Optimizing Microstructure and Property by Ausforming in a Medium-Carbon Bainitic Steel

ISIJ International, Vol. 60 (2020),ISIJ International, No. 9 Vol. 60 (2020), No. 9, pp. 2007–2014

Optimizing Microstructure and Property by Ausforming in a Medium-carbon Bainitic Steel

Guanghui CHEN,1) Haijiang HU,1,2)* Guang XU,1) Junyu TIAN,1,2) Xiangliang WAN1) and Xiang WANG2)

1) The State Key Laboratory of Refractories and Metallurgy, Wuhan University of Science and Technology, Wuhan, 430081 China. 2) Department of Materials Science and Engineering, McMaster University, 1280 Main Street West, Hamilton, ON L8S4L7 Canada. (Received on January 28, 2020; accepted on March 23, 2020)

The transformation behavior and microstructure in a medium-carbon bainitic steel were investigated by combination of metallography and dilatometry. The fine micro-structural units of carbide-free bainite in non-ausformed and ausformed materials were measured by a transmission electron microscope. Mechanical stabilization of austenite in deformed material and its effect on property were analyzed by nanoindentation and tensile tests. Ausforming with a strain of 0.2 at 573 K can not only accelerate bainite transformation, but also improve the comprehensive properties. The strength and ductility of nanostructured bainitic steel can be simultaneously enhanced by ausforming, which should be attributed to the refinement of bainite and the enhanced volume fraction of retained austenite. Compared to the non- deformed material, the mechanical stabilization of austenite can be optimized by ausforming, resulting in good transformation-induced plasticity effects. Also a very important advantage was that, the bainite transformation time could be minimized into practical scale by prior ausforming compared to traditional low-temperature austempering.

KEY WORDS: ausforming; bainite transformation; retained austenite; property; mechanical stabilization.

theory indicates that a high level of substitutional solute was 1. Introduction required at a low-carbon steel to produce nanostructured Automobile industries are continuously challenged to bainite where the difference between the bainite and the Ms reduce weight and improve fuel efficiency due to economic then vanishes.11) Also at low carbon concentrations, experi- and environmental requirements. For that purpose, advanced mentally observations show that the thin platelets of bainite high strength steels were paid great attention to, such as tend to coalesce into coarse grains which are detrimental quenching and partitioning (Q&P) steel,1) medium Mn to strength and toughness.8) The prospects therefore look transformation-induced plasticity (TRIP) steel,2) TRIP aided promising for the design of a medium-carbon nanostruc- bainitic ferrite (TBF) steel,3) etc. Carbide-free bainite usu- tured bainite. ally formed in Si-rich steels accompanying with blocky and As for the control of RA, a multi-step bainitic austemper- film-like retained austenite (RA), which can contribute to ing process was developed to reduce blocky microstructure excellent combination of strength, ductility and toughness.4) and refine the subunits.12) During the first step isothermal To obtain nano-scaled bainite, composition design with holding, a higher austempering temperature was normally high-carbon high-alloy and very low temperature austem- adopted to partially form the relatively coarse bainitic pering were initially utilized.5) However, it took several ferrite plates. Extra carbon content can be rejected from hours even a couple of days to complete transformation, newly formed bainite to surrounding austenite, leading to which is unpractical from the viewpoint of production. its chemical stabilization and a decrease in Ms. Thus, one Also the coarse blocky austenite with low stability usually can use a deep-cold isothermal bainite transformation at transforms partially to brittle martensite during cooling after a lower temperature to obtain finer bainite. The secondly isothermal bainite transformation, resulting in deteriora- formed plates can also divide the untransformed austenite tion of ductility and toughness.6) Many efforts have been into pieces which further facilitates the stability of RA. In done to solve above problems, including optimal composi- addition, an inverted multi-step bainitic austempering route tion and processing design.7–10) Decreasing carbon content was proposed,13) in which the second-step temperature was can shorten the incubation of bainite nucleation, but raise higher than that in the first step. Compared to the previous martensite starting temperature (Ms) simultaneously. The strategy, a higher temperature in the second step can accelerate the bainite transformation. Although both above two * Corresponding author: E-mail: [email protected] multi-step methods contribute to the refinement of bainite DOI: https://doi.org/10.2355/isijinternational.ISIJINT-2020-054 and RA, leading to good comprehensive properties, the

