Quantification of Large Deformation with Punching in Dual Phase Steel and Change of its Microstructure –Part III: Micro-tensile Behavior of Pre-strained Dual-phase Steel

Abstract

The deformation behavior of inhomogeneous microstructures developed by pre-straining was studied by micro-tensile testing to elucidate the cause of low hole expandability of ferrite–martensite dual-phase (DP) steels. Slip bands developed in the ferritic phase, when the DP steel was cold-rolled (CR) at a reduction of 60% in thickness; in the 88% CR microstructure, ultrafine ferrite grains with a strong texture were locally observed. While the nanohardness increased with increasing pre-strain in the ferritic phase, it was constant in the martensitic phase. Tensile tests using micrometer-sized specimens with ferritic and martensitic phases revealed that the ultrafine-grained ferritic microstructure exhibited high yield strength but low ductility when compared to the slip band ferritic microstructure. While a shear type fracture occurred without necking in the former, the latter exhibited a chisel-edge type failure. In the absence of ultra grain refinement by pre-straining, the inhibition of slip transfer by the interphase boundary was a major contributor to the strengthening in the DP steel. The ductility loss of the severely deformed DP steel was attributed to localized strain in the ultrafine-grained ferritic microstructure.

1. Introduction

Dual-phase (DP) steels consisting of soft ferritic and hard martensitic phases are widely used for automotive applications because of their good balance between strength and ductility, and their high work hardenability at the onset of yielding.¹⁾ However, their low hole expandability is often a major drawback. A previous work by Yokoi et al. revealed²⁾ that the local region in the vicinity of the punched hole in a DP steel sheet is severely deformed during the punching process. At the same time, inhomogeneous microstructures develop that can be attributed to the difference in deformability between the ferritic and martensitic phases. In the subsequent hole expansion process, the inhomogeneous microstructures are stretched in the circumferential direction of the hole. As a result, cracking readily occurs at the periphery of the hole. Therefore, the hole expandability was reported to depend on the difference in hardness between the ferritic and martensitic phases.^3,4,5) Hasegawa et al. quantitatively indicated³⁾ that the hole expandability was improved by decreasing the difference in the hardness between the ferritic and martensitic phases. Ishiguro et al. also studied⁴⁾ the relationship between the hardness difference and hole expandability using similar-strength-level steels. They concluded that the reduction of the hardness difference between the ferritic and martensitic phases increases the local elongation, improving the hole expandability.⁴⁾ Thus, from a metallographic perspective, it is found that the microstructural inhomogeneity arising from the difference in the deformability between the ferritic and martensitic phases significantly degrades the hole expandability of the DP steel. In the previous work,²⁾ tensile tests were performed using a punched steel sheet to simulate the hole expansion process. A brittle fracture was observed at the periphery of the punched hole, whereas a ductile fracture occurred in the region far from the hole. In addition, tensile tests using micrometer-sized specimens fabricated from the periphery of the punched hole revealed a shear type fracture without necking, whereas the virgin material specimen exhibited a chisel-edge type fracture.²⁾ These findings implied that microstructural inhomogeneity due to severe deformation during the punching process led to the brittle fracture. The issue that remains is to understand the relationship between the microstructures developed in the vicinity of the punched hole and the mechanical responses. Meanwhile, Nakada et al. quantitatively evaluated⁶⁾ the strain distributions developed in punched and cold-rolled sheets of DP steel using electron backscatter diffraction (EBSD) analysis with a digital image correlation (DIC) method. The resulting local strain mapping demonstrated that the equivalent strain of the large strain band in the cold-rolled sheet was comparable to that of the heavily deformed edge in the punched sheet.⁶⁾ In the present study, the mechanical characteristics of the cold-rolled sheets of DP steel were examined to simulate the severely deformed region of the punched edge face.

