Table 1. Chemical composition of the steel. (wt %)

Transcription

1 Microstructure and Mechanical Properties of Dual-phase Steel Produced by Intercritical Annealing of Lath Martensite* By Naomi MATSUMURA** and Masaharu T OKIZANE*** Synopsis Microstructure and tensile properties of a dual phase (martensite +ferrite) Fe-2.3%Mn-0.05%C-0.03%Nb steel produced by intercritical annealing of specimens with martensite structure have been studied. In order to assess the effects o f prior austenite grain size on microstructure and tensile properties of the dual phase steel, the specimens in a martensitic state with widely different prior austenite grain size were prepared by thermal cycling and thermomechanical processing. Coarse dual phase structure consisting of fibrous martensite and ferrite was obtained by intercritical annealing of the specimens with coarse prior austenite grain size. A characteristic fine dual phase structure consisting of homogeneously dispersed fine martensite particles and fine ferrite grains was obtained by the intercritical annealing of the specimens with ultra fine prior austenite grain size. The fine dual-phase structure was superior in both strength and ductility to the coarse dual phase structure over a wide range of martensite volume fractions examined. It is concluded that better combination of strength and ductility of the dual phase steel is achieved by intercritical annealing of the martensitic specimens with ultra-fine prior austenite grain size which is obtained by the thermomechanical processing. I. Introduction Dual-phase steels whose structures consist of ferrite (a) and martensite (a') have received considerable attention in the past several years because they offer attractive combination of strength and ductility. Various investigations)-91 have been made to obtain clear understanding of the mechanical properties of dualphase steel on the basis of the theories developed for metals with composite structure. At the same time, it has been pointed out that morphology (size, shape, distribution) of the structural constituents is especially important in controlling the mechanical properties of dual-phase steels.lo-14~ Therefore, the initial microstructure and the path by which the final a-a' mixtures is obtained are presumed to exert a profound influence on the mechanical properties of the dualphase steels. In most of the previous investigations, the fundamental procedures for production of dual-phase steels have been performed by cooling austenite (r) into (ce+r) range or by intercritical annealing of steels consisting of a+pearlite. There has been little information about the structure and the properties of dualphase steel produced by intercritical annealing of steels consisting of lath a'. In the present study, microstructure and tensile properties of a dual-phase (a +a') Fe-2.3 %Mn % G-0.03 %Nb steel produced by intercritical annealing of specimens with martensite structure have been studied. In order to assess the effects of prior r grain size on microstructure and tensile properties of the dual-phase steel, the specimens in a martensitic state with widely different prior r grain size were prepared. II. Experimental Procedure 1. Material and Preparation of Specimens The details of chemical composition of the material used is given in Table 1. The plates 7.5 mm and 1.5 mm in thickness were obtained by surface grinding and cold rolling of as received plates about 15 mm in thickness. These plates were austenitized at C for 1 h in vacuum and then iced brine quenched. The 7.5 mm thick plates were cold rolled to the 1.5 mm thick plates (80 % reduction in thickness) having the same thickness as the non-deformed plates. Tensile specimens with a 25 mm gage length and 5 mm wide and the specimens for microscopic examination were cut from the cold-rolled and non-deformed plates prior to the subsequent heat treatments. Figure 1 Fig. 1. Schematic diagrams of heat treatment and rolling schedule of the specimens. (T indicates the intercritical annealing temperature which was varied between 715 C and 800 C.) Table 1. Chemical composition of the steel. (wt %) Originally published in Tetsu-to-Hagane, 70 (1984), 246, in Japanese; Formerly presented to the 105th ISIJ Meeting, April 1983, 5644, at Tokyo Institute of Technology in Tokyo. English version received March 2, ISIJ Graduate School, Ritsumeikan University, Tojiin Kitamachi, Kita-ku, Kyoto 603. Department of Mechanical Engineering, Faculty of Science and Engineering, Ritsumeikan University, Tojiin Kitamachi, Kita-ku, Kyoto 603. (648)

