Structure-Property Correlations of a Medium C Steel Following Quenching and Isothermal Holding above and below the M_s Temperature

Abstract

The processing of advanced multiphase high strength steels often includes isothermal treatments around the martensite start temperature (M_s) for achieving a refined microstructure comprising bainite-austenite and/or bainite-martensite-austenite phase constituents. The objective of this research work was to investigate the structure-property relationship for a medium carbon, high-silicon DIN 1.5025 steel (Fe-0.529C-1.67 Si-0.72Mn-0.12Cr (in wt.%)) following isothermal holding close to the M_s temperature (~275°C) to enable low temperature austenite decomposition. For realizing multiphase microstructures, DIN 1.5025 steel samples were austenitized at 900°C for 5 min and then quenched to the isothermal holding temperatures 350 and 250°C for various times ranging from 5 to 3600 s. Microstructural investigation corroborated the formation of multiphase microstructure comprising tempered martensite, bainite, retained austenite, and fresh martensite in both the samples isothermally held above (350°C) and below the M_s (250°C) temperature. The sample isothermally held at 250°C showed a much more refined microstructure in comparison to that held at 350°C due to the presence of a fraction of initial martensite laths which acted as potential sites for bainite nucleation. Also, the evaluation of mechanical behaviour showed that the best tensile properties in terms of high tensile strength and good ductility were achieved in samples with high volume fractions of both interlath and blocky retained austenite, particularly those isothermally treated at 350°C for 200 s and at 250°C for 600 s, respectively.

1. Introduction

The development of advanced high strength steels (AHSS) depends on the steel type, chemical composition, and the parameters affecting the heat treatment cycles, such as austenitizing time and temperature, cooling rate, and final quenching and holding conditions. Accordingly, various advanced high strength low alloy steels can be designed and developed with multiphase microstructures comprising a mixture of bainite, martensite and retained austenite (RA) phases and with various engineering properties in the end products.^1,2,3) The heat treatment process for achieving these microstructures involves rapid cooling to temperatures slightly above or below the M_s and holding at this temperature in order to facilitate the formation of fine lower bainite. On final cooling, some carbon-enriched untransformed austenite may transform into untempered, (high carbon) fresh martensite (FM), besides retaining the balance austenite stabilized down to room temperature (RT).^4,5,6) At holding temperatures higher than M_s in hypoutectoid steel specimens, the microstructure invariably consists of bainitic ferrite with or without carbides, depending on the alloying elements and austenite decomposition parameters (here, holding temperature and time), and RA in the form of thin interlath films, and/or coarse blocks.^7,8) The properties of these multiphase microstructures are comparable to those of carbide-free bainitic steels and quenched and partitioned (Q&P) steels that have a complex mechanical response to stress.^9,10,11)

The novelty of this metallurgical processing lies in the acceleration of the bainite reaction through the presence of a small fraction of athermal initial martensite (IM) formed a priori on quenching to the desired temperature below M_s.^12,13,14) The formation of IM implies the introduction of martensite-austenite interfaces, which act as preferential sites for bainite nucleation, in addition to the prior austenite grain boundaries.^13,15) Various reports have shown that the final microstructure obtained under these heat treatment conditions cannot be completely bainitic and hence, the final multiphase microstructure contains a martensitic-bainitic matrix in combination with martensite-austenite (MA) islands as well as carbon-enriched austenite.^7,16) Some researches have shown that the yield strength is increased compared to the specimens held at temperatures above the onset of martensite formation.^9,17,18) Research investigations have shown that the mechanical properties of these steel groups heat treated below M_s are influenced by various processes, such as the formation of IM, fine bainitic ferrite formed during isothermal holding, carbide precipitation, if any, during martensite tempering, carbon enrichment of untransformed austenite due to carbon partitioning and finally, transformation of some unstable austenite to FM during final cooling.^9,17,18,19) The role of RA in these multiphase microstructures is very complex and could have positive as well as adverse effects on the final engineering properties depending on the morphology, distribution, and of course the volume fraction of this microphase in the microstructures.^20,21,22)

While fine bainitic ferrite and/or RA are present, besides martensite (both fresh and tempered martensite), in the case of samples isothermally held below the M_s temperature, a sharply different bainite-austenite structure is achieved after isothermal holding above the M_s temperature with the possibility of some FM during final cooling. This research work investigates the mechanical behaviour of a medium carbon high-silicon steel (DIN 1.5025) after the application of different holding times both above (quenching and bainitic holding (Q&B))²³⁾ and below (quenching and partitioning (Q&P)) the M_s temperature and envisages understanding the influence of microstructural characteristics and phase constituents in establishing the structure-property correlation. While the evolution of the microstructures was characterized using different techniques, the mechanical properties were evaluated in respect of hardness and RT tensile properties. The results were further analyzed in respect of morphology and fractions of phase constituents, carbon enrichment of austenite, and formation of carbides to correlate with the properties and the fractographic features of the fractured tensile samples.

