities of ferrite and austenite peaks with equation (1), and the volume fraction of retained austenite Vγ was the average of Vi
13
Results
Formation of austenite during heat treatment in (ferrite+austenite) two-phase region
Figure 3 shows the changes in the volume fraction of austenite measured by XRD after water quenched to room temperature from each heat treatment temperature in the two-phase (ferrite+austenite) region at 620, 650 and 680°C as a function of the holding time. Such a gradual reversion behaviour indicates that it is a diffusional phase transformation process during the intercritical annealing. And the growth of austenite is usually accompanied with the partition of Mn. 16
Changes in volume fraction of austenite after cooling from each heat treatment process in the (ferrite+austenite) two-phase region at 620, 650 and 680°C respectively
When the annealing temperature is low (620°C), the amount of retained austenite at room temperature increased solely with time as shown in Fig. 3. While at 650 and 680°C (Fig. 3), with increasing annealing temperature, the amount of retained austenite at room temperature firstly increased with time and then decreased. However, the amount of retained austenite for 650°C case decreased very slightly, while the amount of retained austenite for 680°C case (Fig. 3) decreased sharply and was much lower than that for the 650°C case and even lower than that for the 620°C case eventually.
The maximum volume fractions of retained austenite obtained under these heat treatment conditions were 26·6 vol.- (620°C), 33·0 vol.- (650°C) and 36·5 vol.- (680°C).
Figure 4 shows the hardness variation with the changes of annealing temperature and time. At 620°C, the hardness decreased solely with annealing time. And at 650 and 680°C, the hardness decreased at first and then increased with increasing annealing time.
Hardness changes after cooling from each heat treatment process in (ferrite+austenite) two-phase region at 620, 650 and 680°C respectively
Observation of microstructure evolution
Microstructure evolution characterised by SEM
The microstructure (examined by SEM) for 0·2C–5Mn steel which was austenised at 850°C for 30 min and followed by water quenching was presented in Fig. 5. It can be found that the microstructure is full martensite lath structure without carbides or austenite, which was also determined by TEM observation in the previous study. 16
Microstructure of 0·2C–5Mn steel austenitising at 850°C for 30 min with water quenching characterised by SEM
The microstructure evolution (examined by SEM) of 0·2C–5Mn steel which was intercritically annealed at 620°C after austenisation at 850°C was shown in Fig. 6. In this intercritical annealing process, the steel was annealed at a temperature located in the lower part of the two-phase region. It can be found that, during the annealing, at the early stage, the carbides precipitated and there were almost no other phases but martensite laths with precipitated carbides. And the amount of carbides increased with increasing annealing time. However, with the duration of annealing (more than 10 min), a solute rich γ phase (austenite) formed and developed predominantly at the martensite lath boundaries and remained stable when water quenched to room temperature 19 and the carbides, which precipitated at the early stage gradually, dissolved into the newly forming austenite leading to the decrease of carbides and the formation of duplex lath structure. And with the dissolving of carbides, the duplex structure became clear. And it can also be found that, although the amount of carbides decreased, it did not totally disappeared eventually. And it is in a very good agreement with the result predicted by the phase diagram as shown in Fig. 1.
Microstructure evolution of 0·2C–5Mn steel annealing at 620°C for a 10 s, b 1 min, c 10 min, d 1 h, e 6 h and f 12 h characterised by SEM
The microstructure evolution (examined by SEM) of 0·2C–5Mn steel intercritically annealed at 650°C after austenisation at 850°C was shown in Fig. 7. It can also be found that the carbides precipitated at the very early stage (only with a few seconds). And compared with that of the 620°C case (Fig. 6a), its amount increased. And the whole microstructure also consisted of martensite lath structure matrix and a lot of precipitated carbides. Austenite can hardly be found. And with the increase of annealing time, the carbides also gradually dissolved into the gradually forming austenite and became very little at last.