2007 © 2020 ISIJ ISIJ International, Vol. 60 (2020), No. 9 procedure is complicated and still takes more than two hours to complete transformation. Fortunately, a low temperature ausforming can be utilized to not only accelerating transformation rate but also promoting the amount of bainite.14,15) Gong et al.16) considered that planar dislocations remaining on the active slip planes at 573 K can assist bainite transformation, accompanied by strong variant selection. Anisotropic dilatation during bainite transformation also illustrates the general rule of variant selection, that is, in each austenite grain, the main variant of which domain the bainite should belong to the same Bain group.17) Therefore, one can employ ausforming process to obtain sufficient nanostructured bainite in an expected transformation time. However, the control of RA by ausforming, which plays an important role in determining the property and performance, is still not clear. The relationship between ausformed bainite Fig. 1. Heat treatment and ausforming routes. (Online version in and property should be further clarified. color.) In the present work, the ~0.43 wt.% C bainitic steel was designed to investigate the effect of ausforming on bainite deformation with a strain of 0.2 at the rate of 1 s −1 and the transformation and property. The morphology of bainite and other was free of strain. The load was promptly removed RA in the deformed material is also discussed in details, and after deformation to keep the sample free of external com- efforts are also made to build a link between property and pressive stress during austempering. The deformed sample ausformed bainite. was then isothermally kept at 573 K for 60 min, while the non-deformed one was held at 573 K for 90 min shown as the blue line in Fig. 1. After isothermal holding, both two 2. Experimental Method cases were quenched to ambient temperature. The dilata- The chemical composition of the steel is 0.43C-1.90Si- tion data during the whole process were recorded by a laser 2.83Mn-0.57Al-0.06Cu (wt.%) with Fe balance. Si was dilatometer. To investigate the tensile property of ausformed added to suppress cementite precipitation during bainite bainite, hot rolling and consequent salt bath treatment tests austempering. Mn was added to increase the hardenability to were carried out using 140 mm × 20 mm × 10 mm blocks. obtain bainite and Al is for accelerating transformation. The The schedule was same as that in the thermomechanical material was refined in vacuum induction furnace and casted simulation test. After ausforming, the hot-rolled sample was into a 50 kg ingot. The ingot is homogenized at 1 523 K for put into a salt-bath furnace and isothermally held at 573 K 5 hours, followed by 8 passes hot rolling into a 12.0 mm for certain time. slab. After hot rolling with finishing temperature of 1 173 K, Microstructures were characterized on Keyence optical the steel plate was air-cooled to ambient temperature. The microscope (OM) and a Nova 400 Nano field emission bainite starting temperature (Bs) and Ms of the tested steel scanning electron microscope (FE-SEM) with an accel- were calculated as 707 K and 516 K, respectively, according eration voltage of 20 kV. The metallographical specimens to the following equations:18) were etched with 4% nital. The fine microstructures were observed using a JEM-2100F transmission electron micro- BPSi 839xxiC 270 F11exp.33 V ..... (1) B i H X scope (TEM). The samples for TEM examination were mechanically ground down to 30 μm thickness and then MKSi 565xxiC 600 F10exp.96 V .... (2) B i H X electrolytic thinned to perforation using an electrolyte where i = Mn, Si, Cr, Ni, and Mo, and the concentration composed of 5% perchloric acid and 95% glacial acetic x in wt.%. acid at ambient temperature, and voltage of 40 V. The volume fractions of RA in the ausformed and non-ausformed Pxii 86xxMn 23Si 67xxCr 67 Ni 75xMo ...(3) B i samples after isothermal transformation were determined using an X’Pert diffractometer with CoKα radiation under Kxii 31xxMn 13Si 10xxCr 18 Ni 12xMo ... (4) B i the following conditions: acceleration voltage, 40 kV; cur- Samples for thermomechanical simulation tests were rent, 150 mA; and step, 0.06°. Tensile specimens were machined to a dumbbell shape with the central cylinder prepared according to ASTM standard and the strain rate of 8.0 mm diameter and 12.0 mm height. The surface of was ~4×10 −3 s −1. Nanoindentation was performed on the samples was conventionally polished to keep the measure- lightly etched specimens using 2 000 μN load for 30 s in a ment face level and minimize the effect of surface rough- triboindenter TI-900 equipped with scanning probe micro- ness. Ausforming and austempering tests were conducted scope. Each test involved a 5 × 5 array of indents. Vickers on a Gleeble-3500 thermal simulator according to the hardness tests were performed on a HV1000A micro- processing schedules shown in Fig. 1. The specimens were hardness tester (0.2 kg-1 960 mN). The average value of at heated to 1 273 K at 5 K s −1 and kept for 15 min, and then least ten individual measurements was calculated, including cooled to 573 K. A cooling rate of 5 K s −1 was utilized several martensite bands and bainite blocks of the micro- to avoid ferrite and pearlite transformation. Subsequently, structure. two routes were designed, i.e. one was applied compressive