2. Materials and Experimental Procedure

A medium carbon commercial grade DIN1.5025 sheet steel of 1 mm thickness having chemical composition of Fe-0.529C-1.67 Si-0.72Mn-0.12Cr (in wt.%) was chosen for this research work. Though a high silicon content (1.67%) is preferred due to its effectiveness in preventing (or at least delaying) carbide precipitation and hence promoting carbon partitioning to the untransformed austenite in order to partly or fully stabilize RA down to RT, the Mn content (0.72%) was somewhat lower than desired to promote enhanced austenite stabilization. Determination of critical temperatures, i.e., the start (A_c1) and end (A_c3) of austenite formation during reheating and the M_s, was considered essential in order to be able to design suitable heat treatment cycles. Consequently, the critical temperatures including A_c1 (765°C), A_c3 (835°C) as well as M_s (275°C) were determined through dilatometer measurements in the Gleeble simulator at the heating and cooling rates of 0.3°C/s and 50°C/s, respectively. Accordingly, the heat treatment cycles were designed by combining Q&B and also usual single step Q&P cycles to achieve multiphase microstructures containing various percentages of bainitic ferrite, martensite, and RA. The proposed heat treated samples were cut with the dimensions of 30×30×1 mm and were first normalized at 900°C for 5 min to achieve a more homogenized and also refined starting microstructure in the samples. The heat treatment cycles involved reaustenitizing at 900°C for 5 min, quenching to 350 and 250°C (temperatures above and below the M_s temperature) in a salt bath for isothermal holding between 5 s to 3600 s followed by water cooling to RT, as shown schematically in Fig. 1. The heat treated samples were mounted, ground, and polished according to ASTM E 3 by using a 3D laser scanning confocal microscope (LSCM; model Keyence VK-X200). Select specimens were subjected to detailed metallography in a field emission scanning electron microscope (FESEM; Zeiss Ultra Plus) equipped with the electron backscatter diffraction (EBSD) technique. The volume fraction and mean carbon content of RA were measured by X-ray diffraction (XRD: Rigaku SmartLab 9 kW equipment) using Co-K_α radiation operated at 135 mA and 40 kV within the 2θ interval angles of 45 to 130° at a step scan size 7.2°/min. The Rietveld WPPF (whole powder pattern fitting) analysis was used to determine the volume fraction of RA. The volume fraction and also the lattice parameter of RA were measured by using a direct comparison approach, comparing the integrated intensities of (111), (200), (220), and (311) of FCC (face-centred cubic) diffracted planes with (101), (002), (112), and (202) of BCC (base-centred cubic) planes, respectively. Also, the carbon content was determined by using the relationship:²⁴⁾   

$$\begin{array}{l}{\mathrm{a}}_{\gamma }{[}^{\mathrm{o}}\mathrm{A}]=3.572+0.033{\mathrm{x}}_{\mathrm{C}}+0.0012{\mathrm{x}}_{\mathrm{M}\mathrm{n}}+0.0056{\mathrm{x}}_{\mathrm{A}\mathrm{l}}\\ \hspace{1em}\hspace{1em}\hspace{1em}+0.00157{\mathrm{x}}_{\mathrm{S}\mathrm{i}}\end{array}$$(1)

Fig. 1.

A schematic of quenching and bainitic holding (Q&B) and usual single step quenching and partitioning (Q&P) heat treatment cycles.

Where a_γ is the lattice parameter of the austenite in nanometers and x_C, x_Mn, x_Al, x_Si are the concentrations in weight percent of carbon, manganese, aluminum, and silicon, respectively. The tensile test specimens were prepared according to the ASTM E8M standard. Tensile tests, three each of Q&B/Q&P experiments, were carried out using a SANTAM tensile testing machine model STM-150 at a constant crosshead speed of 10 mm/min and the corresponding engineering strain rate is about 0.0033 s⁻¹. The values reported are the averaged data of each heat-treated condition. Also, Vickers hardness measurements were made by using a 30 kg load and the time of loading was 15 s for each measurement.

3. Results and Discussion

3.1. Phase Transformation

For better understanding of the mechanisms of phase transformation during the Q&B and Q&P treatments, a detailed microstructural investigation was performed on the samples held isothermally for different durations. Typical laser scanning micrographs of the Q&B samples isothermally held at 350°C in the bainitic region for different times are shown in Fig. 2. The presence of few bainitic laths (dark) indicates that the formation of bainite has just begun at about 5 s (Fig. 2(a)), as the microstructure is essentially martensitic (bright region marked in the figure). Following holding for 30 s (Fig. 2(b)), a significant fraction of bainitic laths (dark) is revealed in the microstructure distributed throughout the matrix (bright), which is essentially a combination of FM and RA, marked by arrows as FM/RA. As the transformation continues with time, extensive bainite formation occurs (Fig. 2(c), 600 s) and is nearly complete in about an hour (Fig. 2(d), 3600 s). During the formation of bainite, carbon is rejected and partitioned from bainitic ferrite into the associated untransformed austenite areas. As a result, the prior austenite gets enriched with carbon and its M_s temperature decreases consequently. On subsequent quenching to RT, a fraction of RA can be expected to be stabilized in the microstructures. Formation of carbides, if any, was not discernable from the LSM micrographs. Though it is difficult to detect RA using laser scanning microscopy, discrete islands of FM/RA microphases can still be identified, as marked in Fig. 2, the volume fraction of FM/RA islands has been estimated to be about 18.2% using image analysis method in the sample held for 600 s (Fig. 2(c)), though it is difficult to distinguish the microstructure of this sample from the one isothermally held for 3600 s (Fig. 2(d)). In fact, as the holding time was increased to 3600 s, more bainitic ferrite plates formed, thus reaching about 90 vol.% of bainite in conjunction with about 10 vol.% of FM/RA islands according to the image analysis report. It has been earlier reported²⁵⁾ that holding for longer durations beyond about 600 s facilitated the formation of carbides, thus resulting in more bainite formation. This also suggests that high silicon in the steel can delay the carbide formation and/or growth, but cannot prevent it during longer holding.

Fig. 2.