Microstructure evolution of 0·2C–5Mn steel annealing at 650°C for a 10 s, b 30 s, c 1 min, d 10 min, e 1 h and f 6 h characterised by SEM
The microstructure evolution (examined by SEM) of 0·2C–5Mn steel intercritically annealed at 680°C after austenisation at 850°C was shown in Fig. 8. In this intercritical annealing process, the steel was reheated to the temperature located in the upper part of the two-phase region which is close to the Ae3 temperature. Therefore, the volume fraction of reversed austenite is very high (Fig. 13). It can be found that the microstructure evolution process is almost the same as the process of 620°C (Fig. 6) and 650°C (Fig. 7) cases at the initial stage (less than 10 min). The austenite formed preferentially along the martensite lath boundaries and a microstructure of parallel laths of alloy rich austenite and alloy lean tempered martensite (ferrite) gradually formed. A large amount of carbides can also be found and gradually dissolved into the new forming austenite. And with increasing annealing time, the duplex structure became the most clear at 10 min as shown in Fig. 7d, which indicated that the amount of austenite reached the maximum at this temperature with this annealing time. And with longer annealing time, the austenite decreased, which agreed quite well with the result as shown in Fig. 3.
Microstructure evolution of 0·2C–5Mn steel annealing at 680°C for a 10 s, b 30 s, c 1 min, d 10 min, e 1 h and f 12 h characterised by SEM.
Microstructure evolution characterised by TEM
The initial microstructure after the austenisation and water quenching was determined as fully martensitic structure both by SEM and TEM. 16 The analysis of the microstructure after intercritical annealing (Fig. 9) indicated that part of the reversed austenite retransformed into martensite during the water quenching. And the final microstructure presents three phases: retained austenite, ferrite and the fresh martensite.
Microstructure of sample annealed for 36 h at 620°C characterised by TEM images
TEM images in Fig. 10 show the microstructure evolution of 0·2C–5Mn steel processed by annealing at 620°C with different times. In the 10 s annealed sample, the microstructure was fully martensitic structure (Fig. 10a). With increasing annealing time, the duplex structure (Fig. 10b and c) became clearer and the amount of austenite increased obviously, while the thickness of austenite lath increased slightly up to about 250 nm, which was still relatively very small. In the 6 h case (Fig. 10b), both granular and rod shaped carbides were found and which certified the results shown in Fig. 6. And in the specimen annealed for 36 h (Fig. 10c), few carbides can be found and instead, more austenite laths developed between martensitic laths, which may indicate that the austenite laths preferred to nucleate at the martensitic laths interface or the packet boundary rather than at the inside of each martensite lath. 13
TEM images showing microstructure evolution in reversion treated specimens at 620°C with a 10 s, b 6 h and c 36 h
The microstructure evolution of 0·2C–5Mn steel processed by annealing at 650°C with different times was characterised by TEM as shown in Fig. 11. The martensite microstructure was found in the 10 s annealed specimen (Fig. 11a) and retained austenite could hardly be found in this nearly full martensite structure. With increasing annealing time, the lamellar duplex structure became clear and the thickness of austenite lath also slightly increased (Fig. 11b and c). And carbides can hardly be found. In the 6 h annealed sample, the microstructure consisted of austenite laths, ferrite laths and/or fresh martensite. And the thickness of austenite lath was around 200 nm.
TEM images showing microstructure evolution in reversion treated specimens at 650°C with a 10 s, b 1 h and c 6 h
The microstructure evolution of 0·2C–5Mn steel annealed at 680°C with different times was shown in Fig. 12. When annealing time was very short, the microstructure was found nearly full martensite with little retained austenite thin film. And with increasing annealing time, the lamellar duplex structure became clear. Comparing the austenite lath of the 10 min case (around 200 nm shown in Fig. 12b) with that of 6 h case (around 400 nm shown in Fig. 12c), it can be found that the thickness of austenite lath also increased with increasing annealing time.
TEM images showing microstructure evolution in reversion treated specimens at 680°C with a 10 s, b 10 min and c 6 h
Discussion
Formation of austenite during intercritical annealing process
Base on the XRD results (Fig. 3), it is found that, at the very first stage (less than 1 min), there is almost no or very little retained austenite forming during all heat treatment cases. And at the second stage, the volume fraction of austenite increased with increasing holding time in all the cases of heat treatment while decreasing at the third stage in the cases of heat treatment at 650 and 680°C. This is generally due to the retransformation of reversed austenite to martensite19–21 (formation of fresh martensite) during the water quenching process.
