Microstructure Evolution in Ferritic Stainless Steels during Large Strain Deformation

Transcription

1 Materials Transactions, Vol. 45, No. 9 (2004) pp to 2821 #2004 The Japan Institute of Metals Microstructure Evolution in Ferritic Stainless Steels during Large Strain Deformation Andrey Belyakov*, Yuuji Kimura, Yoshitaka Adachi and Kaneaki Tsuzaki Steel Research Center, National Institute for Materials Science, Tsukuba , Japan Deformation microstructures were studied in ferritic stainless steels during cold bar rolling and swaging to total true strains about 7. Two steels, i.e. Fe-22Cr-3Ni and Fe-18Cr-7Ni with coarse-grained ferritic and fine-grained martensitic initial microstructures, respectively, were selected as starting materials. Microstructure evolution in the both steels was characterized by the development of highly elongated (sub)grains aligned along the rolling/swaging axis. The transverse size of these (sub)grains in the Fe-22Cr-3Ni steel gradually decreased to about mm with increasing the strain. On the other hand, the transverse (sub)grain size in the Fe-18Cr-7Ni steel decreased to its minimal value of 7 mm with straining to about 3 followed by a little coarsening under further working. The strengthening of worked steels that revealed by hardness tests correlated with the microstructure evolution. The hardness of the Fe-22Cr-3Ni steel increased with cold working within the studied strain range, while that of the Fe-18Cr-7Ni approached a saturation after fast work hardening at strains below 3, leading to an apparent steady-state behaviour. Development of strain-induced (sub)grain boundaries and internal stresses in the steels with different initial microstructures during severe deformation is discussed in some detail. (Received May 11, 2004; Accepted July 12, 2004) Keywords: ferritic stainless steels, severe plastic deformation, strain hardening, grain refinement, grain boundaries, internal stress 1. Introduction Severe plastic deformations aroused a great interest among materials scientists and metallurgical engineers as a specific method for processing of various structural alloys. This interest was mainly motivated by promising structural state with a grain size of submicron scale that could be developed in almost all metals and alloys under large strain plastic working. 1 6) Several techniques were proposed to provide severe plastic working, 7 14) including mechanical milling, torsion under high pressure, multiple forging, equal channel angular pressing, accumulative roll-bonding, etc. The structural changes leading to the submicrocrystalline states during severe deformation could be associated with the formation of strain-induced dislocation subboundaries, namely geometrically necessary boundaries; then a gradual increase in the misorientations between the strain-induced subgrains upon further straining resulted in the development of new ultra fine-grained microstructures. 3,4,15 21) The most of processing methods mentioned above are multiple techniques that based on changing (or reversing) the strain path for each cycle. Such multiple character of straining stimulates the variation of deformation mechanisms operating in each strain pass and seems to play an important role in the formation of rather equiaxed fine grains. Large strain deformation can be also achieved by some conventional methods, such as rolling and drawing, which are simple in applications and can be easily used for processing of ductile metals and alloys. However, during such processing, the formation of strain-induced ultra fine grains was hardly observed in pseudo single-phase metallic materials irrespective of large imposed strains. The severely rolled (drawn) microstructures were characterised by the ribbon-like grains and subgrains, which are highly elongated *On leave from the Institute for Metals Superplasticity Problems, Ufa, Russia in the rolling direction. 17,22 26) The transverse size of the elongated (sub)structural elements was shown to decrease with increasing the drawing strain up to the largest studied strain about 6. This could be simply explained by two reasons: reduction in the cross section of original grains that followed the change of sample geometry, and the development of new strain-induced (sub)boundaries, which aligned parallel to the rolling direction after sufficiently large strains. 3,17,25) It should be noted that the transverse size of structural elements could approach some constant value at large strains following a rapid reduction during preceding deformation. Such behaviour was observed in certain studies on plane rolling of two-phase alloys, when the shear banding frequently operated during plastic working. 27) Another interesting topic, which should be considered for severe plastic deformations, is the strain hardening. In the most of researches, the severe plastic working was shown to result in the strengthening of materials that revealed by hardness and/or tensile tests ,28 30) After fast work hardening at early deformation, the hardening rate generally tended to decrease with increasing strain, approaching, in rare cases, a steady-state-like behaviour at large strains. However, the steady-state behaviour was observed only for torsion straining and/or for processing at relatively high homologous temperatures, when recovery could operate during (or after) deformation. The structural processes developing during large strain deformations could be remarkably accelerated by refinement of initial microstructures. 13,31) During warm multiple deformations, the steady-state flow behaviour corresponding to almost complete evolution of new ultra fine-grained microstructures appeared at smaller strains in the samples with finer initial grain sizes. The aim of the present work is to study the deformation behaviour and to clarify the effect of initial microstructure on the strengthening and the structural changes during severe straining by conventional metalforming methods.

