2021 Volume 61 Issue 2 Pages 582-590
Bainite transformation and the resultant tensile properties of medium-carbon Si-bearing steels having upper bainite with retained austenite were clarified focusing on the effect of prior austenite grain size. Two different conditions (850°C and 1050°C) for austenitization were adopted to obtain the different prior-austenite grain sizes. The bainite structure was obtained by austempering, and the condition of the isothermal holding was decided according to the Time-Temperature-Transformation diagrams. The refinement of prior-austenite grains brought significant shortening of the incubation time of bainite transformation and the morphology of retained austenite grains occasionally appeared more blocky. Meanwhile, the mean grain size of the retained austenite measured by electron back scattering diffraction (EBSD) analysis did not change largely (~3 µm). The samples that were held isothermally at 400°C for 30 min showed fully bainite structure with much amount of retained austenite whose area fraction was ~40%, and these tensile tests of the both samples revealed high strength (1.4 GPa) with adequate ductility (more than 25%). Significant amount of retained austenite disappeared in the breaking samples and this indicates that the transformation induced plasticity (TRIP) occurs. The dependence of prior-austenite grain size on the deformation-induced martensitic transformation could not be found, while the orientation dependence was clearly detected in both the samples. According to these experimental results, the benefit of the refinement of the prior austenite grains is difficult to be discovered with the improvement of the tensile properties. However, it can be found in the shortening of the heat treatment process.
The high strength with large ductility is one of the most demanded properties of structural materials.1) The developing ways of steels with this preferable property can be roughly categorized in two patterns; (1) the usage of higher carbon martensite which has higher strength, and (2) the application of austenite with larger elongation. Since martensite provides high strength just as heat treated state, it is preferentially adopted in production processing such as hot stamping.2) Additionally, the recent study3) clarified that the refinement of prior-austenite grain size of the high carbon martensite steel brings the enhancement of elongation. This report indicates further possibility of the application of martensite. Concerning of austenite, on the other hand, high manganese steel4) has been being studied aggressively. The last trend for the study on austenitic steel goes to the metastable high-entropy alloy in which the chemical composition is optimized to maximize the benefit of transformation induced plasticity (TRIP)5,6,7,8) and/or twinning induced plasticity (TWIP) effects.1,9) These deformation-induced displacive changes are regarded useful to improve ductility.
The combination of these two ways is found in some production process to obtain the multi-phased microstructure consisting of both hard phase and retained austenite. One of these combination processes is the quenching and partitioning (QP) process,10,11) in which solute carbon is rearranged during the holding at the martensite-austenite mixture condition to stabilize austenite. While the QP process has been studied aggressively, previously developed ways for the combination have been studied as well.12,13) In these traditional ways, bainite transformation is utilized with Si-bearing steel to prevent the precipitation of cementite, so as to obtain the retained austenite preferentially.12,13)
One of the early works on the Si-bearing steels was reported by Sinoda et al.14) and this report exhibited that the austempering of a medium carbon steel (0.59wt%C-1.53%Si-0.89%Mn-0.02%Ni-0.14%Cr-Fe) at a temperature ranging from 300°C to 500°C brought the retained austenite in bainite with superior tensile properties. Similar experiments of the austempering were reported by Suzuki et al.15) (0.57%C-1.4%Si-0.75Mn-0.70%Cr-0.08%Ni; JIS-SUP12 and other spring steels) and Santos et al.16)[Santos 2009] (0.56%C-0.58%Mn-1.43%Si-0.47%Cr-Fe) at various cooling rate for the cooling after the austempering. In other works,17,18,19) the austempering was applied after the reheating at inter critical temperature to obtain the mixture of both the ferrite, bainite and the retained austenite. Although these works clarified the superior mechanical properties, the microstructures of these samples were not clarified adequately since these works were published before the spreading of the electron back scattering diffraction (EBSD) measurements. More recently, Caballero et al.20,21) proposed the heat treatment in which the austempering is applied at relatively lower temperature to refine the components of bainite to nano scale. This treatment is characterized by the isothermal holding for long period of time due to the transformation temperature lower than around 350°C. Garcia-Mateo et al.22) proposed the lower alloyed steel (0.98%C-2.9%Si-0.77%Mn-0.45%Cr-0.21%Cu-0.16%Ni-Fe) to obtain the higher strength by the austempering at the low temperature. Królicka et al.23) applied the low temperature austempering with medium carbon spring steel (0.57%C-1.89%Si-0.69%Mn-0.15%Cr-0.19%Ni-0.02%Mo-0.005%V-Fe, ISO 55Si7) with different austenite grain size and they found that the uniformity of prior-austenite grain size is effective to control the toughness. When the applicability of the heat treatment process is considered, the grain refinement of the prior-austenite grain size and the increasing of holding temperature should be desirable to reduce the isothermal holding time. However, the studies on the kinetics of bainitic transformation, microstructure and the tensile strength - ductility balance of the medium carbon Si-bearing steel by the austempering at higher temperature are limited, except for some papers.14,15,16,17,22)
The aim of this work is to clarify the effect of prior-austenite grain size on the bainite transformation, microstructure and mechanical properties of upper bainite with the retained austenite in the medium-carbon Si-bearing steel, so that the present results will indicate that the utilization of upper bainite of medium carbon steel is one of the possible choice for steel to improve the strength - ductility balance.