In order to understand the retransformation behaviours of the intercritically annealed 0·2C–5Mn steel, the equilibrium volume fraction and manganese and carbon contents of reversed austenite were calculated by the commercial thermodynamic calculation software Thermo-Calc with its TCFE6 database (Figs. 13–15). From Fig. 13, it can be seen that the volume fraction of reversed austenite gradually increases with increasing holding temperature, and each levels off at around 36·0 and 47·2, and 63·8 at 620, 650 and 680°C respectively. Figure 14 shows that the manganese (Fig. 14a) and carbon (Fig. 14b) contents of reversed austenite in 0·2C–5Mn steels grade as a function of temperature and both of them decreased with increasing annealing temperature. And the manganese (Fig. 14a) and carbon (Fig. 14b) contents were not only predicated by Thermo-Calc but also estimated by TEM–EDS (for 36 h annealing) and XRD (for 36 h annealing). From Fig. 15, it can be found that the manganese content of reversed austenite at 680°C grades as a function of time and it decreased with increasing annealing time.
Equilibrium volume fraction of reversed austenite as function of temperature predicted by Thermo-Calc software
a manganese and b carbon contents of reversed austenite in 0·2C–5Mn steels grade as function of temperature estimated by TEM–EDS and XRD and also predicated by Thermo-Calc
Manganese of reversed austenite in 0·2C–5Mn steels grade as function of time at estimated by TEM–EDS
When the annealing temperature is low (620°C), the quenching stability of the reversed austenite will be good due to the enrichment of austenite stabilised elements (with ∼10·7 wt-Mn and ∼0·43 wt-C as shown in Fig. 14) in reversed austenite. Therefore, it is difficult for the reversed austenite (forming during the annealing process) to retransform to martensite during the cooling. And it is shown in Fig. 3 that the amount of retained austenite (after cooling to room temperature) increases all the time and does not decrease. And with increasing annealing temperature, the volume fraction of reversed austenite arises (Fig. 13), so the average content of the manganese and carbon elements in the reversed austenite will decrease (Fig. 14). As the higher the annealing temperature is, the less the average content of the manganese and carbon elements in the reversed austenite will be. The quenching stability of austenite will decrease with increasing annealing temperature. Therefore, the amount of retained austenite (after quenching to room temperature) will decrease at last. While above the 650°C (Fig. 3), the amount of retained austenite at room temperature decreased and the amount of retained austenite for 650°C case (Fig. 3) decreased very slightly, which indicated that the quenching stability of reversed austenite at 650°C (with ∼8·35 wt-Mn and ∼0·42 wt-C as shown in Fig. 14) did not decrease very much with longer annealing time. However, the amount of retained austenite for 680°C case (Fig. 3) was much lower than that for the 650°C case and even lower than that for the 620°C case after long time annealing. Therefore, it indicated that the quenching stability of reversed austenite at 680°C (with ∼6·76 wt-Mn and ∼0·31 wt-C as shown in Fig. 14) decreased a lot with longer annealing time as the amount of reversed austenite at 680°C increased very much (approaching 63·8 vol.- as shown in Fig. 13) and the manganese content of reversed austenite decreased a lot as shown in Fig. 15. And water quenching hence resulted in the retransformation of the reversed austenite into fresh martensite. 19 Therefore, the respective maximum volume fractions of retained austenite (after cooling to room temperature) obtained under those heat treatment conditions, which are 26·6 vol.- (620°C), 33·0 vol.- (650°C) and 36·5 vol.- (680°C) as shown in Fig. 3, are less than the volume fractions of reversed austenite (forming during the annealing process), which are 36·0 vol.- (620°C), 47·2 vol.- (650°C) and 63·8 vol.- (680°C) obtained at the same temperature as shown in Fig. 13.
Figure 4 shows the hardness variation with the changes of annealing temperature and time. The decrease in hardness is related to the recovery of martensite and the increase of retained austenite (Fig. 3). However, the increase in hardness results from the retransformation of the reversed austenite to martensite during the cooling process. 22 With increasing annealing temperature and time, the amount of fresh martensite increased. More reversed austenite retransformed into martensite for the 680°C case than for the 650°C case, which resulted that the increase in hardness for 680°C is larger than that for 650°C (Fig. 4).