2 Microstructures in Ferritic Stainless Steels under Severe Deformation Experimental Procedure Two Cr-Ni stainless steels, Fe-02%C-1%Mn- 05%P-01%S-21.99%Cr-3.12%Ni-02%N and Fe- 08%C-1%Mn-04%P-01%S-18.05%Cr-6.99%Ni- 01%N (all in mass%), designated as 22-3 and 18-7 steels, respectively, were vacuum melted and cast into the 20 kg ingots followed by hot forging and homogenization at 1200 C. Then, the steels were hot rolled to 21.3 mm 2 squared bars at 700 C followed by air cooling. The initial hot rolled microstructures are shown in Fig. 1. The 22-3 steel was characterised by coarse-grained ferritic microstructure with a transverse grain size about 0.7 mm (1.4 mm as a subgrain size), and the 18-7 steel obtained a fine martensitic structure with an average (sub)grain size of 230 nm and a fraction of high-angle (sub)grain boundaries about 50% that resulted from! 0 phase transformation at around 100 C. (Note that fraction of retained austenite was about 10%.) For simplicity, a critical misorientation angle between low- and high-angle (sub)grain boundaries was selected to be around 15. Severe plastic working was carried out by rolling to 7.8 mm 2 squared bars followed by swaging from 6.0 to 0.55 mm rods at ambient temperatures, providing a total strain of 7.1. The metallographic analysis was carried out using an optical microscope and JEM-2010F transmission electron microscope (TEM). Almost all microstructural observations were carried out on sections parallel to the rolling/swaging axis. The transverse (sub)grain sizes were measured by a linear intercept method perpendicularly to the rolling/ swaging axis. All the clearly defined (sub)grain boundaries Fig. 1 Initial microstructures of (a) 22%Cr-3%Ni and (b) 18%Cr-7%Ni stainless steels. were taken into account to determine the (sub)grain sizes and the (sub)boundary misorientations. Misorientations of the strain-induced (sub)boundaries were analysed by using a conventional Kikuchi-line technique. 32) The (sub)boundary misorientations were collected from at least four arbitrary selected typical areas for each specimen. A total of (sub)boundaries were analysed for each strain level studied. The dislocation density was measured by counting the individual dislocations in (sub)grain interiors on at least four arbitrary selected typical TEM images for each sample. Textures and internal distortions were studied by X-ray diffraction from [110]-type planes with a RINT-2200/2500 diffractometers using a Cu target. The internal distortions were estimated as the slope of the Hall-Williamson plot. 33) Strain hardening was studied by means of the Vickers hardness test with a load of 3 N. 3. Results 3.1 Strain hardening and deformation microstructures Typical deformation microstructures that developed in the steels under bar rolling/swaging to various strains are shown in Figs. 2 and 3 for the 22-3 and 18-7 samples, respectively. Generally, cold uniaxial deformation results in the evolution of highly elongated grains that aligned along the deformation axis. Deformation substructures depend strongly on the intensity of plastic working. At moderate strain about 2, the deformation substructures in the 22-3 samples with initial coarse-grained ferritic microstructure look like typical coldworked ones 15,17,20) and consist of cell blocks separated by strain-induced geometrically necessary subboundaries such as dense dislocation walls, microbands, etc. (Fig. 2a). Upon further deformation, the strain-induced subboundaries arrange almost parallel to the swaging axis, leading to ribbonlike substructures at rather large strains (Fig. 2b). It is interesting that the severely deformed microstructures after straining to 7.1 that revealed by optical microscopy appear different from those at preceding strains and look like ultra fine-grained ones with a grain size below 1 mm. Formation of ultra fine (sub)grains in severely strained samples is clearly seen in TEM micrographs (Fig. 2c). These (sub)grains are certainly elongated in the swaging direction; however, the presence of them modifies remarkably the ribbon-like deformation substructures developed at preceding strains. In the 18-7 samples with initial fine-grained martensitic microstructure, the general sequence of structural changes during deformation is similar to that in the coarse-grained 22-3 samples. However, the evolution of characteristic ribbonlike substructure in the 18-7 sample takes place at smaller strains (cf. Figs. 2a and 3a). Also, severe deformation results in the development of many transverse dislocation (sub)- boundaries, leading to the evolution of relatively equiaxed ultra fine (sub)grains at strain of 7.1 (Fig. 3c). Effect of the cold working on the hardness and the transverse (sub)grain size is presented in Fig. 4. The 22-3 steel shows fast work hardening at early deformation, then the hardening rate decreases to some value leading to gradual increase in the hardness with further deformation to strains as large as 7.1. On the other hand, the cold deformation of the 18-7 steel is characterised by continuous work hardening