Laser scanning micrographs recorded on Q&B treated samples held at 350°C in a salt bath for: (a) 5 s, (b) 30 s, (c) 600 s, and (d) 3600 s. The symbols of B, FM, and FM/RA stand for bainite, fresh martensite, and fresh martensite/retained austenite microconstituents.

Typical laser scanning micrographs of the samples isothermally held for different times at 250°C, i.e. below the M_s temperature (M_s≈ 275°C), are shown in Fig. 3. As regards holding in the Q&P regime, the structure is essentially martensitic for the sample held for 5 s at 250°C (Fig. 3(a)), including ~20% TM (dark features in the figure), in accord with Koistinen Marburger (K-M) equation²⁶⁾ and ~80% untempered high-carbon FM formed during final cooling to RT with the possibility of a retaining a small fraction of austenite. However, it is not possible to detect RA in this microstructure. By further increasing the isothermal holding to 50 s (Fig. 3(b)), the microstructure is essentially similar to that of the sample held for 5 s, as only TM is discernible in the image. Further holding up to 600 s revealed formation of a small fraction of bainite with fine lath morphology (Fig. 3(c)) and can be distinguished as gray laths from the TM (dark gray). In addition, some FM/RA regions too appeared as bright islands in the microstructure (Fig. 3(c)). Further holding led to the formation of more bainite and at about 3600 s, transformation to bainite seems nearly complete (gray regions) along with the presence of TM and also small islands of FM/RA constituents (white regions), as displayed in Fig. 3(d). Formation of bainite below the M_s temperature is well documented in the literature and, in some cases, in connection with the formation of ultrafine, or nanostructured, bainite with fine interlath austenite.^27,28,29) Bainite formation is accelerated due to the formation of athermal martensite during cooling to 250°C. However, at this magnification, it is not possible to investigate the fine structural details, and hence, other metallographic techniques including FESEM combined with EBSD were used for the purpose.

Fig. 3.

Laser scanning micrographs recorded on Q&P treated samples held at 250°C in a salt bath for: (a) 5 s, (b) 50 s, (c) 600 s, and (d) 3600 s. The symbols of B, TM, and FM/RA stand for bainite, tempered martensite, and fresh martensite/retained austenite microconstituents.

The quantification of the phase fractions in Q&B and Q&P treated samples was performed by using the dilatation results published elsewhere.²⁴⁾ Also, the volume fractions of FM, achieved from the decomposition of a part of carbon-enriched austenite during the final cooling, were estimated from the dilatation data according to the procedure described in a previous study.¹⁴⁾ In Q&P samples partitioned at 250°C, the bainite (B) fractions were estimated based on the volume fractions of the transformed phases, i.e., TM, FM (both obtained from dilatometry), and RA stabilized at RT (measured through XRD). In contrast, for the Q&B samples isothermally held at 350°C, bainite (B) fractions were estimated from the dilatation measurements for different holding times. For this heat treatment condition, the FM fraction was calculated based on the estimated volume fractions of B (obtained from dilatometer measurements), and RA (measured through XRD).

Figure 4 show the evolution of various phase fractions as a function of isothermal holding time following the Q&B and Q&P heat treatment schedules at 350 and 250°C, respectively. As can be discerned, an estimate of various phase fractions for each heat treatment condition can be extracted from these figures. As expected, the volume fraction of TM is constant at both the temperatures, since it depends on the undercooling below the M_s temperature in the case of Q&P treatment. The volume fraction of B increases with holding time for both the Q&B (350°C) and Q&P (250°C) conditions, and correspondingly, a lower volume fraction of FM forms from the carbon-enriched, untransformed austenite during the final cooling to RT. The final fractions of RA stabilized at RT increases initially with holding time to a maximum value of about 18% realized in 200 and 600 s at 350 and 250°C, respectively, beyond which the RA dropped to about 6.9 and 7.8%, respectively, following holding for 3600 s.

Fig. 4.

Variation of various phase fractions as a function of holding time, following isothermal heat treatments at (a) 350°C (Q&B), and (b) 250°C (Q&P), for different times between 5 to 3600 s. The abbreviations for tempered martensite, bainite, fresh martensite, and retained austenite microconstituents are TM, B, FM, and RA, respectively.

Figure 5 shows selected backscattered electron (BSE) micrographs recorded on Q&B samples isothermally held at 350°C. Figure 5(a) shows the representative microstructure of samples held only for 5 s at 350°C, revealing characteristic features of multivariant packets of martensitic crystals. A further increase in isothermal holding to 200 s revealed appreciable bainitic laths with characteristic sheaf-like features, while FM/RA phase constituents were difficult to distinguish and appeared as featureless dispersed islands. Neither was the RA distinguishable from FM nor was it possible to measure its volume fraction from BSE micrographs (Fig. 5(b)) owing to their nearly similar chemical composition, though a good fraction of RA is present also as interlath films between the bainitic and/or martensitic laths. Figure 5(c) presents a representative electron micrograph of the samples isothermally held at 350°C for a long duration of 3600 s and the structure is essentially comprised of FM/RA islands in the bainitic ferrite matrix.

Fig. 5.

The backscattered electron micrographs recorded on Q&B samples isothermally held at 350°C for: (a) 5 s, (b) 200 s, and (c) 3600 s.