Previous studies regarding the mechanical properties of DP steels argued^7,8) that cracking occurs predominantly in the martensitic phase or at the interface between the ferritic and martensitic phases. Therefore, it is necessary to consider the interaction between the ferritic and martensitic phases in order to investigate the mechanical response of the pre-strained sheet of DP steel. In addition, the microstructures inhomogeneously developed in the DP steel are complicated, even though cold rolling is employed instead of the punching process.⁶⁾ These complicated microstructures hinder the understanding of the mechanical characteristics in each phase when conventional mechanical testing is employed. Meanwhile, recent advances in the micro-mechanical testing techniques have helped elucidate the contribution of the micro-constituents to the mechanical properties in several steels and alloys.^{9,10,11,12,13)} Micro-tensile testing was employed in this study to clarify the deformation behavior at the interphase boundary in the inhomogeneous microstructures developed by cold rolling in the ferrite–martensite DP steel.

2. Material and Experimental Methods

The material used in this study was a low carbon steel, composed of 0.14C, 0.005Si, 1.00Mn, <0.002P, 0.0007S, 0.015Al, and 0.0012N (mass%), with the remainder being Fe. Thermo-mechanical treatment was performed to obtain a ferrite–martensite DP microstructure with a martensite fraction of 29 vol.%. The DP steel sheets were cold-rolled at reductions of 30%, 60%, and 88% in thickness (hereafter denoted as CR30, CR60 and CR88, respectively). The details of the sample preparation are described elsewhere.⁶⁾ Both surfaces of these cold-rolled sheets were polished using emery papers and diamond paste, and then they were finished using colloidal silica paste for EBSD analysis. The determination of the crystallographic orientation was performed at a step size of 0.2 μm under an accelerating voltage of 20 kV using field emission scanning electron microscopy (FE-SEM). The EBSD analysis was conducted using crystallographic orientation analysis software (OIM7.1.0), and a clean-up procedure was applied to all EBSD images. Nanoindentation was performed to investigate the hardness variation of each phase due to cold rolling. The samples for nanoindentation were prepared by electrochemical polishing.

Micro-tensile specimens with a gauge section of approximately 20 μm × 20 μm × 50 μm were fabricated from the CR60 and CR88 samples using a focused ion beam (FIB). The loading direction of the tensile specimen was arranged parallel to the rolled surface and the transverse direction of the cold-rolled sheet. On this occasion, an interphase boundary was included in the gauge section of the tensile specimen. The ferritic and martensitic phases were arranged in series for the type-A specimen and in parallel for the type-B specimen. Micro-tensile tests were performed at room temperature in laboratory air at a loading rate of 0.1 μm s⁻¹, which is equivalent to an initial strain rate of 2.0×10⁻³ s⁻¹. Based on the analysis of the shifts of 45 points selected on the optical microscope image during the micro-tensile test, the distributions of the maximum in-plane shear strain were determined. The fracture morphology was observed using SEM.

3. Results and Discussion

3.1. Inhomogeneous Microstructure Developed by Cold Rolling

Figure 1 shows the crystallographic orientation maps color-coded along the normal direction to the sheet surface in the unstrained, CR60, and CR88 samples. In the CR60 sample, slip bands were frequently observed in the ferrite grains, as shown in Fig. 1(b). In contrast, ultrafine grains as well as deformation bands were formed in the ferrite regions of the CR88 sample. The ultrafine-grained microstructures were locally developed in the vicinity of the martensite region, as shown in Fig. 1(c). The average grain size of the ultrafine-grained ferrite microstructure was approximately 0.7 μm. As for the martensite region, there were no significant changes in the microstructure as the strain was increased through cold rolling up to the 88% reduction in thickness.

Fig. 1.

EBSD maps of the unstrained microstructure and of the microstructures that were cold-rolled at thickness reductions of 60% (CR60) and 88% (CR88). Colors in the standard stereographic triangle show the crystallographic orientations along the normal to the plate plane. F and M represent the ferritic and martensitic phases, respectively. (Online vesion in color.)

Ohtani et al. characterized¹⁴⁾ the microstructures developed through tensile loading in the annealed ferrite–martensite DP steel using EBSD analysis. The formation of subgrains with large misorientations, i.e., grain subdivision, occurred by deformation in the ferrite grains; the neighboring martensite significantly influenced the grain subdivision.¹⁴⁾ Therefore, it is suggested that the grain subdivision occurred in the severely deformed region in the ferritic phase, and the occurrence of the grain subdivision might depend on the morphology and distribution of the martensitic phase.