3 2814 A. Belyakov, Y. Kimura, Y. Adachi and K. Tsuzaki Fig. 2 Deformation microstructures in 22Cr-3Ni steel after bar rolling/swaging to total strains of 2.0 (a), 3.2 (b), and 7.1 (c). Selected portions in TEM micrographs are represented in Fig. 6. until a moderate strain around 3, where the hardness approaches some saturation and does not change remarkably at larger strains. The strain dependence of the transverse subgrain size clearly correlates with the strain hardening. The transverse subgrain size smoothly reduces down to about 100 nm in the 22-3 steel with deformation within studied strains, while that in the 18-7 steel reaches its minimum value about 70 nm at strains around 3 followed by little coarsening upon further straining. In other words, the 18-7 steel with initial fine-grained martensitic structure apparently demonstrates steady-state deformation behaviour at strains above 3, when the transverse subgrain size and the hardness do not change with straining. It should also be noted that in the strain range below 56, the transverse grain size is remarkably smaller and the hardness is higher in the 18-7 steel than those in the 22-3 steel, while the deformation microstructures and the hardness at larger strains are almost the same in the both steels. Bar rolling/swaging of the ferritic stainless steels results in development of strong deformation texture, namely, [110] parallel to the rolling/swaging axis, as shown in Fig. 5. This is typical of bcc metallic materials. 25) It should be noted that the difference in the initial microstructures of studied steels does not lead to any remarkable changes in the texture evolution. At small to moderate strains about 2, the texture intensity rapidly increases to rather high saturation value. Upon further deformation, the texture intensity is kept almost constant, although the samples processed to the largest strain of 7.1 can be characterised by somewhat weakened textures. 3.2 Misorientations between strain-induced (sub)grains Figure 6 represents the detailed TEM micrographs with

4 Microstructures in Ferritic Stainless Steels under Severe Deformation 2815 Fig. 3 Deformation microstructures in 18Cr-7Ni steel after bar rolling/swaging to total strains of 2.0 (a), 3.2 (b), and 7.1 (c). Selected portions in TEM micrographs are represented in Fig. 7. misorientations across the strain-induced (sub)boundaries for the deformation microstructures developed in the 22-3 steel during bar rolling/swaging to different strains. The micrographs correspond to those shown in Fig. 2. Deformation substructures evolved at moderate strain of 2 consist of dislocation subboundaries with relatively low-angle misorientations, most of which are below 15. The evolution of the ribbon-like substructure under following straining is accompanied by increase in the misorientations between the highly elongated subgrains, leading to the development of many (sub)boundaries with high-angle misorientations, which are typical of conventional grain boundaries. Then, the severe plastic deformation results in the break down of the ribbonlike (sub)grains by new transverse (sub)boundaries with lowand high-angle misorientations. Similar tendency for the evolution of the (sub)boundary misorientations during severe deformation takes place in the 18-7 steel (Fig. 7). However, comparing the samples processed to the same strains, it is clearly seen that the deformation (sub)structures in the later case are characterised by larger misorientations. The misorientation distributions for the deformation (sub)boundaries that developed during the severe plastic working are shown in Figs. 8 and 9 for the 22-3 and 18-7 steels, respectively. The dashed lines in the figures indicate the theoretical random misorientation distribution. 34) The 22-3 initial coarse-grained ferritic microstructure includes only the low-angle dislocation subboundaries that evolved by previous hot-working. This sharp maximum against lowangle misorientations decreases and spreads out towards larger misorientations during plastic working. After severe deformation to strains above 4, the deformation (sub)boundaries in the 22-3 steel can be characterised by flat-type misorientation distributions with almost equal fractions of (sub)boundaries with different misorientations. On the other