A medium-carbon Si-bearing steel (0.62%C-2.02%Si-0.23%Mn-1.01%Cr-Fe) whose chemical composition is designed based on a spring steel, SAE9254 with the additional bearing of Si to prevent from the carbide precipitation was examined. At the first stage of the experiments, the Time-Temperature-Transformation (TTT) diagram was examined by dilatometry to clarify the range of isothermal transformation temperature at which bainite can be obtained. For the dilatometry, the cylindric sample with a diameter of 8 mm and a height of 12 mm was placed in the vacuum chamber with an induction coil. Temperature of the sample was monitored with a thermo-couple welded on the center-side surface of the samples. The heat patterns examined are shown in Fig. 1(a). Two different temperatures (850°C and 1050°C) were adopted as austenitization temperature (Tγ) to prepare the samples with different mean diameters of the prior-austenite grains. After the asutenitization for 30 sec, the samples were cooled by N2 gas-flow to various holding temperatures and kept for various period of time ranging 100–2000 sec. During the isothermal holding, both the diameter and the height were monitored by two-dimensional length sensor and the volume change, ΔV was examined to determine apparent incubation time.24,25) Some samples were cooled down directly from the austenitization conditions without isothermal holding.
Heat treatments examined for (a) Time-Temperature-Transformation diagram and (b) sample preparation for tensile test.
With the TTT diagrams as shown later (Fig. 2), the heat patterns that were applied to the samples for tensile deformation tests were determined as shown in Fig. 1(b). The samples with a dimension of 14 mm × 14 mm × 150 mm were austenitized at 850°C or 1050°C with an electric furnace and held for 600 sec. Subsequently, the samples were soaked in a salt bath at 400°C, 450°C or 500°C and kept for 1800 sec. Some samples were cooled directly to an oil bath at 168°C from the austenitization conditions without isothermal holding. The heat treated samples were cut to the tensile test pieces with a gage length of 42 mm and a diameter of 6 mm. Tensile test was conducted at ambient temperature at a constant cross head speed of 0.85 mm/min (an initial strain rate of 0.33 × 10−4/sec).
(a) Volume change during isothermal holding and (b) the TTT diagram with austenitization at 850°C or 1050°C. (Online version in color.)
Microstructures were examined by a scanning electron microscope (SEM) with an EBSD measurement system. The observation surface was prepared by mechanical gliding and electro-chemical polishing with a solution of 10 vol% perchloric acid + 90 vol% acetic acid at an applied voltage of 30 V for a polishing time of 1 min at room temperature. In order to obtain clean surface, the polishing was paused every 10 second and cleaned up the polished surface by ethanol. SEM (JEOL 7000F) observation was conducted at a accelerated voltage of 15 kV and EBSD measurements (TSL OIM DataCollection) were done at a scanning pitch of 0.2 μm. BCC and FCC phases were set as the possible phases. Some of the samples were examined by X-ray diffractometry (XRD) (Rigaku MiniFlex) with a Cr-target source operated at 40 kV and 15 mA.