Figure 2 shows the variations in the nanohardness and the Young’s modulus in the ferritic and martensitic phases as functions of the increased strain caused by cold rolling. The plot and error range indicate the average value and the 95% confidence interval, respectively, from 15 measurements. The Young’s modulus in both phases were independent of the strain. The average value of the nanohardness in the ferritic phase was approximately 4.8 GPa for CR88. This value was approximately 1.5 times higher than that of the unstrained sample, which was approximately 3.1 GPa. There is a linear relationship between the nanohardness in the ferritic phase and the reduction in thickness by cold rolling. In contrast, the nanohardness values in the martensite region ranged from 8.7 to 9.0 GPa, and this difference is less than the measurement accuracy. The nanoindentation study using the punched DP steel sheet shows²⁾ that the ferrite region in the vicinity of the punched edge face was work-hardened through plastic flow by punching, whereas the nanohardness in the martensite region of the CR88 specimen was comparable to that of the virgin steel. The nanohardness of the cold-rolled samples exhibited a similar tendency to that of the punched samples. This is reasonably interpreted as indicating the development of inhomogeneous microstructure in the ferrite region caused by severe deformation, as shown in Fig. 1. While the difference in nanohardness between the ferritic and martensitic phases was 3.8 GPa for CR88, which was lower than the value of 5.8 GPa observed for the unstrained sample, the hardness ratio of the ferrite to the martensite was 0.56 for CR88. In addition, in the CR88 samples, the nanohardness values may be noticeably different between the slip band and the ultrafine-grained regions in the ferritic phase.

Fig. 2.

Nanohardness and Young’s modulus in the ferritic and martensitic phases as a function of the reduction in thickness of cold rolling.

As mentioned previously, straining up to an 88% reduction in thickness by cold rolling led to no significant microstructural evolution and little work hardening in the hard martensite region. On the other hand, the soft ferritic phase was inhomogeneously deformed by cold rolling: the deformation bands developed in the CR60 samples, while ultra grain refinement occurred in the CR88 samples. In the following sections, the effect of the microstructural inhomogeneity in the ferritic phase due to pre-straining on the plasticity of the DP steel is discussed by using the tensile testing results of micrometer-sized specimens fabricated from the deformation band and ultrafine-grained regions.

3.2. Stress–strain Behavior of Specimens with Inhomogeneous Microstructures

Figure 3 shows the EBSD maps overlaid on the corresponding FIB images of the gauge section of the micrometer-sized specimens fabricated from the CR60 and CR88 samples. The gauge section of each specimen was composed of the ferritic and martensitic phases. In the ferritic phase, the CR60 and CR88 specimens included the slip band and ultrafine-grained regions, respectively. Although the interphase boundary was inclined with respect to the loading direction in the CR88–A specimen, as shown in Fig. 3(c), the ferritic and martensitic phases were considered to be in series. A previous micro-tensile testing study by Mine et al.¹¹⁾ revealed that the lath martensite steel exhibited habit-plane-orientation-dependent yielding in each packet. Therefore, our specimens were fabricated such that the martensitic phase was included in a single packet to facilitate the crystallographic analysis.

Fig. 3.

EBSD maps overlapped on the corresponding FIB images taken at the initial state of the micrometer-sized specimens. (Online vesion in color.)

Figure 4 and Table 1 show the nominal stress–nominal strain curves and tensile properties obtained by micro-tensile testing, respectively. The strain values were determined using the optical microscopy images, which were dynamically obtained at a frame rate of 8 fps by in-situ observation during tensile testing; however, the strain rate after the onset of the plastic instability was not constant. Therefore, the strain at the final point in the stress–strain curves (Fig. 4) differs from the fracture strain measured using the SEM image after failure (Table 1). We should use the fracture strain (Table 1) to evaluate the ductility. The CR88 specimen exhibited a high yield stress greater than 1200 MPa. The high strength readily satisfied the plastic instability conditions, resulting in premature fracture at a strain of several percent. Unlike the CR88–B with a similar yield stress, the CR60–B specimen exhibited a large fracture strain of approximately 26%.

Fig. 4.

Stress-strain curves obtained for the micrometer-sized specimens.