The corresponding representative BSE micrographs recorded on the Q&P samples isothermally held at 250°C are shown in Fig. 6. A close observation of these images reveals that the prior austenite grain boundaries are easily recognizable because of the nature of the martensitic and bainite transformations largely nucleating at the prior austenite boundaries. On the other hand, the packets of bainite and TM laths are recognizable, as they appear as somewhat dark and light gray areas, respectively. The bright RA with interlath film-like morphology between the bainitic or martensitic laths is decipherable in some micrographs (Fig. 6(c)), besides the presence of whitish FM/RA grains (Figs. 6(b) and 6(c)). The bainite transformation does not start in a short holding time of 5 s, as also seen in the laser scanning micrograph (Fig. 3(a)) and hence, the microstructure essentially consists of TM and FM (Fig. 6(a)). With further holding up to 200 s, a small fraction of bainite can be seen in the microstructure (Fig. 6(b)), distinguishable from TM as described above. The bainitic transformation of balance austenite is practically complete following longer holding of 3600 s at 250°C (Fig. 6(c)) and the microstructure essentially consists of highly acicular bainitic ferrite plates with interlath austenite along with TM and FM/RA, as also noticed in laser scanning microscopy.

Fig. 6.

The backscattered electron micrographs recorded on Q&P samples isothermally held at 250°C for: (a) 5 s, (b) 200 s, and (c) 3600 s.

Typical EBSD images (a combination of image quality (IQ) with phase map (PM)) of the Q&B and Q&P samples isothermally held at 350 and 250°C are shown in Fig. 7. These EBSD figures can provide better information about the location, distribution, and morphological features of RA in the microstructures. Samples held at 350°C for 200 s (Fig. 7(a)) showed a microstructure comprising of bainite (red color), FM (dark red color) and RA (green color) appearing in both blocky as well as interlath morphologies; revealed clearly at a higher magnification, see Fig. 7(a). The mechanical stability of RA is an important aspect of AHSS materials. Factors that affect the austenite stability are the size and the morphology of the RA grains, with the property of the surrounding matrix. Therefore, an investigation was carried out on the transformation behavior of both the thin interlath film as well as blocky shapes of the RA phase, as achieved in the samples isothermally held for 200 s at 350°C (Fig. 7(a)). By increasing the isothermal holding to 3600 s (Fig. 7(b)) at 350°C, RA was stabilized in both blocky and interlath film-like morphologies similarly as seen in the microstructure of sample held for 200 s and both martensite and bainite are present as the surrounding phase of RA phase. On decreasing the isothermal temperature to 250°C (Q&P heat treatment condition) and holding for 600 and 3600 s (Figs. 7(c) and 7(d)), the EBSD images show multiphase microstructures including bainite, martensite, and RA (in both blocky and interlath film like morphologies but with a finer distribution in comparison to that in Q&B heat treated samples (Figs. 7(a) and 7(b)). These figures also indicate that the bainitic areas have appeared as bright red regions due to the higher confidence indexing of bainitic ferrites. In contrast, fresh martensitic areas have poor confidence index than the bainitic regions due to the higher intensity of internal strains and dislocations. As a consequence, the bainitic ferrites are generally much better indexed in EBSD images than the martensitic ones and hence the martensitic areas appear as dark contrasting regions in image quality (IQ) maps. While some blocks of fine RA phase (green) can be easily identified in samples held in the bainitic regime at 350°C (Figs. 7(a) and 7(b)), such grains are very few and much smaller in the case of the Q&P samples held at 250°C well below the M_s temperature and can be hardly seen. This suggests that most of the RA grains are finer than the resolution limit of the EBSD (about 0.1 μm) and a large fraction could be present as finely divided interlath films.

Fig. 7.

EBSD micrographs of samples isothermally held at 350°C for: (a) 200 s, and (b) 3600 s, and at 250°C for: (c) 600 s, and (d) 3600 s. The RA is green, while the bainite and FM phase constituents are bright red and dark red colored, respectively.

3.2. Retained Austenite and its Carbon Content

As revealed in the microstructural analysis using the LSCM and FESEM-EBSD, it is not possible to detect a large fraction of finely divided austenite, particularly the interlath films because of the limits of resolution with these techniques (about 0.08 – 0.1 μm). However, RA forms an important phase in these multiphase steels and hence, its bulk volume fractions in different Q&B and Q&P treated samples was measured by the XRD technique. Since the temperatures of heat treatments are relatively low (250 and 350°C) and the durations of heat treatments are also short (3600 s max), it is unlikely that there would have been any significant movement and partitioning of substitutional elements (such as Si, Mn, etc.). Therefore, the RA content stabilized down to RT, as estimated by XRD technique, is enriched with carbon that influences the lattice parameters of FCC. The average carbon content of RA, therefore, can be estimated based on the lattice parameter calculations and using empirical equations.²⁴⁾ The estimated volume fractions of the RA and their average carbon contents obtained for different Q&B and Q&P samples are plotted against the isothermal holding times, as depicted in Fig. 8. As can be discerned from the figure, the RA fractions initially increased with the holding time, peaking at about 18% after 200 s holding at 350°C above the M_s temperature, while further isothermal holding for longer durations up to 3600 s resulted in a steep drop in RA content to ~7%. The corresponding carbon enrichment, however, showed markedly different behavior, as the carbon content increased with holding time up to 3600 s in these Q&B heat treated samples. The carbon content of the RA phase increased sharply up to about 200 s, beyond which it reached a plateau at about 600 s (1.36%) with the carbon content remaining practically very close reaching a peak value (1.39%) at 3600 s (Fig. 8), though the RA fractions dropped from ~13% at 600 s to ~7% at 3600 s, suggesting loss of carbon due to the formation of carbides. According to these results, partitioning of carbon from bainite to RA occurred all the time up to 3600 s, though the occurrence of carbide precipitation in parallel competed with the bainite transformation and carbon partitioning processes.

