3.1. Microstructure evolution during initial quenching step
The martensite-austenite ratio in Q&P steels is controlled mainly by the initial quenching temperature. However, the stability of RA strongly depends on the partitioning temperature and time [9]. If the quenching temperature is too high, not all the fraction of austenite can be sufficiently stabilized by C diffusion during the partitioning step and it transforms into undesired secondary martensite during final cooling to the room temperature. Therefore, the quenching temperature should be precisely selected. A temperature range of the quenching step was selected based on the Lever-rule method commonly used for Q&P steels [18, 34]. The initial Ms temperature and the volume fraction of martensite formed at the quenching temperature were estimated based on the dilatometry response of the as-quenched sample austenitized at 1100°C, deformed with 50% deformation degree at 900°C and then directly quenched to the room temperature (Fig. 2a). The volume fraction of primary martensite (M1) formed at different quenching temperatures from 260–300°C (Fig. 2b) was determined by applying the lever rule to the as-quench dilatation curve. According to the obtained results, the volume fraction of primary martensite (M1) was in a range from 70 to 90 vol. % and the rest was a volume fraction of untransformed austenite. The target volume fraction of retained austenite in the final microstructure after the Q&P thermal route was about 12–20%, which is close to a fraction of RA usually observed in conventional low-alloy Q&P steels [13, 22, 23]. Therefore, the quenching temperature of 280°C was selected for further investigations for which the volume fraction of primary martensite (M1) and austenite were 80% and 20%, respectively.
3.2. Microstructure evolution during partitioning at different temperatures
Based on the dilatometric curve in Fig. 2b, the fraction of primary martensite (M1) formed during the initial quenching step was 80 vol.%. After initial quenching at 280°C samples were reheated to various partitioning temperatures from 350°C to 450°C for 90s (Fig. 3a). An expansion of samples isothermally held at 350–420°C was noted, which indicates bainite formation during the partitioning step. However, the recorded dilatometric response was relatively small, which means that a fraction of bainite formed during partitioning was limited. It was observed that the fraction of bainite formed in the sample increased as the partitioning temperature decreased. The highest fraction of bainite was observed in the samples partitioned at 350°C and 380°C. During partitioning at 450°C a slight contraction of a sample was recorded, which indicates the formation of carbides in primary martensite (M1). After partitioning at all temperatures martensitic transformation was recorded during cooling. The martensitic transformation becomes less intense with an increase in partitioning temperature.
The quantitative analysis of the microstructure after the full Q&P thermal routes was summarized in Table 2. Obtained results showed an increase in a volume fraction of retained austenite and a decrease in a volume fraction of secondary martensite (M2) with increasing temperature of partitioning process (Fig. 3b). The deviation from linearity of the dilatometric curves upon final quenching to the room temperature indicated the martensitic transformation of low-stable austenite. However, the dilatometric response for all thermal variants was small, which means that small fractions of secondary martensite (M2), 2–6 vol.% (Table 2) were formed in a microstructure. The highest fraction of undesired secondary martensite (M2), about 6 vol. %, was observed in a sample partitioned at the lowest temperature of 350°C. The highest fraction of retained austenite (16.0 vol.%) together with about 2 vol.% of secondary martensite (M2) was revealed for the specimen after partitioning at the highest temperature of 450°C. For this thermal variant the carbon content in RA was also the highest (0.81 wt.%). It resulted from the intensity of carbon partitioning from martensite to austenite, which increases with increasing partitioning temperature. In general, it was observed that a carbon content in RA increased with increasing partitioning temperature (Table 2). Besides the formation of small faction of carbides at 450°C (Fig. 3a), the fraction of carbon available in the solid solution was high enough to stabilize the almost entire fraction of retained austenite.
Figure 4 shows the dilatometric curves for samples isothermally held during 300s at different temperatures from 400°C to 450°C. It was observed that the thermally-activated processes occurring during the partitioning step changed from formation of bainite to precipitation of carbides with increasing isothermal holding temperature (Fig. 4a). It was also observed that the thermal stability of austenite expressed by a decrease in Ms temperature, increased with increasing partitioning temperature (Fig. 4b). The highest carbon content in RA was observed in samples partitioned at 420°C and 450°C, 0.78 and 0.76 wt.%, respectively (Table 2).
Table 2
Summary of the volume fractions of individual phases and carbon content in retained austenite (RA).
QT, °C | PT, °C | Pt, s | Ms1, °C ± 5 | M1, % ± 3 | Ms2, °C ± 5 | RA, vol. % ± 2 | C in RA, wt.% ± 0.2 | M2, % ± 3 |
280 | 350 | 90 | 373 | 85 | 275 | 10.2 | 0.53 | 6 |
280 | 380 | 90 | 270 | 12.0 | 0.58 | 4 |
280 | 400 | 90 | 268 | 12.0 | 0.60 | 3 |
280 | 420 | 90 | 266 | 13.6 | 0.78 | 2 |
280 | 450 | 90 | 264 | 16.0 | 0.81 | 2 |
280 | 400 | 300 | 279 | 14.8 | 0.63 | 1 |
280 | 400 | 900 | - | 14.7 | 0.76 | - |
280 | 420 | 300 | 266 | 13.3 | 0.78 | 2 |
280 | 450 | 300 | 262 | 15.3 | 0.76 | 1 |
Figure 5 shows the SEM images of specimens partitioned at various temperatures. The microstructures were composed of low-C primary martensite (M1) formed during initial quenching and retained austenite in a form of thin laths. Some fraction of blocky secondary martensite (M2) areas were observed in the specimen isothermally held at 350°C as a result of martensitic transformation of low-stable austenite during final cooling to the room temperature. It was observed that secondary martensite (M2) was heterogeneously distributed in the microstructure, which can be related to the segregation of alloying elements typically observed in Mn-containing steels [19]. It is beneficial that the presence of carbides was not observed for all thermal variants, even in a sample partitioned at 450°C, for which dilatometric results (Fig. 3a) showed the formation of small fraction of carbides. It is related to the increased Si and Al contents in steel (Table 1), which delay the formation of carbides in TRIP-aided steels [27, 35, 36].
3.2. Microstructure evolution during partitioning within different times
The duration of partitioning step also affected the microstructural evolution of the investigated steel. Figure 6a shows dilatometric curves for samples quenched at 280°C and partitioned at 400°C for different times: 90s, 300s and 900s. Two stages were observed in the dilatometric curves. The first stage is related to the expansion, which was noted during about 180s. It was related to the formation of bainite in the microstructure. Then, the plateau region was observed. During the second stage after about 360s, the continuous decrease of the sample length was noted, which was related to the formation of carbides in primary martensite (M1). The obtained results showed that the optimal variant of partitioning step at 400°C allowing to obtain the microstructure without carbides was 300s.
It was observed that the fractions of retained austenite after partitioning during 300 and 900s were similar: 14.7 and 14.8 vol.%, respectively, while for the partitioning time of 90s the fraction of RA was lower: 12.0 vol.% (Table 2). It means that partitioning during 300s was sufficient to enrich the austenite in carbon to stabilize almost all its fraction to room temperature. Only 1 vol.% of secondary martensite (M2) formed during final quenching to room temperature was noted for a specimen after isothermal holding at 400°C for 300s (Fig. 6b). The formation of secondary martensite (M2) was not observed for the sample isothermally held at 400°C during 900s. However, the dilatometric response showed the formation of carbides. However, their fraction was small and therefore the amount of carbon in solid solution was sufficient to stabilize the entire fraction of austenite to room temperature. The duration of partitioning step had an influence on enrichment of austenite in carbon because it was noted that as the partitioning time increased, the carbon content in the retained austenite increased from 0.60 wt.% for a specimen isothermally held during 90s to 0.76 wt.% for a specimen isothermally held during 900s.
SEM micrographs in Fig. 7 show that the microstructures of all time variants of partitioning step at 400°C were composed of primary martensite (M1) laths and retained austenite in a form of thin laths. Small areas of secondary martensite (M2) were observed in the sample isothermally held during 90s (Fig. 7a). The presence of small fraction of elongated precipitates inside the primary martensite (M1) laths was observed only for a sample isothermally held during 900s (Fig. 7d).