Table 1. The yield strength (σ_YS), the ultimate tensile strength (σ_UTS), the strain to failure (ε_f), and the yield ratio (σ_YS/σ_UTS) obtained through the micro-tensile testing.

	σYS (MPa)	σUTS (MPa)	εf (%)	σYS/σUTS
CR60-A	723	776	13.9	0.93
CR60-B	1300	1422	25.5	0.91
CR88-A	1851	2010	5.9	0.92
CR88-B	1287	1321	4.7	0.97

Figure 5 shows fracture surfaces of the micrometer-sized specimens. In the CR60–A and CR88–A specimens, in which the ferritic and martensitic phases were arranged in series, fracture occurred in the ferritic phase. Each fracture surface was composed of the ferritic and martensitic phases in the CR60–B and CR88–B specimens. The broken lines in Fig. 5 indicate the interphase boundaries. The CR60–A and B specimens exhibited chisel-edge type failure with significant necking (Figs. 5(a) and 5(b)). In addition, in the CR60–B specimen, the degrees of the inclination in the fracture surfaces were different between the ferritic and martensitic phases, and secondary cracking occurred at the interphase boundary (Fig. 5(b)). In contrast, a shear type fracture without necking occurred in the CR88–A specimen (Fig. 5(c)). This shear type fracture morphology resembles the brittle fracture morphology observed in the vicinity of the punched edge face in the tensile test of the punched sheet.²⁾ In the CR88–B specimen, the morphologies of the fracture surfaces were significantly different between the ferrite and martensite regions, as shown in Fig. 5(d). Surprisingly, brittle shear fracture occurred in the soft ferrite region, whereas the hard martensite region had dimples with necking. A secondary crack observed in the ferrite region corresponded to the pre-existing crack in the initial state. The effect of the secondary crack on the deformation behavior is discussed in section 3.4. The fractographic observation suggests that the shear type fracture in the ultrafine-grained ferrite region led to significant ductility loss in the CR88 specimens.

Fig. 5.

SEM images of the fracture surfaces of the micrometer-sized specimens.

3.3. Balance between the Strength and Ductility in DP Steel

As described in the previous section, the CR60 samples exhibited a good balance between strength and ductility, even though they were subjected to considerable pre-straining. In this section, the deformation mechanism of the DP steel is discussed by comparing the slip transfer behavior at the interphase boundary between the CR60–A and B specimens. The yield stress of the CR60–B specimen was approximately 1.5 times higher than that of the CR60–A specimen, while the yield ratio values (σ_YS/σ_UTS) were almost the same between the CR60–A and B specimens; the values were 0.93 and 0.91, respectively (Table 1). The fracture strain of the CR60–B specimen exhibiting a higher ultimate tensile strength was greater than that of the CR60–A specimen.

Figure 6 shows the (112) and (111) pole figures of the ferritic phase and the (110) and (111) pole figures of the martensitic phase obtained by EBSD analysis in the CR60–A and B specimens in the initial state. The circle and triangle symbols shown in the pole figures represent the normal direction to the slip plane and the slip direction, respectively, in the slip system exhibiting the highest Schmid factor; and the traces indicate the slip planes. The orientation was spread in the ferritic phase because of plastic flow introduced by cold rolling. The {110} <111> and {112} <111> were assumed to be the available slip systems. As previously mentioned, Mine et al. reported¹¹⁾ that the slip system parallel to the habit plane, i.e., the in-habit-plane slip system, was preferentially activated in the low carbon, low alloy steel, and the resolved shear stress at the onset of yielding for the slip system across the habit plane, i.e., the out-of-habit-plane slip system was 1.4 times higher than that for the in-habit-plane slip system. The most favorable slip system in each phase is shown in the pole figures. Considering the habit-plane-orientation-dependent yielding in the martensitic phase, the most favorable slip system of the CR60–A and B specimens corresponded to the in-habit-plane and out-of-habit-plane slip systems, respectively.

Fig. 6.

Pole figures of the ferritic and martensitic phases and the corresponding color-coded maps taken at the initial state of the CR60-A (a and b) and CR60-B (c and d) specimens. SF denotes the number of the highest Schmid factor in each phase. (Online vesion in color.)