Fig. 8.

Average RA fractions and corresponding carbon contents of Q&B (T_Q = 350°C) and Q&P (T_Q = 250°C) specimens isothermally held for different durations.

On the other hand, the Q&P samples isothermally held at 250°C showed somewhat different behaviour in respect of the volume fractions of RA stabilized down to and corresponding average carbon contents. In Q&P samples, the volume fraction of RA increased slowly up to 50 s (4.3–5.7%) similarly as in Q&B condition, beyond which there was a sharp increase in RA content at least beyond about 200 s (~8%) reaching about 17.8% at 600 s and then decreased sharply to 7.8 vol.% at 3600 s (Fig. 8). The drop in RA content beyond about 600 s seems to be due to the formation of carbides, even though the average carbon content of the austenite phase seems to be relatively low (0.90%). Correspondingly, the increase in the carbon content of the RA phase also showed three stages: 0.66–0.75% from 5 to 50 s, 0.75–0.90% from 50 to 600 s, and 0.90–1.20% from 600 to 3600 s (Fig. 8). Whereas the first stage appears to be mainly due to carbon partitioning and isothermal martensite formation (due to stress relaxation in austenite),³⁰⁾ the second stage reveals extensive bainite formation below the M_s temperature, as also revealed in the microstructures (Figs. 3 and 6) and hence, leads to extensive partitioning of carbon to the adjacent untransformed austenite. The third stage marks the sharp drop in RA due to the loss of carbon owing to the formation of carbides, as also described elsewhere.²⁵⁾ Also, in these Q&P samples the main carbon content in the RA phase increased up to 3600 s (1.21%) with an accelerated rate beyond 600 s (0.91%), though the holding time scale is logarithmic). But the amount of carbon content of RA in the Q&P heat treated samples is relatively less than that of the RA phase for Q&B heat treated samples. This behaviour could be due to a lower carbon partitioning rate at 250°C from martensite and/or bainite to the untransformed austenite, though the RA fractions in Q&P samples are somewhat higher beyond 600 s holding time compared to those of Q&B samples.

3.3. Mechanical Behavior during Tensile Testing

Typical engineering stress-strain curves of samples isothermally held for 3600 s at 350 (Q&B) and 250°C (Q&P) are shown in Fig. 9. As can be seen, both heat treated samples are associated with continuous yielding behaviour up to the fracture strength. This tensile behaviour has been attributed to the presence of unpinned dislocations generated during the formation of highly dislocated martensite from RA.^3,31) These unpinned dislocations are assumed to be mobile in the early stage of plastic deformation,³²⁾ thus causing continuous yielding behaviour. Also, another aspect of tensile test results (Fig. 9) showed that the tensile strength (TS) in Q&P specimens was higher in comparison to Q&B samples. This can be attributed to the fine division of phase constituents in Q&P samples (Figs. 7(c) and 7(d)), leading to an extensive strengthening in these specimens, which is in agreement with a previous study.⁵⁾ Formation of a small fraction of martensite prior to the isothermal holding below the M_s temperature leads to an increased number of nucleation sites, thus extensively refining the bainite packet and lath sizes, besides accelerating the kinetics of bainite transformation.¹⁵⁾ Also, in Q&P condition, due to the presence of a high volume fraction of martensite (both TM and FM), the tensile strength is expected to be higher in comparison to that in the Q&B condition with bainite as the main phase constituent in the microstructures (Fig. 4).

Fig. 9.

Typical engineering stress-strain curves of the tensile tested samples isothermally held for 3600 s at 350°C (Q&B) and 250°C (Q&P).

The mechanical response in respect of tensile properties, and the hardness data are listed in Table 1 and also plotted against isothermal holding time in Fig. 10. It is expected that the tensile strength is closely related to the hardness data, as can also be discerned from Table 1.³²⁾ According to the results summarized in Table 1, in both Q&B and Q&P heat treatment conditions, TS decreased with the increasing isothermal holding from 5 s to 3600 s, though the drop in TS is only marginal at 250°C (2545 to 2370 MPa) compared to that at 350°C (2451 to 1750 MPa), which is also depicted in Fig. 10(a). In Q&P condition, tempering of IM with increase in isothermal holding time results in only a marginal drop in tensile strength, as the continued formation of a significant volume fraction of bainite keeps the tensile strength relatively high. In comparison, an enhanced bainite volume fraction in the case of Q&B samples is the most important parameter in determining the tensile strength, as for shorter isothermal holding times, some FM realized during final cooling contributes to relatively higher tensile strengths (Fig. 4). On the other hand, YS (0.2% proof stress) initially decreased in the samples with a consequent increase in the volume fraction of RA (1050 MPa at 200 s and 1230 MPa at 600 s in Q&B and Q&P conditions, respectively). Longer holding times up to 3600 s resulted in an increase in YS up to 1350 and 1670 MPa for Q&B and Q&P conditions, respectively, Fig. 10(a). At 350°C, the hardness drops from 728 HV at 5 s to about 488 HV in 200 s, presumably due to extensive bainite formation and high RA stabilized at RT. Beyond 200 s holding, there is a slight increase in hardness presumably due to more bainite formation and less RA stabilized in the samples, and a small increase in hardness due to carbide formation. Similarly, at 250°C, there is a practically constant hardness in the range 722–730 HV up to about 200 s, beyond which there is a steep drop in the hardness, showing a minimum at 600 s (525 HV30), Table 1. This corresponds to extensive bainite formation below M_s at this point and thus the highest RA fraction (17.8%). There is a small increase in hardness (600 HV; Table 1) beyond this point due to carbide formation facilitating further decomposition of unstable austenite to lower bainite, thus leading to less RA (7.8%) in the steel, Fig. 10(a). Referring to Table 1 and Fig. 10(a), for the Q&B samples isothermally held at 350°C, both the total elongation (TE) and uniform elongation (UE) increased initially up to 200 s isothermal holding as a consequence of highest fraction of RA in this condition, beyond which both TE and UE dropped gradually. On the other hand, for samples held isothermally at 250°C, TE increased with the increase in holding time up to 600 s, beyond which it dropped marginally. However, unlike the TE, UE at 250°C did not show any drop during the entire period of holding up to 3600 s, Fig. 10(a).