5 2816 A. Belyakov, Y. Kimura, Y. Adachi and K. Tsuzaki 5000 Hv / MPa d / nm Fe-18Cr-7Ni Fe-22Cr-3Ni Total Strain Fig. 4 The strain effect on the hardness (Hv) and the transverse (sub)grain size (d) of the 22Cr-3Ni and 18Cr-7Ni steels. Fig. 6 Typical (sub)structures developed in the 22Cr-3Ni steel under bar rolling/swaging to strains of 2.0 (a), 3.2 (b), and 7.1 (c). Numbers indicate the (sub)boundary misorientations in degrees. Fig. 7 Typical (sub)structures developed in the 18Cr-7Ni steel under bar rolling/swaging to strains of 2.0 (a), 3.2 (b), and 7.1 (c). Numbers indicate the (sub)boundary misorientations in degrees. Fig. 5 Fragments of [110] pole figures as related to the rolling/swaging axis for the 22Cr-3Ni and 18Cr-7Ni steels after plastic working to various strains. The contours indicate the orientation distribution densities for levels of etc. hand, the misorientation distribution in the 18-7 initial martensitic microstructure is characterised by two maximums corresponding to low- and high-angle misorientations that resulted from the phase transformation. The both maximums are weaken by the plastic working to a moderate strain about 3, leading to a flat-type misorientation distribution much similar to that developed in 22-3 steel after

6 Microstructures in Ferritic Stainless Steels under Severe Deformation 2817 Fraction, N i / N ε = 3.2 ε = ε = 1.0 ε = 2.0 ε = 4.4 ε = Misorientation, θ Fig. 8 Misorientation distributins for the (sub)grain boundaries evolved in the 22Cr-3Ni steel under bar rolling/swaging to various strains. Fraction, N i / N ε = ε = 1.0 ε = 2.0 ε = 3.2 ε = 4.4 ε = Misorientation, θ Fig. 9 Misorientation distributins for the (sub)grain boundaries evolved in the 18Cr-7Ni steel under bar rolling/swaging to various strains. deformation to larger strains. It is interesting that further deformation of 18-7 steel results in increase in the fraction of high-angle (sub)boundaries with misorientations about 60. Therefore, contrary to the random misorientation distribution, the deformation microstructures in the 18-7 steel that processed by the large strain bar rolling/swaging can be characterised by a specific distribution of (sub)boundary misorientations with relatively large fractions of low-angle misorientations below 15 and high-angle misorientations above 50. Figure 10 shows the crystallographic directions for the (sub)boundary misorientations that evolved in the both steels. Deformation (sub)boundaries with misorientations smaller than 5 were omitted in this figure from methodical considerations. At relatively small strains, the misorientation axes are almost randomly distributed within the standard triangle. After severe deformation, however, the misorientation axes tend to appear near two crystallographic directions, i.e. [110] and [111]. This tendency can be clearly recognised in the 18-7 steel at strains above 4 and seems to be associated with the strong deformation texture (s. Fig. 5). Effect of initial microstructures on the misorientations of deformation (sub)boundaries is summarised in Fig. 11. For the both steels, the fraction of high-angle (sub)grain boundaries (HAB) in deformation structures increases with strain, although the evolution kinetics of the high-angle boundaries are different. In the 22-3 steel, the fraction of HAB increases from almost zero to 7080% with plastic working to a strain above 4, and then, the fraction of HAB of 7080% does not change significantly during further straining. On the other hand, the initial microstructure of the 18-7 steel includes a relatively large fraction of HAB about 50%. This value is kept constant during deformation to a moderate strain of 23 followed by some increase to a saturation of 7080% at strains above 3. Therefore, in the sample with finer initial microstructure, the fraction of HAB is larger after plastic working to a moderate strain (below 4 in the present study). However, the fraction of HAB in the steel microstructures approaches a saturation of 7080% during severe deformation to a sufficient strain irrespective of initial microstructures. 3.3 Dislocation density and internal distortions Commonly, cold working of metallic materials increases the internal stresses, which may be reflected by increasing the density of any lattice defects. However, the TEM micrographs in Figs. 6 and 7 suggest that the density of dislocations in (sub)grain interiors decreases unusually at moderate strains, following a rapid rise during early deformation. The quantitative results for the measurement of the density of interior dislocations and the internal distortions that obtained in the studied steels are presented in Fig. 12. In the both