Figure 7 shows the optical microscopy images and the distributions of the maximum in-plane shear strain in the CR60–A specimen in the micro-tensile test. The distributions of the maximum in-plane shear strain are characterized by an accumulated strain. Slip traces appeared in the ferritic phase at a strain of 0.4% (Fig. 7(b)). The slip traces were inclined at an angle of 49° with respect to the loading direction and corresponded to the (112) [111] slip system exhibiting the highest Schmid factor in the ferritic phase (Fig. 6(a)). In other words, in the CR60–A specimen, plastic deformation commenced by the activation of the primary slip system in the ferritic phase. Necking occurred in the ferritic phase at a strain of 2.7%, and no significant stain concentration was subsequently observed in the martensitic phase (Fig. 7(f)). The ratio of the nanohardness of the ferritic phase to that of the martensitic phase was 0.50 for the CR60 sample (Fig. 2). The yield ratio was 0.93 for the CR60–A specimen. Therefore, the plastic instability conditions were satisfied by decreasing the work hardening rate in the ferritic phase prior to the onset of yielding in the martensitic phase. In the CR60–A specimen, in which the ferritic and martensitic phases were arranged in series, the plastic deformation in the martensite region did not occur, and therefore the stress–strain curve predominantly reflects the deformation behavior of the ferritic phase with the slip band microstructure. Assuming no contribution of the martensite to the plasticity of the CR60–A specimen, the net fracture strain was 22%, which was determined by dividing the fracture strain by the initial length of the ferrite region. This value was equivalent to the fracture strain of the CR60–B specimen.

Fig. 7.

(a–d) Optical micrographs, taken during tensile testing, showing the deformation process of the CR60-A specimen, and (e and f) the corresponding distributions of the maximum in-plane shear strain, γ_max. The ε value indicates the nominal strain. (Online vesion in color.)

Figure 8 shows the progress of the tensile deformation process and the strain distribution in the CR60–B specimen. Slip traces appeared in the ferrite region at a strain of approximately 1.1% (Fig. 8(b)). Slip deformation transferred within the ferrite region (Fig. 8(c)) while maintaining a steady stress level after slight work hardening. The flow stress gradually decreased after the slip traces appeared in the martensite region (Fig. 8(d)). The slip traces that primarily appeared in each phase corresponded to the slip system exhibiting the highest Schmid factor shown in Fig. 6.

Fig. 8.

(a–e) Optical micrographs, taken during tensile testing, showing the deformation process of the CR60-B specimen, and (f–h) the corresponding distributions of the maximum in-plane shear strain, γ_max. The ε value indicates the nominal strain. (Online vesion in color.)

The resolved shear stress values at yielding were calculated for the primarily operative slip systems in the ferrite crystal to exclude the effect of the crystallographic orientation. The resolved shear stress at yielding for the (112) [111] slip system was determined to be 351 MPa for CR60–A and 628 MPa for CR60–B. A comparison of the tensile testing results between CR60–A and B suggested that the higher yield stress in the latter was attributed to the inhibition of the slip transfer by the interphase boundary. The high fracture strain of 26% was retained for the CR60–B specimen, despite the significant strengthening. The strain distributions in the CR60–B specimen indicate that plastic deformation occurred uniformly in both the ferritic and martensitic phases. The plastic deformation of the martensitic phase concurrent with the ferritic phase largely contributed to the superior ductility of the DP steel.

3.4. Strain Concentration in the Ultrafine-grained Ferritic Microstructure

Ikeda et al. studied¹⁵⁾ the tensile behavior of cold-rolled DP steel, which was the same material that we used in the present study, using bulk specimens with gauge section dimensions of 10-mm length, 1-mm width, and 0.5–1.0-mm thickness. While the unstrained sample exhibited a ductile fracture with strong necking, shear localization resulted in premature fracture in the CR88 sample.¹⁵⁾ As described in section 3.2, the micro-tensile tests revealed that shear type fracture led to significant ductility loss in the CR88 specimens. In this section, the plastic behavior in the CR88–A and B specimens is analyzed to clarify the mechanism of the strain concentration in the ultrafine-grained ferritic microstructure.