Figure 2 shows the results of the dilatometry. Some examples of the volume change during isothermal holding were shown in Fig. 2 (a). The relative volume change (ΔV/V0) is evaluated by the normalization of the examined change, ΔV by an initial volume, V0. At any condition of the austempering, the volume expansion occurs after the incubation period. When compared at the same holding temperature, the higher austenitization temperature brings the longer incubation period, while no change is found in the apparent maximum volume after the long isothermal holding for 1000 s. This indicates the refinement of prior-austenite grain size provides no significant change in the contents of transformed phases. The apparent incubation time was determined as the volume change reached to 0.1% and the data was plotted as the TTT diagram in Fig. 2(b). At both the austenitization temperatures, the starting curves of the transformations consist of double C-curves between which the bay was found around 520°C. The double C-curves in TTT diagram was found typically in some alloyed steels26,27) and the upper limit of the lower C-curve indicates bainite start temperature, Bs. At the transformation temperature below the Bs, the prior-austenite grain refinement makes the incubation time shorter with the higher transformation temperature. This temperature dependence is discussed at the final paragraph of this subsection.
In order to clarify the microstructural change by bainite transformation at the isothermal holding, the EBSD measurements were conducted with the samples held at 400°C for various periods. The results of the measurements at the austenitization temperature of 850°C and 1050°C were shown in Figs. 3 and 4, respectively. These figures show the spatial distributions of the BCC and the FCC phases separately, and the area fraction of each phases are denoted at the top-right corner of the corresponding maps. The black line indicates the prior-austenite grain boundary. The prior-austenite grain boundaries were determined with respect to the morphology of microstructure and the distribution of the orientation of BCC phase.28,29,30,31) The twin boundaries were also taken in account as prior-austenite grain boundaries. The grain growth during the austempering was hardly found and the mean prior-austenite grain sizes are 11 μm at Tγ = 850°C and 50 μm at Tγ = 1050°C. These grain sizes were determined by the intercept method with random lines.
Orientation color maps of BCC (a,b,c) or FCC (d,e,f) phases in the samples isothermal holding at 400°C for (a,d) 50 sec, (b,e) 100 sec or (e,f) 400 sec following after the austenitization at 850°C. The orientation parallel to the normal direction of the observation plane was illustrated by color key shown in the triangle at the top right of this figure. Black line shows the prior γ grain boundary. Area fractions of the component phases are denoted with each map. (Online version in color.)
Orientation color maps of BCC (a,b,c) or FCC (d,e,f) phases in the samples isothermal holding at 400°C for (a,d) 50 sec, (b,e) 100 sec or (e,f) 400 sec following after the austenitization at 1050°C. (Online version in color.)
In the sample austenitized at 850°C, both the BCC and the FCC grains with elongated shape, which was typically observed in the high magnification images attached at the top side in Fig. 3, was observed even at the shortest holding time (50 sec). The FCC grains in one of prior-austenite grains show the orientations very similar to each other. On the other hand, various orientations of the BCC phases were found and a prior-austenite grain consists of several blocks in one of which the BCC grains has the same orientations. This distribution is similar to a martensite in quenched carbon steels.7,28,29,30) These mixture of both BCC and FCC phases can be regarded as bainitic ferrite and retained austenite. Some part in the sample held for 50 sec has little FCC grains and it contains martensite structure, where the untransformed part of austenite changed to martensite during the final cooling. The area fraction of martensite appeared decreasing with increasing the holding time due to the progress of bainite transformation, and the retained austenite grains look to become somehow larger.