7 2818 A. Belyakov, Y. Kimura, Y. Adachi and K. Tsuzaki Fe-22Cr - 3Ni ε = 1.0 Fe-18Cr-7Ni ρ / m ε = 4.4 X-Ray / 10-3 ε Fe-18Cr-7Ni Fe-22Cr-3Ni Total Strain ε = 7.1 Fig. 10 Inverse pole figures of the (sub)grain boundary misorientation axes for the 22Cr-3Ni and 18Cr-7Ni steels after bar rolling/swaging to various strains. Fraction of HAB (%) Fe-18Cr-7Ni Fe-22Cr-3Ni Total Strain Fig. 11 Effect of bar rolling/swaging on the fraction of high-angle deformation (sub)boundaries (HAB) in the 22Cr-3Ni and 18Cr-7Ni steels. steels, the dislocation density quickly increases to its maximum after deformation to relatively small strains of 12. Then, the dislocation density remarkably decreases at moderate strains around 4 followed by repeated increase under further severe working. In spite of similarity in the variations of dislocation densities in both steels during Fig. 12 The dislocation density in (sub)grain interiors () and the internal distortions (" X-Ray ) in the 22Cr-3Ni and 18Cr-7Ni steels upon plastic working by bar rolling/swaging. deformation, the distinguishing changes in the density of interior dislocations occur faster in the 18-7 than in the 22-3 steel. Qualitatively, the evolution of the internal distortions during the plastic working is also the same in the both steels. Contrary to the dislocation densities, the internal distortions increase gradually with straining. The rate of increase of internal distortions is almost the same in the both steels and it is roughly a constant within the all strain range studied. However, the level of internal distortions in the 18-7 steel is higher than that in the 22-3 steel for all analysed samples. This may be associated with the initial martensitic structure in 18-7 steel that is certainly characterised by the presence of residual stresses. Therefore, the present plastic working by bar rolling/ swaging to large strains results in the evolution of increasing internal stresses, irrespective of apparent decrease in the dislocation density at moderate strains. Decreasing the dislocation density that accompanied by the development of high internal stresses was also mentioned in some studies on grain refinement by severe plastic deformation. 4,10,19,31,35) The origin of such high internal stresses could be attributed to the large lattice curvatures associated with a non-equilibrium state of strain-induced (sub)grain boundaries. Figure 13 shows high resolution TEM image of triple junction of deformation (sub)boundaries that observed on the transverse section of the 18-7 steel sample processed to a strain of 3.2. This sample is characterised by a relatively low density of interior dislocations (s. Fig. 12). It is clearly seen in Fig. 13 that the lattice planes near the boundary junction are smoothly bent through about 3.5. Such lattice curvature could be accommodated with a dislocation wall with a spacing between dislocations (l) about 4 nm, which should correspond to a rather high density of interior dislocations () of 2: m 2 within a (sub)grain size (d) about 100 nm, i.e. ¼ðldÞ 1. 36) It should be noted that such an imaginary dislocation density is remarkably higher than that counted experimentally in the strain-induced (sub)grains. The lattice curvature shown in Fig. 13 may be considered as a specific

8 Microstructures in Ferritic Stainless Steels under Severe Deformation 2819 Transverse (Sub)grain Size, d / µm 1 Fe, 6T m (Embury et al. [22]) Fe, 6T m (Langford, Cohen [23]) Al, 1T m (Liu et al. [26]) Cr, 9T m (Ball et al. [24]) Cu, 2T m (Embury et al. [22]) Fe-22Cr-3Ni, 6T m Fe-18Cr-7Ni, 6T m Fig. 13 High resolution TEM image of the [110] bent lattice planes near a triple junction of (sub)grain boundaries in the 18Cr-7Ni steel processed by bar rolling/swaging to a total strain of 3.2. crystallographic defect called the wedge disclination, 37,38) and then, the internal stress can be roughly estimated as follows: ¼ 0:1!G, or ¼ G, where! and G are the Franc vector (rotational distortion) and the shear modulus, respectively. Therefore, some decrease of internal stresses that may result from decreasing the density of interior dislocations is compensated by the strong internal distortions associated with the strain-induced (sub)boundaries, leading the internal stresses to gradual increase during large strain deformation Total Strain Fig. 14 Relationship between the large strain deformation and the transverse (sub)grain size in various materials. 