Table 1. Tensile properties obtained on Q&B and Q&P samples along with the corresponding hardness data.

Steel	TS (MPa)	YS (MPa)	UE (%)	EL (%)	TS*EL (GPa%)	YS/TS	Hardness (HV30 kg)
350C-5s	2451	1450	4.8	4.8	11.80	0.6	728
350C-200s	2175	1050	8.9	17.9	38.94	0.48	488
350C-600s	1935	1350	7.2	15.7	30.40	0.70	500
350C-1h	1755	1350	5.1	15	26.30	0.77	530
250C-5s	2545	1395	4.4	4.4	11.20	0.55	722
250C-200s	2470	1275	7.6	7.6	18.80	0.52	730
250C-600s	2390	1230	8	11.9	28.50	0.52	525
250C-1h	2327	1670	8.5	9.6	22.40	0.72	600

(Abbreviations: TS = tensile strength, YS = yield strength, UE = uniform elongation, TE = total elongation, HV = Vickers hardness)

Fig. 10.

Mechanical response exhibited by the Q&B and Q&P heat treated samples as a function of isothermal holding time. The abbreviations TS, YS, EL, and UE stand for tensile strength, yield strength, total elongation, and uniform elongation, respectively.

The results of another parameter TS*EL, an important property for automotive sheet applications, is shown both in Table 1 and also in Fig. 10(b) as a function of isothermal holding time. Accordingly, the best properties in respect of formability in Q&P specimens is related to the samples partitioned for 600 s at 250°C (28.50 GPa%) with the microstructure characterized by a fine division of TM, bainite and RA phase constituents (Fig. 7(c)). On the other hand, in Q&P condition, a higher volume fraction of RA phase (~18%) is achieved in the microstucture (Fig. 7(c)) of the sample isothermally held for 600 s at 250°C which transformed to martensite during tensile straining and improved the tensile strength and elongation by TRIP effect. Two different EBSD images of this sample following tensile testing are shown in Fig. 11 with the measurements made close to the fractured ends. A comparision of the EBSD images after and before tensile straining, presented in Figs. 11 and 7(c), respectively, clearly revealed that most of the RA phase (both blocky and thin film morphologies) transformed to martensite during tensile straining and only a small amount of RA (green region) is seen in some locations (Fig. 11). It should be noticed that the XRD did not reveal any discernible RA phase in the fractured samples, which may be well below the detection limit of XRD (1–2%) and needs further study by TEM. Anyhow, limited EBSD measurments made on the tensile tested sample held for 600 s at 250°C, suitably exemplified the strain induced transformation of the RA phase following tensile deformation.

Fig. 11.

EBSD micrographs of tensile tested samples isothermally held at 250°C for 600 s. The RA is green, while the bainite and FM phase constituents are bright red and dark red colored, respectively.

However, a study of TS*EL for the Q&B samples showed that the best behaviour is achieved in the samples isothermally held for 200 s (38.94 GPa%), which comprises a multiphase microstructure including a high fraction of refined lower bainite and FM/RA islands (as shown in Fig. 7(a)). Hence, the presence of a relatively softer RA phase adjacent to the FM and bainite laths in these samples inhibits crack propagation and growth due to the transformation induced plasticity effect (TRIP) occurring during the tensile deformation, and therefore resulting in high tensile strength with good elongation. RA phase in this sample (Fig. 7(a)) has both thin film-like and blocky morphologies with different mechanical stabilities and hence, it transforms to martensite over a large strain, thus attributing improved mechanical properties.

Another parameter shown in Fig. 10(b) and also Table 1 is the yield ratio (YS/TS), which is an important parameter in the evaluation of the strain hardening potential. A lower YS/TS ratio means a higher strain hardening capacity in the material. The ratio varies in the range 0.48 to 0.77 for the samples held at 350°C and the corresponding ratios for the samples held at 250°C is marginally different in the range 0.52 to 0.72, which is in agreement with the previous studies.^5,10,33) In order to understand the behavior, two microstructural aspects viz., the presence of martensite as well as the mechanical stability of the RA phase have been considered. The presence of TM phase in Q&P samples will have two opposite effects. While the presence of TM in Q&P samples increased the YS compared to just bainitic ferrite, it is still softer than the FM that may form in the final quenching, owing to its lower carbon content and reduced dislocation density. Mechanical stability of RA is another effective parameter that is influenced by various factors including chemical composition, size, morphology as well as the strength of its surrounding phases.^34,35) The microstructural study revealed the presence of a relatively higher fraction of blocks or islands of RA phase in Q&B samples (Figs. 7(a) and 7(b)) compared to the Q&P samples (Figs. 7(c) and 7(d)), thus influencing the strain hardening behaviour during tensile straining, Fig. 9.