When the higher temperature was chosen as the austenitization temperature, bainite transformation was suppressed relatively as shown in Fig. 4. In the samples that were austenitized at 1050°C and austempered at 400°C for 50 sec (a,d) or 100 sec (b,e), the elongated retained austenite grains within the bainite was found at very limited areas while fine austenite grains with a diameter less than 1 μm was scattered at the area out of bainite. The area with fine austenite in the FCC phase map is well consistent with that showing lath martensite in the BCC map. When the holding time reaches to 400 sec (c,f), the elongated austenite grains was covered in whole the sample, indicating no martensite was evolved. These microstructural observations confirm the acceleration of bainite transformation by austenite grain refinement, which is well consistent with the dilatometric measurements shown in Fig. 2. When compared at the same holding time, the morphology of the bainite in the larger prior-austenite grains in Fig. 4(f) looked more elongated than those in the smaller prior-austenite grains in Fig. 3(f), although the size of these grains does not change largely. This difference has been similarly reported in the study on the low carbon steels by Sugimoto et al.32) and Królicka et al.23)
It was reported33) that the austenite grain size effect on the kinetics of bainite transformation varies depending on the chemical composition. According to the previous report,33) the austenite grain size effects were decided by the differences between the influence on nucleation and that on growth. When the austenite grain refinement is effective to increase the nucleation site and the growth rate is slow, the bainite transformation is accelerated by the grain refinement. The same situation should be occurred in this work, and the austenite grain refinement is useful to reduce the process time for the austempering. The kinetics of bainite transformation was attempted to be modeled with the assumption that the density of nucleation site can be decided by prior-austenite grain size.33) Considering the prior-austenite grain size in this work, the smaller grain size is about 1/2.5 of the larger grain size. This change of grain refinement brings the 2.5 times increase of the density of grain boundaries; whereas, as shown in Fig. 2(b) at around 400°C to 500°C, the incubation time of the sample with smaller grain size is around one-tenth smaller than that of the coarse grain size. In addition, the effect of austenite grain size on the incubation time of bainite transformation depends on the transformation temperature. The higher the transformation temperature is, the shorter the incubation time becomes. This implies that the change of the incubation time is governed not only by the grain size itself but also by the temperature dependence of driving force for transformation. One of the possible reasons can be discussed as follows: the austenite grain boundary energy is assumed one of the driving force for transformation besides the chemical free energy;34,35) however, the temperature dependence of the grain boundary energy should be smaller than that of the difference in free energies between the parent and the resultant phases at the relatively lower temperature than the melting point as theoretically demonstrated by Lee et al.35,36) Although the grain boundary energy is much smaller than the chemical free energy difference between the two phases, the relatively large grain boundary energy at higher temperature provides the change in energy barrier against which the nucleation overcomes by thermal activation. Consequently, the incubation time was presumably shortened significantly by the grain refinement at higher temperature.
3.2. Tensile Test and Austempering TemperatureAccording to the results related to the TTT diagrams shown before, the temperatures ranging from 400°C to 500°C were chosen as the austempering condition for the preparation of tensile test pieces. These temperatures are included in the range where the bainite transformation occurs firstly. The austempering for the tensile test pieces was conducted by box furnace so that the holding time, 30 min was different from the heat treatment for the TTT diagrams. Consequently, the experiments started from the further examinations of the microstructures before the tensile test.
Figure 5 shows the phase distribution maps of the samples austempered at various temperatures for 30 min. Red and green areas indicate the BCC and the FCC phases, respectively. The white line illustrates the prior-austenite grain boundaries. The mean prior austenite grain sizes of the samples austenitized at 850°C and 1050°C are 16 μm and 40 μm. Both of the 400°C austempered samples shows the fully bainite structures with the retained austenite. These characters of microstructure are similar to the results in Figs. 3 and 4 in the previous subsection. The retained austenite grain size was ~3 μm in both of the 400°C austempered samples when the diameter was measured by random line intercept method using the phase maps. Meanwhile, the samples austempered at 450°C or 500°C included the area where the retained austenite grain was absent. This area showed martensite which was evolved from both the austenite without any transformation and the austenite surrounded by bainitic ferrite but in which carbon was not concentrated. This is because the long incubation time provides large non-transformed area which changes martensite or pearlite.
Phase maps in the samples austenitized at 850°C (a,b,c) or 1050°C (d,e,f) followed by isothermal holding at 400°C (a,b), 450°C (b,e) or 500°C (c,f) for 1800 sec. The white line indicates the prior γ grain boundary. (Online version in color.)
The observation of secondary images in the SEM was conducted to examine especially the appearance of carbide, as shown in Fig. 6. The morphology of carbide was influenced mainly by the austempering temperature. The samples austempered at 400°C (a,b) include the bainitic ferrite with small number of carbides whose size is around 100 nm or less. The limited precipitation of carbide is due to the Si-bearing as mentioned in the introduction.12,13) The number of carbide becomes fewer in the samples austempered at 450°C (c,d) and some of these shows the morphology different from the 400°C austempered samples. This heterogeneous aspect agrees with that found in Fig. 5. The sample austempered at 500°C partially shows pearlite lamellar and lath martensite structure which evolved after the later starting of bainite transformation. This suggests that the temperature just below Bs is not preferable to obtain the fully bainite structure with uniform distribution of retained austenite.