4. Discussion The present results suggest that the microstructure evolution in ferritic steels during severe plastic working is greatly affected by the initial microstructures. The general sequence of the structural changes taking place under large strain deformation follows common one reported for various metallic materials ) However, refinement of the initial microstructure can significantly accelerate kinetics of the change of almost all microstructural parameters during plastic working. Using the initial fine-grained martensitic microstructure, it was possible to attain the steady-state-like behaviour during cold working at relatively small strains above 3. The strain dependence of the transverse (sub)grain size (d) that obtained in the present study is represented in Fig. 14 in semi-logarithmic coordinates with reference to some arbitrary selected literature data for almost pure single-phase metals with relatively large initial grains ,26) Assuming the shape change of the (sub)grains during bar rolling/swaging corresponds to that of the whole sample, the log d should vary linearly with strain with the slope of 0:5 (the slope equals 1 for a plane-strain compression/rolling). 23) It is clearly seen in Fig. 14 for almost all data that the transverse (sub)grain size follows the shape change of the bulk sample during processing only at relatively small strains around 1, after rapid decrease at early deformation, when many dislocation subboundaries evolve in deformation substructures. Upon further straining, the transverse (sub)grain size Fig. 15 Schematic drawing of the (sub)structural changes taking place under severe deformation by bar rolling/swaging. decreases much slower than the change of the sample shape. Moreover, in the 18-7 steel, the transverse (sub)grain size becomes almost constant at strains above 3. This suggests a specific mechanism of microstructure evolution that operates in ribbon-like substructures at large strains during cold rolling/swaging. Adjacent boundaries of highly elongated (sub)grains may pinch off at sufficiently large reductions in cross section of samples, leading the ribbon-like (sub)grains to rupture, as schematically illustrated in Fig. 15. During severe deformation, such structural mechanism operating concurrently with subgrain coalescence by the plastic hinge mechanism 23) can result in the constant transverse (sub)grain size as well as the development of somewhat elongated fine grains after large strains. It is interesting that the fraction of high-angle (sub)grain boundaries in the processed steels saturates at about 7080% after deformation to strains about 4 and, contrary to expectations, does not increase with further straining. Figure 16 combines the point-to-point misorientations with

9 2820 A. Belyakov, Y. Kimura, Y. Adachi and K. Tsuzaki 60 cumulative 3000 Misorientation, θ point-to-point Distance, d / nm high-angle boundaries low-angle boundaries Fig. 16 Point-to-point and cumulative (point-to-origin) misorienations in the 18Cr-7Ni steel (sub)structure evolved at a strain of 3.2. The data correspond to those in Fig. 7b (right side, from top to bottom). point-to-origin (cumulative) ones that developed in the 18-7 steel at strain of 3.2. The data were collected from Fig. 7b, from top to bottom on the right-hand side. It is clearly seen that the (sub)grains may have mutually opposite misorientations. The fraction of high-angle misorientations among the cumulative misorientations is even smaller than that for the point-to-point misorientations in Fig. 16. In other words, the (sub)grains located far away from each other may have the close crystallographic orientations. Therefore, decrease in the number of (sub)grains in the cross section of sample, i.e. disappearance of some (sub)grains by the structural mechanism discussed above, should not necessary lead to increase in the fraction of high-angle (sub)boundaries. Another interesting feature in the misorientation distribution for the strain-induced (sub)grain boundaries after severe deformation is a large fraction of misorientations of about 60. The same feature was observed in some other studies on deformation microstructures after large strain processing. 