drag effects can bring isothermal bainite transformation 3. Results and Discussion into stable stage,20,21) leading to remaining of austenite after 3.1. Original Microstructure and Property long isothermal holding. Therefore, blocky M was obtained A small metallographic specimen was selected from the during the cooling process after bainite transformation. The center of hot-rolled slab and the original SEM micrograph of growth of bainite can section residual austenite, thus one the steel was shown in Fig. 2(a). The microstructure consists can see that the marked M block was cut into four parts by of granular bainitic ferrite and carbides, which shows micro- three sheaves. Ausforming can stimulate dense nucleation structure characteristic of medium-temperature transition. of bainite, resulting in multiple-crossed morphology in Fig. Figure 2(b) shows the phase transition under equilibrium 3(d). conditions calculated by Thermal-calc with the database of TCFE-6. The starting temperature of γ→α transformation 3.3. TEM Results is 1 081 K, and cementite begins to precipitate at 1 004 K. The fine microstructure of BF was confirmed by examin- After hot rolling, the steel was air cooled to ambient tem- ing the microstructure at higher magnifications using TEM. perature, leading to intermediate phase transformation. Figure 4 shows a typical morphology of bainitic plates However, the present work aims to produce nanostructured in the non-ausformed steel, in which most of them are in bainite. Thus a larger cooling rate should be considered. the form of clusters of parallel plates, emanating from the The tensile strength and total elongation of the test steel are prior austenite grain boundaries shown by yellow arrows. 1 016 MPa and 23.3%, respectively, which can be taken as Thin film-like RA is trapped between adjacent BF plates reference.

3.2. Microstructure after Ausforming and Austemper- ing Figure 3 shows the microstructures of the non-ausformed and ausformed samples after isothermal transformation for different times. Both the samples comprise of needle- like bainite and matrix of martensite, as shown by arrows in Figs. 3(a) and 3(b). The deformed sample seems to contain more bainite compared to the non-deformed one in spite of the shorter isothermal holding time. From the high-resolution SEM micrographs in Figs. 3(c) and 3(d), a sheave of bainitic ferrite (BF) consists of several subunits with film-like RA between them. Blocky RA also canbe seen in both cases. Large blocks of martensite (M, dotted box in Fig. 3(c)) still exist in the non-deformed materials even undergoing 90 min isothermal transformation, while the microstructure of the deformed one was divided into pieces. Carbon diffusion from BF into untransformed austenite during the progression of bainite transformation has been proved by many researchers. Base on T0 theory, bainite transformation ceases when carbon concentration of untransformed austenite reaches the critical value where the Fig. 3. (a) and (c) OM and SEM micrographs of the non-deformed free energy of austenite is equal to that of ferrite.19) In addi- sample after isothermal holding at 573 K for 90 min; (b) tion, both partition local equilibrium (PLE) and negligible and (d) OM and SEM micrographs of the ausformed sample after isothermal holding at 573 K for 60 min. (Online partition local equilibrium (NPLE) modes proved that solute version in color.)