SEM images of the samples austenitized at 850°C (a,b,c) or 1050°C (d,e,f) followed by isothermal holding at 400°C (a,b), 450°C (b,e) or 500°C (c,f) for 1800 sec.
Figure 7 shows the nominal stress – nominal strain curves of the samples austempered at various temperatures. The results of the oil quenched samples with martensite structure were shown by dotted lines for a reference. The strength of the samples austempered at 450°C and 500°C exhibits higher than 1.0 GPa but the elongations are limited to a few %. In these cases, the refinement of the austenite grain size contributes some improvement of elongation. The similar grain size effect has been already reported by Wang et al.3) with their study on the medium-carbon martensitic steel. The samples austempered at 400°C exhibit the high strength of 1.4 GPa with preferentially large elongations lager than 27%. The significant improvements of elongation can be found in the samples austempered at 400°C. However, a few % of elongation was extended by the refinement of prior-austenite grain size and less change of strength was found, so that it can be pointed out that the improvement of the tensile properties by the austenite grain refinement is limited as far as the present study is concerned.
Nominal stress - nominal strain curves of the sample austempered at various conditions. The dotted curves are the data of the quenched samples with martensite microstructure. (Online version in color.)
Concerning with the fracture appearance, all the samples were fractured without necking as it is suggested by the stress – strain curves without the stress reduction after the maximum of nominal stress. This means that the facture mechanism is different from ductile fracture. As shown in Fig. 8, the fracture surfaces of the oil quenched samples (a,e) shows coarse faceting with a kind of river pattern. The river patterns were also found in all the austempered samples and some of them includes shallow dimple patterns37) at limited areas. These morphologies imply brittle-like fracture due to the brittleness of the bainitic ferrite or martensite. When compared at the same austempering temperature, it can be found that the finer prior-austenite grain size provides the finer faceting planes on the fracture surfaces.
SEM images of the fracture surfaces of the samples austenitized at 850°C (a,b,c,d) or 1050°C (e,f,g,h) followed by oil quenching (a,d) or isothermal holding at 400°C (b,f), 450°C (c,g) or 500°C (d,h) for 1800 sec.
As shown in the stress – strain curves in Fig. 7, the samples austempered at 400°C shows larger elongations. To explore the reason for these large elongation, further characterization was conducted with those samples, especially focused on retained austenite. The most characteristic feature of the 400°C austempered samples has the significant amount of retained austenite. The carbon content of the retained austenite is one of the dominant factors to decide the stability of austenite.32,38) To determine the carbon content, XRD profile was obtained and the results were shown in Fig. 9. All of the diffraction peaks can be indexed as FCC or BCC phases whose lattice parameters are 0.287 nm and 0.362 nm, respectively. These peaks of the samples with different size of the prior austenite grains are well coincident to each other. This means the same carbon contents in both the samples. According to the report by Scott et al.39) about the relation between the lattice constant and the carbon content in the retained austenite, the solute carbon content in the retained austenite in this work can be evaluated as ranging from 1.0% to 1.4%, and this is similar to that in the low carbon low alloyed TRIP steels studied by Sugimoto et al.32)
XRD profiles of the samples isothermally held at 400°C for 1800 sec with two different prior-austenite grain sizes. (Online version in color.)
The carbon content of retained austenite has been studied and the measurement results were compared to the carbon content represented by T0 condition in which the free energy of austenite is equal to that of ferrite.40) According to the thermomechanical calculation,41) the carbon content at the T0 condition decreases with increasing Mn-content, indicating the change of carbon content in the retained austenite in the samples of this study. However, T0 indicates the upper limit of carbon content in bainitic ferrite.40) The distribution of carbon content changes according to microstructure. Actually, the study on the bainite steels with medium carbon content42) reported the carbon content lower than that of this work. It indicates that the discussion on the mechanism which controls the carbon content in retained austenite needs more experimental examinations.