2 Transactions ISIJ, Vol. 24, 1984 (649) shows the schematic diagrams of heat treatment and rolling schedule used in the present study to obtain the dual-phase structures with various morphologies. In the case of series A, non-deformed specimens were directly offered to the intercritical annealing at the several selected temperature, T, between 715 C and 800 C. In the case of series B, non-deformed specimens were austenitized at 880 C (just above A3 temperature) for 20 s and iced brine quenched to produce the martensitic structure with finer prior r grain size than that in the case of series A, and were then offered to the similar intercritical annealing. In the case of series C, the cold-rolled specimens were austenitized at 880 C for 20 s and iced brine quenched to obtain the a' structure with extensively finer prior r grain size. Such a kind of thermomechanical processing, which consists of severe cold rolling of low carbon lath a' and subsequent rapid heating at the temperature just above A3 point, is the most successful method for r grain refinement.15~ Moreover, it has been pointed out that such an effect of thermomechanical processing is remarkable when it was applied to Nb-bearing steels16~ as that used in the present study. In the schedule mentioned above, heating at 880 C and intercritical annealing (holding at T for 15 min followed by iced brine quenching) were carried out using a neutral salt bath. 2. Metallography For metallographic examinations by optical microscope and scanning electron microscope, surface layer of the specimens were removed to avoid the effects of decarburization during heat treatment using the solution of HF(5%)+H202(85%)+H20(10%). The standard etchant 3 % nital and Le Pera's solutionl7) were used to reveal the microstructure and the morphologies of the constituents of the dual-phase specimens. The volume fraction of a' constituents (fm) was determined by using an image analyser for optical micrographs. A part of the specimens (intercritically annealed at 740 C) in each series were deformed to the strain ranges before or after the commencement of local necking. These specimens were also examined by scanning electron microscope and transmission electron microscope (TEM). For transmission electron microscopy, 150 ~cm thick foil were prepared by the chemical polishing. 3 mm diameter disks were cut from these foils and then jet polished by the solution of HC104 5 ml+ch3cooh 75 ml. The final electrolytic polishing was carried out using the solution of HC104 5 ml+ CH3COOH 20 ml. 3. Tensile Test Tensile tests were performed at room temperature in an Instron type machine with a crosshead speed of 0.5 mm/min. Yield strength was taken at 0.2 % offset, and uniform elongation was determined by the point at which the load began to decrease, i.e., the onset of necking. III. Experimental Results 1. Metallography Microstructures of the specimen of series A to C before intercritical annealing were shown in Photo. 1. They are considerably different from each other due to the wide differences in their prior r grain size. The results of examination of these specimens by TEM, however, showed that they consist of typical low carbon lath a' with no remarkable difference in their structure. Mean prior r grain diameter, dr, of the specimens of series A, B and C were determined to be 405, 12 and 4 pm, respectively. Typical dualphase C for structure 15 min and revealed etched in by the Le Pera's specimen solution17~ held at were 740 shown in Photo. 2, in which a' appears white while a appears tan. In the specimen of series A, coarse dualphase structure consisting of fibrous a' and a was obtained (Photo. 2(a)). Such a morphology is similar to that reported by Kim and Thomas.14> On the other hand, in the specimen of series C, a characteristic fine dual-phase structure consisting of homogeneously dispersed fine a' particles (ua'=1.4 pm) and fine a grains was produced as shown in Photo. 2(c). In the specimens of series B, structure of a+a' mixture seems to be a intermediate feature between those obtained in the specimens of series A and C (Photo. 2 (b)). Average diameter of a grain in the specimens of series B and C were determined to be 4.4 and 2.8 pm, respectively. (In the case of series A, a grain size could not be evaluated.) Microstructure of the same specimens etched by 3 % nital were shown in Photo. 3 which clearly indicates the structural differences among the specimens of series A to C. The dual-phase structures produced by holding at higher temperature such as 760 C were shown in Photo. 1. Optical micrographs (etched by 3 % nital) of as quenched structure observed in the s pecimens prior to the intercritical heat treatment.

3 (650) Transactions ISIJ, Vol. 24, 1984 Photo. 2. Optical micrographs of intercritically annealed (at 740 C) structure. (etched by Le Pera's solution17)) Photo. 3. Optical micrographs of intercritically annealed (at 740 C) structure. (etched by 3 % nital) Photo. 4. Optical micrographs o f intercritically annealed (at 760 C) structure. (etched by Le Pera's solution17)) Photo. 4. Little morphological change caused by the increase in holding temperature is recognized in each series except for slight increase in the connectivity of a' constituent due to increase of fm. Transmission electron microscopic examination of the dual-phase specimens prepared by holding at 740 C reveals the existence of some sub-boundaries in the a region as shown in Photo. 5. These sub-boundaries are considered to be formed by the recovery process of lath boundaries of original a' during holding at 740 C.18) A small amount 4f residual r (rr) was also observed in each specimens. There have been many reports19-23) on the behavior of rr in dual-phase steel. In the present study, X-ray diffraction241 was carried out to determine the volume fraction of rr in these specimens. However, the content of rr was too small (less than 3 %) to determine the value of volume fraction quantitatively, ti 2. Tensile Properties The uniform elongation (su) and the total elongation (8T) obtained from the results of the tensile tests agreement with the previous 1nvestlgations,11'12,14,25~ of the specimens of each series were summarized, respectively, in Figs. 2 and 3, as a function of fm. In both su and ~T decrease almost linearly with increasing fm in each series. At given value of fm, series C gives the highest value of CU and ET, over the whole range of fm examined. Figure 4 shows the tensile strength (0B) and the yield strength (60,2) as a function of fm. At given value of fm, series C indicates the highest value of 0B, while series A indicates the lowest value of 0B. In consistent with the previous investigations,3,s,11,12,25-27~ 0B increases almost linearly with increasing fm in each series. 00,2 also increases with increasing fin. In this case, however, the degree of increase in series C is larger than in those of series B and A. Such a trend is probably related with the difference in the connectivity of a' constituent with increasing fm in respective series. Further studies are required about this point. Yield-to-tensile ratio, Qo.2l QB, in series A, B and C were evaluated to be in the range of , and , respectively.

4 Transactions ISIJ, Vol. 24, 1984 (651) Fig. 2. The relation between uniform elongation volume fraction of martensite (fm). and the Fig. 3. The relation between total elongation and the volume fraction of martensite (fm). Photo. 5. Transmission electron micrographs obtained from the intercritically annealed (at 740 C) specimens. 3. Metallography of the Strained Specimens Photograph 6 shows the scanning electron micrographs obtained from the specimens held at 740 C and deformed to the maximum homogeneous strain (5.4, 7.0 and 11.3 %, respectively, for the specimens of series A, B and C) where the local necking was not still commenced. In series C, a' particles seem to be almost undeformed (Photo. 6(c)), while more or less deformed a' constituent are observed in series A and B (Photos. 6(a) and (b)). Similar examination for necked region of the specimens deformed to the higher strain levels showed that a' particles in series C are also deformed to some extent (Photo. 7). Void formation which occurs mainly around small particles of nonmetallic inclusions or a-a' interface25~ was also observed in each specimen. However, the number of void formed in series C is very small in comparison with those formed in series A and B. Iv. Fig. 4. The value of 0.2 % offset flow stress (60,2) and tensile strength (d'b) as a function of the volume fraction of martensite (fm). Discussion 1. Effects of Prior r Grain Size on Dual phase Structure There is a remarkable difference in the morphology of dual-phase structure between the specimens of series A and C in the present study. Such a difference may be caused by the marked difference in their prior r

5 (652) Transactions ISIJ, Vol. 24, 1984 Photo. 6. Scanning electron micrographs o to the maximum homogeneous btained from the elongation (ru). intercritically annealed (at 740 C) specimens deformed up (Arrows indicate the tensile direction) Photo. 7. Scanning electron micrographs obtained from specimens. the necked region of the intercritically annealed (at 740 C) Photo. 8. Optical micrographs showing the c hange in structure of series A and C d urm g intercritical heat treatment at 740 C. (etched by Le Pera's solutionl7)) grain size which presumably have an effect on the behavior of r formation during intercritical annealing. From this point of view, extensive studies were carried out by means of optical microscopy of the specimens quenched after holding at 740 C for various short time duration, in order to reveal the difference in the behavior of r formation in the original a' between the specimens of series A and C. The results obtained are shown in Photo. 8. The behavior of r formation in each series can be deduced from a' which was transformed (by quenching) from r formed during holding at 740 C. In the case of series A (specimens with coarse prior r grain size: dr=405 hem), precipitation of r occurs preferentially along prior r grain boundaries and packet boundaries during the most early stages of holding (Photo. 8(a)). These precipitates grow in size and increase in number, and are combined with each other to form fibrous precipitates. At this stage, prior r grain boundaries and packet boundaries which are believed to provide the most favorable sites for r nucleation seem to be saturated, and other type of r precipitates which have somewhat acicular morphology are found to be aligned along block or lath boundaries of the original a' (Photo. 8(b)). Although these boundaries provide less favorable sites for r nucleation comparing with prior r grain boundaries, such

6 Transactions Is", Vol. 24, 1984 (653) acicular precipitates also increase in number and grow to form coarse precipitates having fibrous morphology during prolonged holding (Photo. 8(c)). It can be recognized that the growth of such two types of fibrous r precipitates in series A produce the final dual-phase structure (Photo. 2(a)) consisted of totally fibrous a' and a. In the case of series C (specimens with ultra-fine grain size: dr=4,gym), fine r particles appear along prior r grain boundaries during verly early stages of holding (Photo. 8(d)). These particles grow in size and increase in number with increasing holding time as shown in Photos. 8(e) and (f). During such a process, precipitation within prior r grains are scarcely observed. Namely in this case saturation of r at the most favorable sites for nucleation was not attained during the early stages of holding because of the remarkably large total prior r grain boundary area depending upon the ultra-fine prior grain size of the specimens. Moreover, mean interparticle distance of r precipitates distributed along the prior r grain boundaries is short enough to form the r by nealy equiaxial growth without further nucleation of r within the prior r grains. Thus, the final dual-phase structure of series C (Photo. 2(a)) shows the characteristic fine structure consisting of homogeneously dispersed fine a' particles and fine a grains. 2. Tensile Strength and Ductility Figure 5 shows total elongation as a function of tensile strength for the dual-phase specimens of series A to C. It is notable that series C show superior elongation and higher tensile strength in comparison with series A and B. These results are probably due to the morphological features of series C. High tensile strength may be caused by the fine a grain size. Tanaka et a1.28> reported that yield stress, flow stress and tensile strength of dual-phase steel obey Hall- Petch relation (a=a2+kd-1~2), and the grain size dependency, K, of tensile strength has comparatively high value. Photograph 9 shows the transmission electron micrographs obtained from the specimens held at 740 C and deformed to the range of tensile strain (5.4, 7.0 and 11.3 %, respectively, for the specimens of series A, B and C) where the local necking has not still commenced. Many dislocation pile-ups are observed at a sub-grain boundaries in each specimen. These results suggest that the formation of a sub-boundary previously pointed out in Photo. 5 is beneficial to increase the strength of dual-phase specimens of each series through the effects of grain refinement. As mentioned above, high tensile strength of the dual-phase specimens of series C can be attributed to their fine grain size. In such a specimen with fine a grain size, a constituent itself presumed to have high strength. It has been pointed out that at a given strength of the dual-phase structure, the strength of the a constituent is the most important factor in governing the ductility.29f In fact, dual-phase specimens with fine a grain size (series C) show superior elongation. Recently, on the other hand, Balliger and Gladman6~ applied Ashby's work-hardening theory30~ to Fig. 5. The relation strength. between total elongation and tensile Photo. 9. Transmission electron micrographs obtained from the intercritically annealed (at 740 C) specimens deformed up to the maximum homogeneous elongation (eu).

7 ( 654) Transactions ISIJ, Vol. 24, 1984 dual-phase steels based on their observed fact that, during tensile deformation, a' island do not deform until strain well in excess of the maximum uniform strain is reached. They concluded that the workhardening rate of the dual-phase steel is dependent on (fm/d)'/2, where d is the mean a' island diameter. Thus they demonstrated that optimization of the properties of dual-phase steel may be achieved by refining the distribution of a', since this lead to increase workhardening rates and hence increase maximum uniform ductility. Lanzillotto and Pickering27> also analysed the flow stress and work-hardening characteristics of dual-phase steels using the Ashby's dispersion-strengthening model30~ modified by Brown and Stobbs31 and obtain the similar conclusion given by Balliger and Gladman.s~ The fact that, in the present study, the specimens of series C, in which mean a' particle diameter is remarkably small (e.g., d=1.4 pm in the specimen held at 740 C) show considerably large elongation is well supported by the conclusion that derived from their analysis. From these consideration, it is emphasized that the better combination of strength and ductility can be achieved by producing the dual-phase structure consisting of fine a' particles dispersed homogeneously and fine a grain as these in the specimens of series C in the present study. V. Conclusions Microstructure and the mechanical properties of dual-phase Fe-2.3 %Mn-0.05 %C-0.03 %Nb steel prepared by heating the specimens in a martensitic state with various prior austenite grain size into (ferrite+ austenite) region were tested. The results obtained were summarized as follows : (1) The specimen with large prior austenite grains (405 pm) produced the coarse dual-phase structure consisting of fibrous martensite and ferrite. In contrast to such a morphology, the specimens with ultrafine prior austenite grains (4 pm) produced the characteristic fine dual-phase structure consisting of homogeneously distributed fine martensite particles and the fine ferrite grains. (2) The specimens with fine dual-phase structure were superior; they showed higher strength and, at the same time, higher ductility compared with the specimens of coarse dual-phase structure having the same volume fraction of martensite. (3) The sub-boundaries presumably formed by prior lath boundaries were found in the ferrite matrix of the dual-phase structure prepared by the present method. The presence of such sub-boundaries may be a cause of strengthening of the present dual-phase steels. Acknowledgements The authors are grateful to Central Laboratory, Kobe Steel, Ltd., for providing the steel used in the present study. This work was ported by Grant-in-aid ( ) for search from the Ministry of Education. REFERENCES partially Scientific sup- Re- 1) Y. Tomota, K. Kuroki, T. Mori and I. Tamura : Mat. Sci. Eng., 24 (1976), 85. 2) K. Araki, Y. Takada and K. Nakaoka : Trans. ISIJ, 17 (1977), ) R. G. Davies : Met. Trans., 9A (1978), ) G. R. Speich and R. L. Miller: Proc. of Symp, on Structure and Properties of Dual Phase Steels, RIME, New York, (1979), ) H.K.D.H. Bhadeshia and D. V. Edmonds: Metal Sci., 14 (1980), 41. 6) N. K. Balliger and T. Gladman : Metal Sci., 15 (1981), 95. 7) C. M. Wan, S. N. Yie, M. T. Jahn and S. M. Kuo: J. Mat. Sci., 16 (1981), ) P. Ostrom: Met. Trans., 12A (1981), ) I. Tamura, Y. Tomota and M. Ozawa: Proc, of the 3rd Int'l Conference on the Strength of Metals and Alloys, I, Inst. Metals and Iron Steel Inst., Cambridge, (1973), ) J. Y. Koo and G. Thomas: Mat. Sci. Eng., 24 (1976), ) J. Y. Koo and G. Thomas: Met. Trans., 8A (1977), ) G. Thomas and J. Y. Koo: Proc. of Symp, on Structure and Properties of Dual Phase Steels, RIME, New York, (1979), ) J. Y. Koo and G. Thomas : Script. Met., 13 (1979), ) N.J. Kim and G. Thomas: Met. Trans., 12A (1981), ) M. Tokizane, N. Matsumura, K. Tsuzaki, T. Maki and I. Tamura: Met. Trans., 13A (1982), ) M. Tokizane and N. Matsumura: Tetsu-to-Hagane, 67 (1981), ) F. S. Le Pera: J. Metals., 32 (1980), ) N. Matsumura and M. Tokizane: Trans. JIM., 23 (1982), ) Narasimha-Rao, V. Bangaru and A. K. Sachdev: Met. Trans., 13A (1982), ) K. Kunishige, M. Takahashi, S. Sugisawa and Y. Masui: Tetsu-to-Hagane, 65 (1979), ) A. Nakagawa, J. Y. Koo and G. Thomas: Met. Trans., 12A (1981), ) H. Takechi: J. Japan Soc. for Technology of Plasticity, 21 (1980), ) T. Furukawa, H. Morikawa, M. Endo, H. Takechi, K. Koyama, 0. Akisue and T. Yamada : Tetsu-to-Hagane, 68 (1982), ) R. L. Miller: Trans. ASM, 57 (1964), ) A. R. Marder : Met. Trans., 13A (1982), ) L. F. Ramos, D. K. Matlock and G. Krauss: Met. Trans., 10A (1979), ) C.A.N. Lanzillotto and F. B. Pickering: Metal. Sci., 16 (1982), ) T. Tanaka, M. Nishida, K. Hashiguchi and T. Kato : Proc, of Symp, on Structure and Properties of Dual Phase Steels, RIME, New York, (1979), ) R. G. Davies: Met. Trans., 9A (1978), ) M. F. Ashby: Phil. Mag.,14 (1966), ) L. M. Brown and W. M. Stobbs : Phil. Mag., 23 (1971), 1185.