Fig. 2. (a) SEM micrograph of initial materials before heat treatment; (b) phase diagram of equilibrium calculation, γ refers to face-centered cubic (fcc) austenite and α is body-centered cubic (bcc) ferrite. (Online version in color.)

Fig. 5. Plate-shaped bainite formed in the non-ausformed sample, (a) bright field image and (b) corresponding dark field image taken from an austenite reflection, α refers to bainite plate and γ is RA. (Online version in color.)

Fig. 4. Typical TEM micrographs showing full morphology of bainite in the non-ausformed sample transformed at 573 K for 90 min. (Online version in color.) with longer and straight feature, as shown in Fig. 5. The dark features lying within the plates and at the angle to the axis of these plates in Fig. 5(a) are similar in appearance to typical low bainitic carbides. However, they are actually regions of RA, as evidenced by dark field micrograph of Fig. 5(b). The bainitic plate is composed of sub-units and the films of austenite have sometimes a typical wavy morphology characteristic in high Si steels, which are dispersed between the sub-units of BF. The sub-units do not appear to be entirely separated by RA. The extent of transformation in some plates can be so large as to make the individual plate indistinct, for example, Fig. 4. In addition, different crystallographic orientation of plates shown by red line can be seen in one prior austenite grain. Figure 6 presents a characteristic microstructure that was achieved when prior-deformation of 20% was applied followed by isothermal transformation at 573 K for 60 min. The general morphology is different from that in the non-deformed sample, that is, the less crystallographic orientations of sheaf appear. In local area, BF sheaves Fig. 6. TEM Micrograph of BF sheaves align more or less in a align more or less in a common direction. It may due to common direction in ausformed sample. (Online version the variant selection of bainite in ausformed austenite, in color.) which has been theoretically and experimentally proved by many works.16,17,22,23) BF exhibits the orientation relationships with respect to the parent austenite, scattered around Kurdjumov-Sachs (K-S), Nishiyama-Wassermann (N-W) and Greninger-Troiano (G-T) relationships.24) There are different crystallographic bainitic variants that can form from a given orientation of the parent austenite. Ausforming can increase the probability of certain variants to form since the energy that is needed to transform from austenite to that crystallographic variant is lower than for others. When only several of the variants have grown from each prior austenite grain, it is said that there is variant selection, resulting in the micro-structural alignment. The width and length of bainite unit were quantitatively measured and averaged on different fields of TEM observations, where hundreds of units were measured for each thermomechanical and isothermal history. The variation of the average width and length of bainitic unit, aspect ratio and respective standard deviation for the different thermo- Fig. 7. Average width and length of bainite plate unit in different mechanical conditions are summarized in Fig. 7. For the thermo-mechanical conditions. (Online version in color.)

© 2020 ISIJ 2010 ISIJ International, Vol. 60 (2020), No. 9 non-ausformed material, the average width and length is approximately horizontal dilatation at the final stage of each ~0.724 μm and ~4.398 μm, respectively, while ~0.151 μm case in Fig. 8(a) as one hundred percent of individual bainite in width and ~0.830 μm in length are presented for the aus- reaction. The incubation period (the time to complete 5% formed sample. For both width and length of unit, it can be bainite transformation) for the bainite transformation in the seen that the value decrease significantly in the sample with ausformed sample is only about 287 seconds, which is obvi- a strain of 0.2 when compared with the non-deformed one. ously shorter than the non-ausformed sample (about 1 120 The results demonstrate that microstructures formed during seconds). Also the time to complete bainite transformation prior deformation limit the growth of bainite unit. However, is much shorter in the ausformed case. The accelerated it is of interest to note that the aspect ratio (width/length) transformation kinetics can be attributed to the enhanced remains unobvious changed for different conditions. nucleation efficiency and facilitated variants by deformation. It is clearly seen that the grain boundaries serve as the nucleation site of growth of bainitic plate in non-deformed 3.5. Volume Fractions of Different Phases sample, while in prior-deformed sample to produce the fine The XRD diffractograms for the non-deformed and intersected platelet structure characteristic of BF sheaves, deformed materials are given in Fig. 9. The volume frac- two conditions are necessary to be satisfied. One is a number tion of RA was calculated based on integrated intensities of of intragranular nucleation events are required, another is (200)α, (211)α, (200)γ, (220)γ, and (311)γ diffraction peaks, that nucleation gives rise to multiply oriented platelet vari- according to Eq. (5):25) ants so that the chaotic interlocking structure is evident in 1 Figs. 3(d) and 6. Additional nucleation sites within grain can Vi ...... (5) 1 GI / I be induced by dislocations formed by the plastic deforma- 10) tion of the austenite. The nucleation site is increased in where Vi is the amount of austenite for each peak, Iα and deformed sample but the growth space is restrictive, result- Iγ are the integrated intensity of ferrite and austenite peaks, ing in that each nucleus then transforms a smaller size due respectively, and G is constant for each peak. The carbon to interaction of neighbouring platelets. content in RA was estimated using Eq. (6):26)