It can also be found (Fig. 3) that the time needed for the different heat treatment processes to get the maximum volume fractions of retained austenite decreased significantly with increasing annealing temperature. In the 620°C case, the time needed for obtaining the maximum volume fraction of retained austenite (∼26·6 vol.-) is about or even more than 36 h. While in the 680°C case, the time needed for obtaining the maximum volume fraction of retained austenite (∼36·5 vol.-) is only about 5–10 min. It may be explained that the driving force (obtained through Thermo-Calc) for reversed austenite transforming and the diffusion coefficient of the manganese in austenite increased with increasing annealing temperature as shown in Figs. 16 and 17 respectively.
Driving force for reversed austenite transforming obtained through Thermo-Calc
Diffusion coefficient of manganese in austenite
Microstructure evolution during intercritical annealing process
Both the SEM and TEM results indicated that the microstructure evolution process was affected by the heat treatment parameters significantly. And the initial microstructure after the austenisation at 850°C and water quenching was determined as fully martensitic structure both by SEM and TEM. 16 For the SEM results shown in Figs. 6–8, as the solid solution of carbon in the ferrite phase is quite low (<0·0218 wt-), when annealing, in all cases (620, 650 and 680°C), the carbides precipitated from the martensite matrix, which was oversaturated of carbon. The higher the temperature was, the faster the carbides precipitated. With increasing annealing time, the austenite nucleated and grew up. As the solid solution of carbon in austenite is relatively high (the maximum approaching 0·45 wt- as shown in Fig. 14b), the formation of austenite needed a lot of carbons, which leads to the gradually dissolving of carbides into the austenite. The amount of carbides decreased with increasing austenite. And at high annealing temperature, the evolution results shown in Fig. 7 indicated that there was still no enough quenching stability for the reversed austenite even all the carbides dissolved into the austenite, as the amount of reversed austenite was so high that only a small amount of ferrite was present, which means that the average alloy content of austenite was very small. Water quenching to room temperature hence resulted in the retransformation of austenite into fresh martensite. 21 Therefore, the amount of retained austenite decreased.
Figures 10–12 show the TEM results of the microstructure evolution of the samples annealed at different temperatures with different times. In the very short annealing time stage (∼10 s), it can be found that the microstructure is still fully martensitic structure and nearly no retained austenite can be found for all annealing temperatures. It is in a good agreement with the SEM results shown in Figs. 6–8. In this stage, the martensite was only simply recovered as the driving force for recrystallisation was not high enough for ferrite recrystallisation. Without any deformation, the stored energy for recrystallisation is indeed not sufficient, though the martensitic structure has a high dislocation density. 23 The second step is the nucleation of austenite at lath boundaries, packet boundaries and prior austenite grain boundaries. 24 The third is the growth of austenite and the partitioning of manganese in austenite. 16 Austenite with lath-like structure may grow mainly along lath boundaries. Ferrite also appears as lath-like as it is located between two adjacent austenite laths. They also indicate that, when increasing the annealing temperature (from 620 to 680°C), due to the retransformation of reversed austenite to fresh martensite 21 when water quenched to room temperature, the amount of retained austenite decreased.
Conclusions
Intercritical annealing treatments were applied on the 0·2C–5Mn steel. Austenite reverted transformation process was observed during the intercritical annealing. The amount of the retained austenite and the evolution of microstructure at different temperatures with increasing annealing time were studied. The results were generally summarised as follows:
After austenisation at 850°C for 0·5 h and water quenching, a full martensite structure was found in the specimen. During the following annealing process, it gradually transformed into the austenite and ferrite duplex structure. After the quenching, as part of the reversed austenite retransformed into fresh martensite, the final structure became a complex ultrafine structure of retained austenite, ferrite and fresh martensite.
The annealing time and temperature are the two key parameters to obtain and control the final compound structure. At the relative lower annealing temperature (620°C close to Ae1 temperature), the austenite volume fraction increased continuously with increasing annealing time. During the 650°C annealing process, the austenite volume fraction increased with increasing annealing time firstly, but then slightly decreased. While at the relative higher annealing temperature (680°C close to Ae3 temperature), the austenite volume fraction increased sharply up to the maximum 36·5 vol.- with only 5–10 min and then very quickly decreased with increasing annealing time.
Under the different heat treatment conditions, the maximum retained austenite volume fractions were obtained about 26·6, 33·0 and 36·5 vol.- at 620, 650 and 680°C respectively.
Footnotes
Acknowledgements
This research is supported by the National Natural Science Foundation of China (nos. 51371057 and 51171087) and National Basic Research Program of China (973 Program, no. 2010CB630803).