To examine the deformation-induced martensitic transformation related to the TRIP behaviors, the EBSD measurements of the samples fractured by the tensile test were conducted and the results are shown in Fig. 10. The EBSD measurements were conducted at a place from which the fracture surface was distanced in longer than 5 mm. When comparing between these maps and the phase maps before tensile test in Figs. 5(a), 5(b), it is clearly found that the area fractions of the FCC phase in both the samples with different grain sizes largely decreased by the tensile test. This is due to the deformation-induced martensitic transformation of the retained austenite. The reduction of FCC phase does not change largely with the change of the prior-austenite grain size. This implies that the prior-austenite grain size has little influence on the kinetics of deformation-induced transformation.
Orientation color maps of BCC (a,b) or FCC (c,d) phases in the samples fractured by tensile test. The austempering was conducted at 400°C for 1800 sec with the austenitization at 850°C (a,c) or 1050°C (b,d). The orientation parallel to the tensile axis is illustrated. (Online version in color.)
The orientation dependence of the deformation-induced transformation can be found in the inverse pole figures in Fig. 11. These triangles show the probability of the tensile axis of the samples before and after the tensile test. The samples before the tensile test (a,c), no significant concentration of the orientation was found, indicating random distribution. On the other hand, both the samples after fracture, the orientation of the tensile axis strongly gathers around [111]. This concentration is difficult to be explained only by the crystallographic rotation by plastic deformation. One of the reasons is the amount of the elongation (~30%) which is not enough to move all the orientations to only one [111] orientation. In addition, the recent work43) clarified that the orientation rotation in the retained austenite has no specific rotation axis. Consequently, the texture change of the austenite by the tensile deformation indicates that the austenite grain with a tensile orientation parallel to [111] is difficult to have the deformation-induced transformation.
Inverse pole figures showing the distribution of orientation parallel to the tensile axis of the samples austenitized at 850°C (a,b) or 1050°C (c,d) and isothermally held at 400°C for 1800 sec. The results before the tensile test (a,c) or after the fracture by tensile test (b,d) were shown and the cross marks indicates the orientation showing the maximum intensity (max). (Online version in color.)
The most significant finding of the experimental results should be the superior strength - ductility balance of the sample austempered at 400°C as found in Fig. 7. These samples have fully bainite structure consisting of mainly bainitic ferrite and ~40% area fraction of retained austenite (Fig. 5). These contents are similar to those found previously in a low alloyed TRIP steels.14,17) Consequently, the strength - ductility balance obtained in this work was compared to the data reported in the previous studies on Si-bearing bainitic steels (TRIP steels) as well as the quenched and tempered medium-carbon steels, as shown in Fig. 12. In this figure, the marks circled by the dotted bold line are the data of the samples austempered at 400°C. The data of the quenched and tempered martensite examined by Kimura et al.44) were exhibited for a reference. In their work, the condition (sample size and test speed) of the tensile test is just the same as in this work. The data of the Si-bearing steels reported by Shinoda et al.14) (sheet steel, ~0.04/sec) and Matsumura et al.17) (sheet steel, ~0.003/sec) were also plotted. Additionally, data from more recent reports by Suzuki et al.15) (round bar, 0.0028/sec), Santos et al.16) (round bar, ~0.001/sec) and Gartia-Mateo et al.,22) (round bar, 0.004/sec) are added. The dotted lines indicate the contour lines designating the product of tensile strength and elongation for the guide of the strength - ductility balance. The data for the present work show better strength - ductility balance than that of the quench and tempered steels. When compared with the other Si-bearing steels, the best datum (400°C austempering) of the present work is compatible to the Shinoda’s data and some of the Matsumura’s data is superior to the present work. This should be due to the difference in the heat treatments: the process in the Shinoda’s work16) was conducted with fully austenitization while the Matsumura’s process included the intercritical heating before the austempering. The intercritical heating provides ferritic structure partially and resultant microstructure can be regarded as multiphased structure composed with bainite as well as retained austenite. In addition, the data reported by Gartia-Mateo et al.22) represented further superior strength-ductility balance with the strength over 2.0 GPa exploiting their process in which the austempering was conducted at lower temperatures (220°C or 250°C). Consequently, the superiority of these data indicates the possibility of further improvement of the strength - ductility balance of the medium-carbon steel with multi-phased structure including bainite with retained austenite.
Strength - ductility balance of the data shown in this work and the previous works on the medium carbon steels. (Online version in color.)
Another finding concerning of the strength-ductility balance is the fact that the effect of prior-austenite grain size is not significant in the present study. According to the previous work,45) retained austenite in bainite steel reduces the area of single facet on the fracture surface and this reduction contributes to the improvement of toughness. As shown in Figs. 8(d), 8(f), the area of unit facet of the sample austenitized at 850°C looks a little smaller than that in the sample austenitized at 1050°C. In other words, no big change was found between the samples with different prior-austenite grain sizes. This is because the refinement of prior-austenite grain size is not accompanied by the significant refinement of retained austenite grain size as found in Fig. 5.
4.2. Dependence of Orientation and Grain Size on Deformation-induced Transformation during Tensile TestThere are some studies concerning the crystallographic orientation dependence on the TRIP behaviors. Blondé et al.46) reported the results of the in-situ X-ray diffraction of the retained austenite during their tensile deformation and they clarified that the deformation-induced transformation preferentially occurs in the retained austenite grains with the tensile orientation nearly parallel to [001]. This is well consistent with the results of this study.
The change in the texture of the retained austenite by tensile deformation should be brought by the orientation rotation with the dislocation slip and/or the deformation-induced transformation. According to the previous works, the tensile deformation of the FCC metals provides the texture of both [111]//tensile axis and [001]//tensile axis when the deformation-induced transformation does not occur.47,48,49) Considering this information, the lack of the [001]//tensile axis in the present study should be mainly due to the deformation induced transformation.
The orientation dependence of the stability of retained austenite grains against deformation-induced transformation has been studied by several researchers.50) Some of them discussed with the character of martensitic transformation itself.51,52) De Knijf et al.53) discussed the orientation dependence using the quantitative theory based on the study by Patel and Cohen,51) while Kato and Mori54) explained the orientation dependence based on the model expressed by Bogers and Burgers52) in which the transformation dislocations from FCC to BCC phases are considered. On the other hand, the effect of the plastic deformation and the dislocation substructure formed in austenite matrix on the deformation-induced transformation has been discussed as well.55) The accumulation of dislocation is able to suppress the martensitic transformation due to the stress evolved by dislocation substructure.56) This effect was also considered in the study of the orientation dependence against the deformation-induced martensitic transformation from FCC to HCP phases in the 30%Mn austenitic steel.57) These variety of the discussions for the orientation dependence implies that additional data about the transformation behavior, such as the change of texture during tensile deformation., should be necessary. The discussion on this issue will be given in our future paper.
The difference of the area fraction of the retained austenite found in Figs. 5 and 10 indicates little effect of the prior-austenite grain size on the kinetics of the deformation-induced transformation. This is probably because there are no significant change of the retained austenite grain size and also because the deformation-induced transformation behavior seems to be decided mainly by the crystallographic orientation. It should be notable that these insensitivities of the prior -austenite grain size on the TRIP effect does not undermine the engineering advantage that the medium carbon Si-bearing steel with bainite structure has superior mechanical properties.
The medium-carbon Si-bearing steels was austempered at the temperature for upper bainite transformation in order to obtain the high strength - large ductility. The significant results shown in this work is as followings.
(1) The austenite grain refinement accelerated the bainite transformation and it is useful to reduce the process time to obtain full bainite structure.
(2) At the temperature below but near the bainite start temperature (450°C and 500°C), the kinetics of the bainite transformation was relatively suppressed and pearlite evolved partially. This heterogeneity deteriorated the tensile properties. These temperatures were not appropriate to obtain fully bainite structure.
(3) When compared the samples with fully bainite structures by austempering at 400°C, the prior-austenite grain refinement changed the shape of retained austenite grain from elongated to blocky morphology, while the grain size did not change largely.
(4) The sample austempered at 400°C performed preferential strength - ductility balance. The fraction of retained austenite was reduced by the tensile test, indicating the TRIP effect.
(5) The crystallographic orientation texture of the retained austenite in the as-austempered samples showed random distribution, while the retained austenite with [111] orientation parallel to the tensile axis of the sample preferentially untransformed to martensite after the tensile test, regardless of the prior -austenite grain size.
This study is based on work supported by a Grant-in-Aid for Scientific Research (ID: 17H03433, 20H02488) through the Japan Society for the Promotion of Science (JSPS). We acknowledge the supports of Dr. Hiroto, Mr. Iida, Mr. Kobayashi and Mr. Hibaru in the National Institute for Materials Science for the experiments.