3.4. Dilatometry Analysis CM 3.. 5780 0 00095 n ...... (6) Figure 8 shows the dilatation results of the non-ausformed 0.. 0056Al 0 0015Cu /. 0 033 and ausformed samples along the diameter direction during isothermal bainite transformation at 573 K. There is a dis- where Cγ is the carbon content in RA and αγ is the lattice tinct difference between the final amounts of dilation in two parameter of RA, which is determined by the position of cases, shown in Fig. 8(a). For the non-deformed sample, three austenite peaks using Cohen’s method, and Mnγ, Alγ the dilatation can represent the real amount of bainite trans- and Cuγ represent the concentration of individual elements formation, while it is inappropriate for the deformed one (mass-%) in RA. Based on microstructure results, the due to variant selection. According to the micro-structural volume fractions of bainite were calculated using Image result, the amount of bainite in deformed austenite is lightly Pro-Plus software, and the detail was described in Ref. 27). larger than that in the non-deformed sample. However, the After multiple measurements, the average results are shown dilatation result shows a noticeable increase. Anisotropic in Table 1. The amount of bainite can be promoted by dilatation during bainite transformation in ausformed aus- ausforming with a strain of 0.2 at 573 K, although the tenite in nanostructured bainitic steel can be caused by vari- isothermal holding time is shorter than the non-ausformed ant selection. He et al.17) also reported that the ausformed sample. Moreover, the volume fraction of RA was increased sample shrunk in the axial direction but expanded in the to ~23.2% by deformation, indicating that the stability of radial direction. Nevertheless, the transformation rates can austenite can be enhanced by ausforming. The higher carbon be calculated based on dilatation data using tangent method, content of RA in the non-deformed sample is mainly due and the result was given in Fig. 8(b). One can assign the to the long bainite transformation time, which allows more

Fig. 8. (a) Dilatation change versus time during isothermal holding reflecting bainite transformation, and (b) transformation rates in non-ausformed and ausformed samples, showing a much faster kinetics of bainite transformation after prior deformation.

Fig. 9. X-ray diffraction diagrams, (a) non-ausformed sample and (b) ausformed with a strain of 0.2. (Online version in color.)

Fig. 10. (a) Image showing hardness tests on BF and M, and (b) Vickers micro-hardness results of non-ausformed and ausformed samples. (Online version in color.)

Table 1. Volume fractions of different phases (%) and carbon content in RA (wt.%).

Sample BF M RA Cγ Non-ausformed 53.8 ± 2.4 38.4 ±1.8 7.8 ± 0.8 1.18 ± 0.08 Ausformed 63.3 ± 2.5 13.5 ±1.3 23.2 ±1.5 0.96 ± 0.08

carbon diffusion into adjacent austenite.

3.6. Mechanical Properties Vickers micro-hardness tests were carried out on different phases, illustrated in Fig. 10(a), in which BF and M were located in different areas. It is obvious that the area of the collapsed rhombus of M is small than that of BF. The average results were plotted in Fig. 10(b). The hardness of Fig. 11. Engineering stress-strain curves of non-ausformed and ausformed samples. (Online version in color.) bainite in the ausformed sample is 563±34 HV0.2, larger than that in the non-ausformed one (532±29 HV0.2). On the contrary, the hardness of M decreases slightly after defor- can significantly increase the strength by microstructure mation with a strain of 0.2, which should be attributed to the strengthening due to isothermal bainite transformation. lower carbon content in the residual austenite ~0.96 wt.%. According to Bhadeshia and Young,28,29) the strength of The nominal engineering stress-strain curves were given bainite and martensite can be calculated as follows in Eq. in Fig. 11. The ultimate tensile strengths (UTS) of the (7): non-deformed and deformed samples are ~1 521 MPa 1 and ~1 755 MPa, respectively. It indicates that ausform- i kL K 05. ....(7) Fe B i SS C  3 ppt Dd ing with a strain of 0.2 at 573 K can increase the strength of bainitic steel. Furthermore, the total elongation (TE) of Where σFe is strengthening contribution of pure annealed i the ausformed material ~18.1% is better than that of the bcc iron, σ SS is solid solution strengthening due to substi- non-ausformed one ~14.7%. Compared with the original tutional solutes, σC is solid solution strengthening due to material, both non-ausforming and ausforming processes carbon, σppt is precipitation strengthening from carbide par-

1 ticles and kL 3 is the lath size strengthening component The dislocation-rich BFs were observed in dark field image including the constant kϵ, and L3 is mean linear intercept of taken from a ferrite reflection, as shown by arrows in Fig. 05. 32) laths. KDdρ is the dislocation strengthening component 12(b). It was reported by He et al. that intensive mobile of the model, where the constant KD = 0.38 μb for bcc dislocations can not only increase strength but also improve metals, μ is the shear modulus and b is the burgers vector ductility, which was realized by deformed and partitioned (~0.25 nm). Compared to the non-ausformed sample, the (D&P) process. In the present work, BF and RA with high bainite laths in the ausformed sample were much finer, lead- dislocation density were simultaneously produced by aus- ing to a high strength. The dislocation density is calculated forming and austempering processes, resulting in a high based on the TEM measurements, 2.51 × 1015 m −2 and product of strength and elongation ~31.8 GPa% (UTS×TE). 5.32 × 1015 m −2 for non-deformed and deformed samples, Finally, nanoidentation tests were performed to investi- respectively. Based on Eq. (7), the calculated contribu- gate the mechanical stabilization of RA in non-ausformed tion to strength due to refinement of BF by ausforming is and ausformed samples. The typical load-displacement about 603 MPa, and that due to enhanced dislocation is curves were plotted in Fig. 13, in which the terraces pointed ~174 MPa. One may note that the difference between the by arrows signify the transformation from RA to M. It is UTSs is just ~234 MPa, smaller than the combined calcula- interesting to find that, the loads to induce M transformation. The reason should be attributed to the larger amount tion in non-deformed sample are all above ~1 500 μN of hard martensite in non-deformed sample. From Table 1, (Fig. 13(a)), while it mostly ranges 250~1 000 μN for the the carbon content of non-deformed RA is the higher as ~1.18 wt.%, resulting in enhanced hardness of martensite as well as strength.

3.7. Mechanical Stabilization In low-temperature banitic steel, the strength is mainly related to the fine scale of the BF plates and martensite (hard phases), ductility is mostly controlled by the RA (soft phase).30) In order to take full advantage of TRIP effect, the stability of RA, i.e., its capability to transform to martensite under stress, must be neither too low nor excessively high. Before that, the stabilization of austenite during isothermal transformation should be considered first, which directly determine the volume fraction of RA at ambient temperature. The stabilization of ausformed austenite depends on carbon content and dislocation caused by deformation, in which the latter plays a more important role,31) leading to the larger amount of RA in the deformed sample. Figure 12 shows the fine structures of BF and RA in the ausformed samples. From Fig. 12(a), the phase of γ randomly distributes among α-plates, which mostly seems to be in a coarse size than that in non-ausformed sample (Fig. 5(b)). Actually, film-like RA also exists between aus- Fig. 12. TEM micrographs showing bainite plates formed in ausformed BF, which can be clarified by dark field image taken formed sample, (a) bright field image, (b) corresponding from an austenite reflection in Fig. 12(c). With a strain of dark field image taken from a ferrite reflection, (c) dark 0.2 at 573 K, high-density dislocation was introduced into field image taken from an austenite reflection, and (d) corresponding diffraction pattern. (Online version in the parent austenite, which then inherited by bainitic plate. color.)