Figure 9 shows the pole figures of the ferritic and martensitic phases obtained from EBSD analysis in the initial state in both CR88 specimens. Because a strong texture developed in the ultrafine-grained ferrite microstructure, the Schmid factors for the operative slip systems were determined based on the average orientation of the texture. Figure 10 shows the strain distributions at a given strain during the tensile process and the fracture morphology of the CR88–A specimen. Although clear slip traces were not observed in the CR88–A specimen, the strain distribution maps reveal strain concentration in the ultrafine-grained ferrite region (Figs. 10(a) and 10(b)). In addition, the fracture position coincided with the strain concentration region (Fig. 10(c)). The orientation of the formed fracture surface corresponded to the slip system exhibiting the highest Schmid factor, which was determined based on the texture of the ultrafine-grained ferrite microstructure (Fig. 9(a)).

Fig. 9.

Pole figures of the ferritic and martensitic phases and the corresponding color-coded maps taken at the initial state of the CR88-A (a and b) and CR88-B (c and d) specimens. SF denotes the number of the highest Schmid factor in each phase. (Online vesion in color.)

Fig. 10.

(a and b) The distribution of the maximum in-plane shear strain, γ_max, and (c) the fracture morphology of the CR88-A specimen. The ε value indicates the nominal strain. (Online vesion in color.)

Mine et al. studied¹⁶⁾ the tensile behavior of ultrafine-grained Fe–0.01 mass% C steel processed by high pressure torsion (HPT). They indicated¹⁶⁾ that similar shear type fracture occurred in the ultrafine-grained iron at a high flow stress greater than 1100 MPa. When the work hardening rate was less than the flow stress, strain localization facilitated the shear type fracture in the ultrafine-grained ferrite region. The premature shear type fracture may be attributed to the combined effects of high strengthening and the reduction of the work hardenability through the ultra grain refinement by inhomogeneous deformation in the DP steel.

Figure 11 shows the strain distributions during the tensile test of the CR88–B specimen. Just as the CR88–A specimen, shear type fracture prevailed in the CR88–B specimen (Fig. 5(d)). The orientation of the shear fracture surface was not consistent with those of available slip systems in the texture of the ultrafine-grained ferrite microstructure. Close observation shows a crack in the strain concentration region in the initial state (Fig. 11(a)). Therefore, the stress concentration may be attributed to the pre-exiting crack. It should be noted that yielding occurred primarily in the martensitic phase despite its high strength. The fractographic observation of the CR88–B specimen showed plastic deformation in the martensite region (Fig. 5(d)). Nevertheless, strain localization in the ultrafine-grained ferrite region mainly contributed to the premature fracture. Cracking within the martensitic phase and at the ferrite/martensite interphase boundary is believed to be the main contributor to the premature fracture in DP steel.^7,8) In the case where the DP steel was heavily deformed, e.g., in the punching process, the presence of a crack facilitated the strain concentration into the ultrafine-grained ferrite, which resulted in the significant ductility loss.

Fig. 11.

(a) Optical micrograph of the top surface of the CR88-B specimen at the initial state, and (b and c) the distribution of the maximum in-plane shear strain γ_max. The ε value indicates the nominal strain. (Online vesion in color.)

In summary, a micro-tensile testing study using the CR88 specimens having a microstructure comparable to the one developed in the vicinity of the punched hole suggests that the strain concentration in the ultrafine-grained ferritic microstructure promoted the shear fracture to cause significant loss of ductility.

4. Conclusions

Micro-tensile testing and nanoindentation were employed in this study to clarify the relationship between the locally developed microstructure and the mechanical behavior in a pre-strained ferrite–martensite DP steel. The pre-strain was introduced by cold rolling up to an 88% reduction in thickness to simulate the heavily deformed region in the punched sheet. The conclusions can be summarized as follows:

(1) The nanohardness increased with increasing pre-strain in the ferritic phase, whereas it was constant in the martensitic phase.

(2) In the martensitic phase, no significant change was observed in the microstructure by cold rolling up to an 88% reduction in thickness. Slip band microstructure developed in the ferritic phase when the DP steel was cold-rolled at a reduction of 60% in thickness. Moreover, ultrafine ferrite grains with a strong texture were locally observed in the region adjacent to the martensite in the 88% cold-rolled DP steel.

(3) In the slip band ferrite microstructure, the yield stress was increased through the inhibition of the slip transfer by the interphase boundary. The plastic deformability of the martensite plays an important role in the enhanced balance between the strength and ductility of the DP steel.

(4) The specimen with the ultrafine-grained ferrite microstructure exhibited significantly high yield and flow stresses, but small uniform and fracture strains. This was attributed to the shear fracture in the ultrafine-grained ferrite region, which might result from the combined effects of high strengthening and the reduction in the work hardenability.

After examining the results of a series of studies,^2,6) we conclude that when ferrite–martensite DP steel is subjected to the punching process, plastic deformation localizes in the soft ferritic phase, which forms an ultrafine-grained ferrite microstructure at a much smaller equivalent plastic strain than in the case of a homogeneous microstructure. In the subsequent hole expansion process, the strain localization in the ultrafine-grained microstructure facilitates void formation. Therefore, the decreased proportion of the void growth process to the whole fracture process results in the premature fracture. This explains why the hole expandability of the DP steel is degraded.

Acknowledgement

The present work was supported in part by a Grant-in-Aid for Scientific Research (A) 15H02302 from the Japan Society for the Promotion of Science (JSPS).

References

• 1)   N. J.  Kim and  G.  Thomas: Metall. Trans. A, 12 (1981), 483.
• 2)   T.  Yokoi,  H.  Shuto,  K.  Ikeda,  T.  Tsuchiyama,  N.  Nakada,  T.  Ohmura,  Y.  Mine and  K.  Takashima: ISIJ Int., 56 (2016), 2068.
• 3)   K.  Hasegawa,  K.  Kawamura,  T.  Urabe and  Y.  Hosoya: ISIJ Int., 44 (2004), 603.
• 4)   T.  Ishiguro,  Y.  Yoshida,  N.  Yukawa,  T.  Ishikawa,  H.  Yoshida and  N.  Fujita: Tetsu-to-Hagané, 97 (2011), 38.
• 5)   H.  Shirasawa,  Y.  Tanaka and  K.  Korida: Tetsu-to-Hagané, 74 (1998), 120.
• 6)   N.  Nakada,  K.  Ikeda,  H.  Shuto,  T.  Yokoi,  T.  Tsuchiyama,  S.  Hata,  H.  Nakashima and  S.  Takaki: ISIJ Int., 56 (2016), 2077.
• 7)   J.  Zhang,  H.  Di,  Y.  Deng and  R. D. K.  Misra: Mater. Sci. Eng. A, 627 (2015), 230.
• 8)   T.  Sirinakorn,  S.  Wongwisws and  V.  Uthaisangsuk: Mater. Des., 64 (2014), 729.
• 9)   Y.  Mine,  K.  Hirashita,  M.  Matsuda,  M.  Otsu and  K.  Takashima: Corros. Sci., 53 (2011), 529.
• 10)   Y.  Mine,  H.  Fujisaki,  M.  Matsuda,  M.  Takeyama and  K.  Takashima: Scr. Mater., 65 (2011), 707.
• 11)   Y.  Mine,  K.  Hirashita,  H.  Takashima,  M.  Matsuda and  K.  Takashima: Mater. Sci. Eng. A, 560 (2013), 535.
• 12)   Y.  Mine,  H.  Takashima,  M.  Matsuda and  K.  Takashima: Mater. Sci. Eng. A, 618 (2014), 359.
• 13)   Y.  Mine,  R.  Maezono,  H.  Oda,  M.  Yamasaki,  Y.  Kawamura and  K.  Takashima: Mater. Trans., 56 (2015), 952.
• 14)   S.  Ohtani,  T.  Morikawa,  K.  Higashida and  S.  Hashimoto: Tetsu-to-Hagané, 95 (2009), 28.
• 15)   K.  Ikeda,  S.  Hata,  H.  Nakashima,  H.  Shuto and  T.  Yokoi: CAMP-ISIJ, 168 (2014), 295, CD-ROM.
• 16)   Y.  Mine,  S.  Matsumoto and  Z.  Horita: Corros. Sci., 53 (2011), 2969.