3.4. Fracture Behaviour

Figures 12 and 13 show fractographic features of various Q&B and Q&P heat treated samples following tensile testing at RT. It can be seen from Figs. 12(a) and 13(a) that the samples isothermally held for 5 s at 350 and 250°C, respectively, showed surface features typical of brittle fracture and at higher magnification, revealed finer details in respect of the occurrence of cleavage facets, cracks, and voids of varying sizes. Obviously, this has concurrence with the formation of large fractions of FM in both the samples, as also revealed in the microstructures (Figs. 5 and 6) and corroborated by high yield and tensile strengths and low ductilities (Table 1). However, both shallow dimples and cleavage facets typical of quasi-cleavage fracture, were observed in the fractographs of the samples held for 200 and 600 s at 350 and 250°C, as shown in Figs. 12(c) and 13(c), respectively. By further increasing the isothermal holding time to 3600 s for the Q&P and Q&B samples, the formation of dimples marking the occurrence of microvoid coalescence becomes distinctly apparent, as shown in Figs. 12(e) and 13(e), respectively. A previous study^36,37) in this regard showed that the blocky RA exhibited lower mechanical stability, especially in the centre of the grains due to a low carbon concentration, obviously as a consequence of slower diffusion rate in FCC austenite. Therefore, it is considered that the blocky RA is prone to strain-induced martensitic transformation (TRIP effect) at a relatively small plastic strain. Besides, due to the mechanical heterogeneity (strength mismatch) between the FM and blocky RA, cracks are more likely to initiate under tensile loading preferentially at the interfaces of the martensite and RA, and eventually propagate along these interfaces,³⁸⁾ but the strain-induced transformation of RA can blunt or deviate the cracks as well and will determine the ductility accordingly. Referring to Figs. 7(a) and 7(b), for Q&B treated samples, the RA is present as both blocky grains as well as thin interlath films depending on the isothermal holding times. It, therefore, transformed to martensite over a large strain range (Table 1) and helped improve the ductility and also, revealed more dimples (ductile fracture) in the fracture surface.

Fig. 12.

SEM fractographs of the tensile tested Q&B samples heat treated at 350°C for different isothermal holding times: ((a), (b)) 5 s, ((c), (d)) 200 s, and ((e), (f)) 3600 s.

Fig. 13.

SEM fractographs of the tensile tested Q&P samples heat treated at 250°C for different isothermal holding times: ((a), (b)) 5 s, ((c), (d)) 600 s, and ((e), (f)) 3600 s.

4. Conclusions

The constitutive relationships between the microstructures and mechanical properties of different multiphase microstructures obtained through the quenching and isothermal holding for different times at temperatures above (350°C- Q&B) and below (250°C- Q&P) the M_s temperature in a medium C DIN 1.5025 sheet steel (Fe-0.529C-1.670Si-0.721Mn-0.120Cr) have been analyzed. The main conclusions of the study are the following:

(1) A detailed microstructural investigation revealed that in both the Q&B (350°C) and Q&P (250°C) samples, isothermally held above (350°C) and below (250°C) the M_s temperature, respectively, multiphase microstructures comprising bainite- martensite- retained austenite were achieved with different fractions of these phase constituents, with or without carbide formation.

(2) In the Q&B samples isothermally held above the M_s temperature (350°C), the bainite content in the microstructure of the specimens increased with increasing isothermal holding time up to 3600 s, though the RA content decreased to about 7% as a consequence of carbide formation beyond about 600 s of isothermal holding.

(3) Microstructural analysis of Q&P samples heat treated at 250°C showed a fine division of phase constituents, particularly the RA (mostly as films) than the Q&B heat treated samples at 350°C, though carbide formation occurred beyond about 600 s of isothermal holding leading to a reduced RA fraction (7.8%) in the samples held at 3600 s, for instance.

(4) The study of tensile behavior revealed that the best properties in respect of TS*EL were obtained for the Q&B and Q&P samples isothermally held for 200 and 600 s at 350 and 250°C, respectively.

(5) The RA phase with both blocky and interlath film-like morphologies in these differently heat treated Q&B and Q&P samples has a considerable effect on the stress-strain behaviour. In particular, the samples with a high volume fraction of RA displayed high elongations due to the improved strain hardening capacity over a wide strain mainly due to TRIP effect.

Acknowledgments

The funding of this research activity under the auspices of the Genome of Steel (Profi3) by the Academy of Finland through project #311934 is gratefully acknowledged. S. Pashangeh expresses her gratitude to the Ministry of Science Research and Technology in Iran for funding a research visit to the University of Oulu, Finland to conduct this research work. Authors would like to express their gratitude to Dr. Sumit Ghosh and Mr. Tun Tun Nyo for extending help in conducting a part of the metallographic work.

References

• 1)   P.  Huyghe,  L.  Malet,  M.  Caruso,  C.  Georges and  S.  Godet: Mater. Sci. Eng. A, 701 (2017), 254.
• 2)   N.  Tsuchida,  S.  Okura,  T.  Tanaka and  Y.  Toji: ISIJ Int., 58 (2018), 978.
• 3)   S.  Yan,  X.  Liu,  W. J.  Liu,  T.  Liang,  B.  Zhang,  L.  Liu and  Y.  Zhao: Mater. Sci. Eng. A, 684 (2017), 261.
• 4)   R. C.  Giacomin and  B. A.  Webler: ISIJ Int., 59 (2019), 858.
• 5)   A.  Navarro-López,  J.  Hidalgo,  J.  Sietsma and  M. J.  Santofimia: Mater. Sci. Eng. A, 735 (2018), 343.
• 6)   E. P.  Da Silva,  D.  De Knijf,  W.  Xu,  C.  Föjer,  Y.  Houbaert,  J.  Sietsma and  R.  Petrov: Mater. Sci. Technol., 31 (2015), 808.
• 7)   S. M. C.  Van Bohemen,  M. J.  Santofimia and  J.  Sietsma: Scr. Mater., 58 (2008), 488.
• 8)   I. A.  Yakubtsov and  G. R.  Purdy: Metall. Mater. Trans. A, 43 (2012), 437.
• 9)   J.-C.  Hell,  M.  Dehmas,  S.  Allain,  J. M.  Prado,  A.  Hazotte and  J.-P.  Chateau: ISIJ Int., 51 (2011), 1724.
• 10)   X. Y.  Long,  J.  Kang,  B.  Lv and  F. C.  Zhang: Mater. Des., 64 (2014), 237.
• 11)   J.  Sun,  H.  Yu,  S.  Wang and  Y.  Fan: Mater. Sci. Eng. A, 596 (2014), 89.
• 12)   R. T.  Howard and  M.  Cohen: Trans. AIME, 176 (1948), 401.
• 13)   H.  Kawata,  K.  Hayashi,  N.  Sugiura,  N.  Yoshinaga and  M.  Takahashi: Mater. Sci. Forum, 638–642 (2010), 3307.
• 14)   L.  Zhao,  L.  Qian,  J.  Meng,  Q.  Zhou and  F.  Zhang: Scr. Mater., 112 (2016), 96.
• 15)   A.  Navarro-Lopez,  J.  Sietsma and  M. J.  Santofimia: Metall. Mater. Trans. A, 47 (2016), 1028.
• 16)   D.  Kim,  J. G.  Speer and  B. C.  De Cooman: Metall. Mater. Trans. A, 42 (2011), 1575.
• 17)   W. T.  Zhao,  X. F.  Huang and  W. G.  Huang: Mater. Sci. Technol., 32 (2016), 1374.
• 18)   J.  Feng,  T.  Frankenbach and  M.  Wettlaufer: Mater. Sci. Eng. A, 683 (2017), 110.
• 19)   J.  Tian,  G.  Xu,  M.  Zhou and  H.  Hu: Steel Res. Int., 89 (2018), 1700469.
• 20)   A.  Arbor: J. Heat Treat., 1 (1980), 24.
• 21)   L.  Zhao,  N. H.  Van Dijk,  E.  Bruck,  J.  Sietsma and  S.  van der Zwaag: Mater. Sci. Eng. A, 313 (2001), 145.
• 22)   A.  Mostafapour,  A.  Ebrahimpour and  T.  Saeid: Int. J. Iron Steel Soc. Iran, 13 (2016), 1.
• 23)   S.  Pashangeh,  H. R.  Karimi Zarchi, S. S. Ghasemi bandkooki and M. C. Somani: Metals, 9 (2019), 492.
• 24)   A. K.  Behera and  G. B.  Olson: JOM, 71 (2019), 1375.
• 25)   S.  Pashangeh,  M. C.  Somani,  S. S. G.  Banadkouki,  H. R. K.  Zarchi,  P.  Kaikkonen and  D. A.  Porter: Mater. Charact., 162 (2020), 110224.
• 26)   S. M. C.  Van Bohemen and  J.  Sietsma: Metall. Mater. Trans. A, 40 (2009), 1059.
• 27)   H.  Bhadeshia: Proc. R. Soc. A, 466 (2010), 3.
• 28)   W.  Li,  H.  Gao,  Z.  Li,  H.  Nakashima,  S.  Hata and  W.  Tian: Int. J. Miner., Metall., Mater., 23 (2016), 303.
• 29)   B.  Avishan,  C.  Garcia-Mateo,  S.  Yazdani and  F. G.  Caballero: Mater. Charact., 81 (2013), 105.
• 30)   M.  Somani,  D.  Porter,  L.  Karjalainen and  D.  Misra: Metall. Mater. Trans. A, 45 (2014), 1247.
• 31)   A.  Kumar,  S. B.  Singh and  K. K.  Ray: Mater. Sci. Eng. A, 474 (2008), 270.
• 32)   S.  Pashangeh,  S. S. G.  Banadkouki and  M. C.  Somani: J. Mater. Res. Technol., 9 (2020), 5007.
• 33)   N.  Tsuchida,  S.  Okura,  T.  Tanaka and  Y.  Toji: ISIJ Int., 58 (2018), 978.
• 34)   Z. P.  Xiong,  A. A.  Saleh,  R. K. W.  Marceau,  A. S.  Taylor,  N. E.  Stanford,  A. G.  Kostryzhev and  E. V.  Pereloma: Acta Mater., 134 (2017), 1.
• 35)   P.  Huyghe,  L.  Malet,  M.  Caruso,  C.  Georges and  S.  Godet: Mater. Sci. Eng. A, 701 (2017), 254.
• 36)   S.  Yan,  X.  Liu,  W. J.  Liu,  T.  Liang,  B.  Zhang,  L.  Liu and  Y.  Zhao: Mater. Sci. Eng. A, 684 (2017), 261.
• 37)   J.  Hidalgo,  K. O.  Findley and  M. J.  Santofimia: Mater. Sci. Eng. A, 690 (2017), 337.
• 38)   L.  Zhao,  L.  Qian,  Q.  Zhou,  D.  Li,  T.  Wang,  Z.  Jia,  F.  Zhang and  J.  Meng: Mater. Des., 183 (2019), 108123.