Fig. 13. Nanoindentation results of (a) non-ausformed and (b) ausformed samples, and five typical load-displacement curves were selected corresponding to five indents in upper-left image, terrace means martensite transformation induced by stress. (Online version in color.)

2013 © 2020 ISIJ ISIJ International, Vol. 60 (2020), No. 9 ausformed one (Fig. 13(b)). For the very rare indent, the present work. TEM work was conducted at Canadian Centre load increases to ~1 780 μN shown by red line in Fig. for Electron Microscopy (CCEM), Hamilton, ON, Canada. 13(b). It indicates that the RA in non-ausformed sample is too stable due to high carbon content and mostly film-like Data Availability morphology. The tensile fracture happened before the TRIP Both the raw data and the processed data required to effect was produced, leading to a limited ductility. Another reproduce these findings can be provided by the authors. important reason is that the volume fraction of RA is just ~7.8% which is very small. Bhadeshia33) et al. reported that REFERENCES tensile failure in nanostructured bainite occurred when the 1) P. Huyghe, S. Dépinoy, M. Caruso, D. Mercier, C. Georges, L. Malet RA content is diminished to about 10%. In this case, only and S. Godet: ISIJ Int., 58 (2018), 1341. 2) J. Hu, W. Q. Cao, C. Y. Wang, H. Dong and J. Li: ISIJ Int., 54 the ausformed sample can have TRIP effect, which seems (2014), 1952. corresponding to the nanoidentation result. In total, aus- 3) K. Sugimoto, T. Hojo and J. Kobayashi: Mater. Sci. Technol., 33 (2017), 2005. forming with a strain of 0.2 at 573 K can not only increase 4) C. Hofer, H. Leitner, F. Winkelhofer, H. Clemens and S. Primig: the strength but ductility as well, which can be attributed to Mater. Charact., 102 (2015), 85. refinement of BF and large volume fraction of RA. Also a 5) C. Garcia-Mateo, F. G. Caballero and H. K. D. H. Bhadeshia: Mater. Sci. Forum, 500–501 (2005), 495. very important advantage was that, the bainite transforma- 6) Y. Li, W. Li, W. Q. Liu, X. D. Wang, X. M. Hua, H. B. Liu and X. tion time could be optimized into a practical scale. J. Jin: Acta Mater., 146 (2018), 126. 7) C. Garcia-Mateo, F. G. Caballero and H. K. D. H. Bhadeshia: ISIJ Int., 43 (2003), 1821. 8) H. K. D. H. Bhadeshia: Proc. R. Soc. A, 466 (2009), 3. 4. Conclusions 9) G. M. A. M. El Fallah and H. K. D. H. Bhadeshia: Mater. Sci. Eng. A, 746 (2019), 145. The transformation behavior and microstructure evolu- 10) G. H. Chen, G. Xu, H. S. Zurob, H. J. Hu and X. L. Wan: Metall. tion in designate medium-carbon nanostructured bainitic Mater. Trans. A, 50 (2019), 573. 11) H. K. D. H. Bhadeshia: In: Howe, J. M., Laughlin, D. E., Lee, J. K., steel were investigated by metallographic and dilatometry. Dahmen, U., Soffa, W. A. (Eds.), Solid-Solid Phase Transformations, Mechanical stabilization of austenite in deformed material TMS-AIME, Warrendale, PA, USA, (2005) 469. 12) X. L. Wang, K. M. Wu, F. Hu, L. Yu and X. L. Wan: Scr. Mater., and its effect on property were analyzed by nanoindentation 74 (2014), 56. and tensile tests. The conclusions can be drawn as follows: 13) G. H. Gao, H. R. Guo, X. L. Gui, Z. L. Tan and B. Z. Bai: Mater. (1) Ausforming with a strain of 0.2 at 573 K can not Sci. Eng. A, 736 (2018), 298. 14) W. Gong, Y. Tomota, M. S. Koo and Y. Adachi: Scr. Mater., 63 only accelerate bainite transformation, but also improve (2010), 819. the comprehensive property which shows enhanced com- 15) H. J. Hu, H. S. Zurob, G. Xu, D. Embury and G. R. Purdy: Mater. Sci. Eng. A, 626 (2015), 34. bination of strength and ductility, i.e. ultimate strength 16) W. Gong, Y. Tomota, Y. Adachi, A. M. Paradowska, J. F. Kelleher ~1 755 MPa and total elongation ~18.1%. and S. Y. Zhang: Acta Mater., 61 (2013), 4142. (2) Low temperature bainite can be refined by ausform- 17) J. G. He, J. Du, W. Z. Zhang, C. Zhang, Z. G. Yang and H. Chen: Metall. Mater. Trans. A, 50 (2019), 540. ing due to additional nucleation sites induced by dislocation. 18) S. M. C. van Bohemen: Mater. Sci. Technol., 28 (2012), 487. However, the aspect ratio (width/length) remains unobvious 19) H. K. D. H. Bhadeshia and D. V. Edmonds: Acta Mater., 28 (1980), 1265. changed for different conditions. Micro-structural alignment 20) Y. Xia, G. Miyamoto, Z. G. Yang, C. Zhang and T. Furuhara: Acta of bainite sheaves can form in deformed austenite due to Mater., 91 (2015), 10. 21) H. D. Wu, G. Miyamoto, Z. G. Yang, C. Zhang, H. Chen and T. variant selection. Furuhara: Acta Mater., 133 (2017), 1. (3) The simultaneous increase of strength and ductil- 22) Y. Kimura and T. Inoue: ISIJ Int., 56 (2016), 2047. ity can be attributed to refinement of BF and more volume 23) H. Beladi, V. Tari, I. B. Timokhina, P. Cizek, G. S. Rohrer, A. D. Rollett and P. D. Hodgson: Acta Mater., 127 (2017), 426. fraction of RA. Compared to the non-deformed material, the 24) T. Moritani, N. Miyajima, T. Furuhara and T. Maki: Scr. Mater., 47 mechanical stabilization of austenite can be optimized by (2002), 193. 25) C. Y. Wang, J. Shi, W. Q. Cao and H. Dong: Mater. Sci. Eng. A, 527 ausforming, resulting in better TRIP effects. (2010), 3442. 26) D. J. Dyson and B. Holmes: J. Iron Steel Inst., 208 (1970), 469. 27) H. J. Hu, G. Xu, L. Wang, Z. L. Xue, Y. L. Zhang and G. H. Liu: Acknowledgement Mater. Des., 84 (2015), 95. The authors gratefully acknowledge the financial supports 28) H. K. D. H. Bhadeshia: Bainite in Steels, 3rd ed., CRC Press, London, (2015) 304. from the National Nature Science Foundation of China 29) C. H. Young and H. K. D. H. Bhadeshia: Mater. Sci. Technol., 10 (Nos. 51704217, 51874216), China Postdoctoral Science (1994), 209. Foundation (2017M622533), the Major Projects of Tech- 30) C. Garcia-Mateo and F. G. Caballero: ISIJ Int., 45 (2005), 1736. 31) H. J. Hu, G. Xu, L. Wang, M. X. Zhou and Z. L. Xue: Met. Mater. nological Innovation in Hubei (No. 2017AAA116), Hebei Int., 21 (2015), 929. Joint Research Fund for Iron and Steel (E2018318013). The 32) B. B. He, B. Hu, H. W. Yen, G. J. Cheng, Z. K. Wang, H. W. Luo and M. X. Huang: Science, 357 (2017), 1029. authors also thank Prof. Hatem Zurob at McMaster Uni- 33) H. K. D. H. Bhadeshia: Proc. 1st Int. Symp. on Steel Science (IS3- versity, Canada for the suggestions and corrections on the 2007), ISIJ, Tokyo, (2007), 17.