26,39) This may result from the strong [110] fibre texture. In fact, the misorientation between two grains rotating about [110] through 6090 can be represented by the equivalent minimal misorientation about 60 with the misorientation axes close to [111]. Therefore, the resulting inverse pole figures for the misorientation axes should display two maximums around [110] and [111] crystallographic directions. Finally, let us consider the strengthening mechanism for the ferritic stainless steels processed by large strain bar rolling/swaging. Two mechanisms are generally discussed to explain the hardening by a large strain deformation ,40 42) The substructural strengthening mechanism results in a power-law relationship between the strength/hardness increment and the subgrain size with the size exponent of 1. Another one is the grain strengthening mechanism leading to the similar type grain size dependence for the strengthening, but the grain size exponent is 0:5. The latter is usually referred as the Hall-Petch relationship. In spite of some debates, it is generally agreed that the strengthening mechanism depends on various parameters of deformation Hv = Hv - Hv O / MPa (sub)structures, i.e. (sub)grain size, misorentations between (sub)grains, (sub)boundary character, etc. The relationship between the hardening (Hv) and the (sub)grain size that obtained in the present study is presented in Fig. 17. The Hv O of 1800 MPa was taken from the fully annealed 22-3 steel with the grain size above 1 mm. It is clearly seen in Fig. 17 that all the worked steel samples obey the (sub)grain size exponent about 0:5. Therefore, it can be concluded that the strengthening of the ferritic stainless steels during severe working follows the Hall-Petch-type relationship, irrespective of some differences in the misorientations between the strain-induced (sub)grains. 5. Conclusions Hv O = 1800 MPa d -1 / µm -1 The microstructure evolution during severe plastic deformation by bar rolling/swaging up to large strain of 7.1 at ambient temperature was studied in two Cr-Ni stainless steels, Fe-22%Cr-3%Ni and Fe-18%Cr-7%Ni, with initially ferritic and martensitic microstructures, respectively. The main results could be summarised as follows: (1) Microstructure evolution in the both steels was characterized by the development of highly elongated (sub)grains aligned along the rolling/swaging axis. The steel with initial martensitic structure demonstrated faster kinetics of the structural changes, i.e. decreasing the transverse (sub)grain size, increasing the fraction of high-angle (sub)- boundaries and the internal stresses, than the samples with initial ferritic microstructure. (2) The transverse size of highly elongated (sub)grains in the initially ferritic steel gradually decreased to about mm and the hardness increased to about 4000 MPa during deformation to the largest strain of 7.1. On the other hand, the transverse (sub)grain size in the initially martensitic steel decreased to its minimal value of 7 mm with straining to about 3 followed by a little coarsening under further working; and the hardness rapidly increased to about 4500 MPa followed by a steady-state-like behaviour at strains above 3. (3) In the both steels, the fraction of high-angle misorientations between the strain-induced (sub)grains increased 1 Fe-18Cr-7Ni Fe-22Cr-3Ni Fig. 17 Relationship between the transverse (sub)grain size (d) and the hardening (Hv) for the bar rolled/swaged stainless steels. 10

10 Microstructures in Ferritic Stainless Steels under Severe Deformation 2821 to 7080% with straining to 34 and then did not vary significantly upon further working. The misorientation distribution of deformation (sub)boundaries that evolved by severe straining was characterised by a remarkably large fraction of high-angle misorientations about 60. (4) The work-hardening followed the grain strengthening mechanism, the Hall-Petch relationship, in the all studied samples regardless of some differences in the (sub)grain boundary characteristics. Acknowledgements The authors are grateful to Drs. N. Sakuma, T. Hibaru, S. Kuroda, M. Kobayashi, Metallurgical Processing Group and Dr. T. Kanno, Welding Metallurgy Group, Steel Research Center, National Institute for Materials Science, for their assistance in the materials processing. One of the authors (A.B.) would also like to express his thanks to the National Institute for Materials Science for providing a scientific fellowship. REFERENCES 1) C. Suryanarayana: Int. Mater. Rev. 40 (1995) ) G. A. Salishchev, R. G. Zaripova, A. A. Zakirova and H. J. McQueen: Hot Workability of Steels and Light Alloys-Composites, ed. by H. J. McQueen, E. V. Konopleva and N. D. Ryan, (TMS-CIM, Montreal, 1996) pp ) F. J. Humphreys, P. B. Prangnell, J. R. Bowen, A. Gholinia and C. Harris: Phil. Trans. R. Soc. Lond. 357 (1999) ) R. Z. Valiev, R. K. Islamgaliev and I. V. Alexandrov: Prog. Mater. Sci. 45 (2000) ) D. G. