1 Introduction

Quenching and Partitioning (QP) is a heat treatment firstly proposed by Speer et al. in 2003 which is meant to introduce a fraction of metastable retained austenite into a martensitic matrix.[1] The goal is to obtain a set of mechanical properties that cannot be obtained using traditional routes, by exploiting a multiphase microstructure. Retained austenite is indeed a soft phase, which increases ductility and undergoes a strain-induced transformation into martensite when deformed (the TRIP effect), thus delaying the onset of necking.[2,3] The final result is a material with high tensile properties and improved ductility.[4] In the QP process, the steel is heated until full austenitization and then quenched between Ms and Mf. This step is followed by partitioning, which consists of isothermal holding at the quenching temperature (in this case it is called single step) or at a higher temperature (double step), which allows carbon to diffuse from the supersaturated martensite to the austenite, stabilizing it at room temperature.[1] Not only the amount of retained austenite, but also its stability is important to maximize the effect of the treatment. Depending on its dimension, morphology, and orientation, a dissimilar stability can be obtained, leading to differences in the final properties, and a fine, stable, retained austenite dispersion is reported to be the most preferable condition for austenite effectiveness.[5,6,7]

Carbon diffusion from martensite to austenite is not the only phenomenon occurring during partitioning, as competitive mechanisms such as carbide precipitation happen at the same time: as it reduces the amount of carbon effectively at disposal for partitioning, it must be as controlled as possible in order to maximize the effectiveness of the treatment.[8,9,10,11]

The first studies on QP were performed on low-carbon steel grades with the aim of obtaining the third-generation advanced high strength steels (AHSS) suited for automotive applications, such as body panels and structural components, and are known for their excellent combination of strength, toughness, and formability.[8,12,13,14] The effectiveness of the QP treatment is related to the chemical composition of the alloy.[15,16,17,18] The addition of silicon to the microstructure (>1 pct wt) is reported to help austenite stabilization delaying cementite formation, leading to a positive effect on the final result of the treatment, even if it is not capable of suppressing transition carbide formation, while manganese can help austenite stability and hardenability.[19,20,21,22,23] Steels such as QP980 and QP1180 are featured by the presence of these elements in their compositions.[24,25,26]

However, quenching and partitioning also shows great potential also for other applications, as the set of properties that can be obtained, such as high tensile strength and improved ductility, are not obtained with traditional routes. Nevertheless, few studies have been conducted on commercial medium-carbon steels. The main drawback is the difficulty in controlling the competitive phenomena that occur during the treatment because of non-tailored composition. Some studies have shown promising results on commercial steels, as fractions of retained austenite were successfully stabilized, and interesting tensile properties coupled with good elongation at break were obtained. Hong et al. obtained on a commercial low silicon boron steel a maximum UTS of 1731 MPa, coupled with a 10.8 pct elongation.[27] Han et al. worked on B1500HS hot stamping steel, reaching UTS of 1600 MPa and elongation of 10 pct,[28] while Kumar et al. obtained for a medium-carbon steel an UTS of 2091 MPa coupled with 12 pct elongation.[29] These results are promising for studying the feasibility of QP on other classes of commercial steels, verifying its wide range applicability, and exploiting outstanding mechanical properties that are desirable for subsequent industrial applications.

30MnV6, in particular, is a commercial medium-carbon microalloyed steel containing small amounts of niobium and vanadium and is commonly used in high strength applications, such as the construction of bolts, fasteners, and mining equipment. Its chemical composition is characterized by a carbon concentration that allows the generation of martensite through quenching, whereas the presence of manganese is desirable to stabilize the retained austenite. Nevertheless, the addition of vanadium, which is typically present in microalloyed steels to enhance strength and toughness, might form carbides, effectively reducing the amount of carbon present for partitioning.[30,31,32] The achievement of high tensile properties can therefore be interesting for this material to reduce the section and thus the weight of components, as requested especially by automotive industry, while an increased hardenability is desirable for the plastic deformation processes typical of the production cycle. This material is already currently industrialized, and even if its composition is not optimized for the QP treatment, such treatment could be performed with promising performance.

The goal of this work is to design and perform Quenching and Partitioning treatments on commercial 30MnV6 steel and to evaluate the effect of the treatment on the final tensile properties of the material. A preliminary test on the effect of an intercritical mixed ferritic–austenitic starting microstructure on the final properties is performed.

2 Materials and Methods

The composition of the studied 30MnV6 steel is reported in Table I measured with a Bruker Q4 Tasman quantometer.

Table I 30MnV6 Composition
Full size table

Critical temperatures were computed through Thermocalc 2020b, using the equilibrium calculator feature and the Martensite model embedded in the software. The range for the times has been selected according to values found in literature for another steel grade with similar carbon content.[27]

After normalization at 850 °C, air cooling, and subsequent austenitization at 850 °C, quenching and partitioning treatments have been performed using muffle ovens; for the partitioning treatments, salt baths were used. Four different single-step QP treatments were performed as reported in Table II.

Table II QP Treatments Parameters
Full size table

For the intercritical condition, the material was heated up to 850 °C, then slowly cooled in oven to the intercritical region between A1 and A3 (750 °C) in order to obtain a mixed austenite–ferrite microstructure; then, it was quenched and partitioned at 250 °C for 10 minutes. This sample is named I-750-QP.

XRD analyses were performed through a Rigaku SmartLab SE using a D/Tex Ultra 250 as a detector. X-ray worked at 40 kV and 40 mA, with a Cu Kβ filter 1D. Scan range went from 35° to 120° with step size of 0.02° and a scan speed of 0.5°/min. Rietveld analysis was performed on the integrated software Smartlab Studio II.

EBSD (Electron Backscattered Diffraction) analyses have been done to investigate the presence of retained austenite using a Carl Zeiss EVO 50 equipped with FEG: specimens have been smoothly polished through silica gel to obtain a smooth surface.

Tensile tests were conducted, following the standard ASTM E8-M, with a MTS100 machine, with 2 mm/min crosshead speed and initial gauge length of 50 mm on subsize samples.

3 Results

Critical temperatures were calculated through Thermocalc simulations using as input data the chemical composition of the alloy made by OES (Table I), and the initial grain size, hypothesized as about 20 µm. The results are in good agreement with the established well-known empirical formulae,[33] and the actual values found are shown in Table III.

Table III Critical Temperatures
Full size table

X-ray diffraction tests are shown in Figure 1. The peaks of FCC γ-phase and α/α’ (BCC/BCT) phase are present in all the four different cases: the presence of the former ones suggests that a fraction of retained austenite is present in the microstructure at room temperature, meaning that it was successfully stabilized through the heat treatment.

Fig. 1
figure 1

XRD spectra (a) 220-10, (b) 220-20, (c) 250-10, (d) 250-20. In all the different cases, α/α’ Fe and γ-Fe peaks are present

Full size image

Retained austenite fractions were computed starting from X-Ray diffraction spectra, through Rietveld analysis. Carbon amount in austenite is computed through Dyson Holmes equation.[34] Crystallite size is calculated through Scherrer equation.[35] Results are shown in Table IV.

Table IV Retained Austenite Fractions
Full size table

The fraction of stabilized retained austenite is between 4.5 and 6.3 pct, which is lower than expected by the theoretical CCE model. However, the hypotheses of the model are not completely satisfied in the experimental campaign. For instance, full partitioning is not achieved due to the low temperature and carbide formation is not suppressed due to the insufficient amount of silicon of the alloy: these conditions decrease the effectiveness of austenite stabilization, thus reducing its final amount.

EBSD analyses were performed to observe the morphology and distribution of retained austenite into the microstructure.

EBSD analyses in Figure 2 show the presence of a fraction of retained FCC austenite (in yellow) into a martensitic microstructure (in red). Regarding morphology, it can be noticed that austenite appears in a fine dispersion, both in elongated and blocky shape. Distribution and morphology of austenite are observed as not only the amount of retained austenite is crucial, but also other factors which influence its stability and therefore the effect on mechanical properties, namely, grain size, morphology, and crystallographic orientation. Big grains are reported to transform in the first stages of plastic straining, globular grains are more stable than lamellar ones, and depending on the orientation, grains can either rotate to accommodate deformation or directly transform into martensite.[5,6,7]

Fig. 2
figure 2

EBSD phase and IPF maps: (A, E) 220-10, (B, F) 220-20, (C, G) 250-10, D, H) 250-20

Full size image

Tensile tests, reported in Figure 3, show that the different QP treatments have introduced high tensile properties, as UTS reaches 1620 to 1675 MPa. Tensile curves are featured by a high hardening, as shown by the high UTS/YS ratio (Table V). Elongation at break is high, reaching 14.8 pct in 220-20 sample.

Fig. 3.
figure 3

Engineering stress–strain curves comparison

Full size image
Table V QP and Reference QT Tensile Tests
Full size table

Instantaneous hardening coefficients were computed, in the plastic stretch between YS and UTS, using the following relationship:

$$ n_{{{\text{inst}}}} \left( i \right) = \frac{{\log \left( {\sigma_{i} } \right) - \log \left( {\sigma_{i - 1} } \right)}}{{\log \left( {\varepsilon_{i} } \right) - {\text{log}}\left( {\varepsilon_{i - 1} } \right)}} $$

The values of ninst were computed between YS and UTS because the presence of retained austenite is reported to delay the onset of necking, thus displacing UTS to higher deformation levels.[3,36,37]

Figure 4 shows the hardening coefficients in relation to the true plastic deformation. The values start from high values (0.41 to 0.48) and progressively decrease increasing the plastic deformation, in all the cases, moving asymptotically to value of about 0.1. Specimens 220-10, 250-10, and 250-20 show two different slopes in the evolution of ninst, while 220-20 case instead is featured by a more homogeneous decrease in the hardening coefficient.

Fig. 4
figure 4

Instantaneous hardening coefficients evolution along plastic deformation computed in the uniform elongation region between YS and UTS

Full size image

True strain values corresponding to uniform elongation limit (UTS) move from 0.047 of the 220-10 condition to 0.076 of 250-20.

The presence of two hardening stages has already been observed previously in AHSS steels with high percentages of martensite.[38,39,40] Initial hardening coefficients range from 0.41 to 0.48, while final values are of the same order of magnitude of 0.1. The initial hardening coefficient is higher for a less tempered martensite, which is coherent with the considerations of Hidalgo et al.[39] and Findley et al.[41] which state that in the first stages of deformation, hardening coefficient is more linked to the dislocation density of the martensitic matrix than to austenite presence, as confirmed also by Ebner et al.[42]

A comparative QT sample was tested. This sample is subjected to a tempering treatment at 250 °C for 20 minutes. This is meant to introduce a martensitic microstructure featured by the same tempering of the QP sample, while the difference between the two samples is the presence of retained austenite in QP sample. While UTS is very similar, as expected by the similar martensitic matrix, elongation at break is increased in the QP sample. Especially uniform elongation is enhanced, thanks to the presence of retained austenite which delays the onset of necking (Figure 5).

Fig. 5
figure 5

Engineering stress-strain curves comparison: 250-20 sample and corresponding 250-20-QT sample

Full size image

However, during partitioning, martensite is tempered; varying partitioning time and temperature, different levels of tempering are obtained, leading to different properties: tempering increases the ductility of martensite. Therefore, as the deformability of the material is linked both to the presence of retained austenite and to the tempering level, these two simultaneous effects are analyzed.

The effect of tempering on the final deformation of the material is investigated. EBSD analysis, through the calculation of the KAM (Kernel Average Misorientation), can help in the evaluation of the tempering level, as a more tempered martensite is less distorted, and shows therefore a distribution centered on lower values of misorientation.[39] As expected indeed, in Figure 6, martensite results less distorted, and therefore more tempered in the case 250-20 with respect to the case 220-10. This means that martensite, in the latter case, has undergone a slightly more intense tempering process.

Fig. 6
figure 6

Kernel Average Misorientation (KAM) analyses. Increasing tempering level, a slight shift of the distribution to lower angles is observed

Full size image

Nevertheless, the small difference in the tempering level of martensite between the different cases is not high enough to justify those differences in the uniform elongation obtained. Consequently, the increase of uniform elongation observed can be related to austenite stability and amount, and probably to a lesser extent, to tempering of martensite: these two effects are simultaneous and cooperative and contribute to the increase of the plastic stretch before UTS. To support this hypothesis, Voce parameter (1/εc ) was calculated for the 220-10 and 250-20 case, following the procedure explained by Angella et al.[43]: lower value of this parameter is correlated with a higher austenite stability.[44,45] The two calculated values of 1/εc are 47.5 for specimen 220-10 and 26.5 for 250-20, respectively, so the latter case features a more stable austenite.

The results of the tensile tests show that the UTS obtained is high and coupled with an improved ductility. However, the modeling of mechanical properties may require lower strengths than those obtained through the treatments presented. The idea is so to introduce another soft phase into the microstructure, which can decrease the overall tensile strength of the material. Ferrite can exploit these features, and moreover, as its carbon solubility is low, it does not trap the carbon necessary for the partitioning treatment. Thus, the study of the effect of a mixed ferritic-austenitic starting microstructure on the tensile properties of the material is presented.

A comparative specimen (called I-750-Q) was subjected to the same initial heat treatment (heating up to 850 °C, slow cooling to 750 °C) and then, instead undergoing quenching and partitioning, it was directly quenched. The aim of this sample was to be a reference for evaluating the effect of quenching and partitioning on the final properties of the material. Results are shown in Table VI and Figure 7.

Table VI Tensile Tests for Intercritical Conditions
Full size table
Fig. 7
figure 7

Engineering stress-strain curves comparison for intercritical conditions

Full size image

The difference between I-750-Q and I-750-QP is evident in Figure 7: while the first shows almost no ductility (1.7 of A pct), the latter is featured by an enlarged plastic stretch. Partitioning results therefore crucial for obtaining a microstructure with improved and sufficient ductility, which is not given by the starting ferritic–martensitic microstructure.

The comparison between the two QP specimens instead shows a similar A pct, but a very different hardening behavior: I-750-QP condition reaches UTS for higher strains than the 250-10 one, while the latter reaches higher UTS (1620 vs 1326 MPa). UTS is higher as a martensitic matrix is stronger than a mixed martensitic–ferritic one.

The analysis of the instantaneous hardening coefficients, presented in Figure 8, can help in the evaluation of this behavior. The trend evolution of the ninst between the two different cases is very similar, but some differences are noticed: as the hardening coefficients are evaluated in between YS and UTS, it is clear that the uniform deformation reached by the I-750-QP sample is higher with respect to the 250-10 one. The initial value of the hardening coefficient is higher in I-750-QP sample (0.58) with respect to the 250-10 one (0.44). Moreover, in the first case, the decrease in the hardening coefficient appears to be less steep than in the latter: this is probably due to the presence of ferrite, whose deformability enhances the instantaneous hardening rate with respect to the other case.

Fig. 8
figure 8

Instantaneous hardening coefficients for I-750-QP and 250-10 samples

Full size image

4 Discussion

The feasibility of Quenching and Partitioning treatments on the selected alloy was assessed. In different cases, higher tensile properties coupled with high ductility were obtained with respect to the QT condition.

The trends of ninst are generally characterized by a progressive decrease, which is related to the microstructural evolution occurring in the QP specimens during the deformation process. In the first stages after YS indeed, hardening coefficients are very high as it starts the plastic deformation process of the hard martensitic matrix and the deformation and the strain-induced transformation of austenite into fresh martensite; the initial values in particular are mostly related to the state of martensitic matrix and its dislocation density, which is directly related to its tempering level.[6,41,42,46]

Increasing the deformation level, progressively martensite hardenability decreases as the dislocation density reaches saturation[42]; meanwhile, as the strain-induced transformation of austenite proceeds starting from the less stable to the more stable grains,[6,47,48] the amount of untransformed retained austenite decreases, and so does the coefficient moving toward low values (0.1). The fraction of stable retained austenite at higher strains is also responsible for the delay of necking, increasing the stretch of uniform elongation.[3,36,37] The comparison between 250-20 and 250-20-QT sample shows that the presence of retained austenite consistently enhances not only the total elongation, but also the uniform elongation. This trend is similar to the one observed by De Moor et al.[46], Finfrock et al.,[4] and Ye et al.[48] for QP samples, and He et al.[49] and Chiang et al.[47] for DP steels.

Although a slightly different QP treatment can lead to different amounts and stabilities of retained austenite,[36] it must be noted that also different tempering levels of martensite are obtained, which affect the properties of the material,[39,50] as the overall hardening behavior of the material is given by the concurrent effect of the strain-induced transformation of retained austenite and the hardenability of the martensitic matrix. KAM analysis shows a slight difference in the tempering level between the different cases, as a lower misorientation is measured for a higher tempering level. This observation is consistent with the initial values of the instantaneous hardening coefficient shown in Figure 4, where the one of the less tempered 220-10 sample is higher than the 250-20 one, as they are linked to the state of the martensitic matrix.[39,42] However, the differences observed in KAM are too small to explain the important variations in the extent of uniform elongation. Consequently, the differences in austenite stability were the leading factor for discriminating the samples. Coherently with this hypothesis, the computation of Voce parameters (1/εc) for 220-10 and 250-20 samples shows increased stability in the latter case, which is featured by a lower 1/εc value.[43,44,45] A more stable austenite indicates that the strain-induced transformation into martensite is activated only at higher strains, delaying necking to higher deformation levels.[3,41] It is well reported that long partitioning times can lead also to competitive mechanisms such as iron carbide formation[10] and austenite decomposition into bainite.[42] Therefore, by reducing the amount of carbon at disposal for partitioning, the kinetics of these transformations are deeply influenced by temperature as they are governed by diffusion.[9] Single-step treatments are conducted at lower partitioning temperatures with respect to double step and consequently are featured by slower carbon diffusion kinetics. Under these conditions, longer partitioning times are required to obtain stable austenite than those observed for double-step treatments.[51] The material shows high mechanical properties owing to the martensitic microstructure, and also work hardening of austenite. In addition, vanadium forms carbides formed at high temperatures, which improve the mechanical properties by dislocation and grain boundary pinning.[52] However, vanadium carbides reduced the amount of carbon that effectively acted as an austenite stabilizer; therefore, the time required to achieve sufficient stability was higher. Nevertheless as the amount of carbon in retained austenite is increased with respect to its concentration in the alloy, partitioning occurred (Table IV). The EBSD phase maps (Figure 2) show that austenite grains are always small (less than 2 µm), indicating that a sufficient carbon concentration has been introduced only to this extent, and the smaller the carbon-rich austenite grains, the strongest the effect on the final properties of the material.[3,5,50,51]

The intercritical specimen shows a different behavior, both compared to the directly quenched and 250-10 condition, featured by the same partitioning treatment. The starting intercritical condition before quenching is characterized by low-carbon ferrite and high-carbon austenite. When quenched, high-carbon austenite is transformed into stronger martensite with respect to the QP condition, as the carbon concentration in martensite is higher. Therefore, the starting hardening coefficient for the intercritical condition is higher than that for the QP condition [3. 41, 42]. In both cases, progressive the deformation and transformation of austenite occur; however, in the intercritical condition, ferrite hardening slows down the decrease in the hardening coefficient.[15] The effect of this mixed microstructure also leads to a longer plastic stretch before UTS and to a lower YS and UTS with respect to the full QP condition.

The comparison between I-750-QP and I-750-Q highlights the effect of partitioning on the ductility. The first specimen showed a very high elongation at break with respect to the second, featured by non-tempered martensite, which broke with almost no elongation. Two main factors might have influenced this phenomenon: the tempering of the martensitic matrix and the effect of retained austenite. During partitioning indeed, carbon diffuses from the martensitic matrix, tempering it, to austenite stabilizing it. This results in an outstanding increase in ductility with respect to the fully quenched condition. The comparison with the 250-10 specimen instead shows the effect of ferrite on the tensile properties, as the two conditions feature by the same heat treatment, but a different starting microstructure. The presence of a fraction of ferrite, which is softer than martensite, gives a lower YS and UTS with respect to a fully martensitic matrix but enhances the resistance to necking, as the UTS is reached for a higher plastic deformation level.

5 Conclusions

In conclusion, the Quenching and Partitioning feasibility on commercial 30MnV6 steel was assessed. This work confirms the applicability of quenching and partitioning treatment on a commercial microalloyed steel which was not specifically designed for the above treatment, introducing property sets that are not achievable with more traditional treatments. The outcomes of this work are as follows:

  • Different single-step quenching and partitioning treatments are capable of effectively stabilizing a fraction of retained austenite (4.5 to 6.3 pct) in the microstructure.

  • Tensile properties are significantly influenced by the treatment, in comparison to the quenched and tempered condition: the UTS reaches 1620-1675 MPa, and a high A pct is obtained (14 pct) in the 250-10 and 220-20 case. The highest UTS/YS ratios are found for the 250-10 and 250-20 specimens. Regarding tensile properties, partitioning at 250 °C for 10 minutes appears to be the most promising treatment for the selected alloy. A comparison between 250-20 and 250-20-QT sample shows that the austenite presence enhances uniform elongation, delaying the onset of necking.

  • The hardening coefficients start from high values (0.41 to 0.48) and exhibit a progressive decrease with increasing deformation level. The initial values are mostly related to the tempering of martensite, and a slight decrease is observed with increasing tempering level. It is proposed that the extension of the plastic stretch is given mainly by austenite stability, and, to a lesser extent, by the tempering level of martensite. Longer partitioning times provide more time for carbon diffusion, leading to more stable austenite grains where the strain-induced transformation is delayed to higher strains, resulting in a higher uniform elongation.

  • The quenching and partitioning treatment performed starting from a mixed austenite–ferrite microstructure shows reduced YS and UTS with respect to the quenched and partitioned condition featured by the same partitioning parameters, but a longer plastic stretch, as expected because of the presence of a softer phase in the microstructure. The analysis of the hardening coefficient shows that the presence of ferrite increases the hardenability of the material and increases the stretch of the uniform elongation with respect to the reference quenched and partitioned condition.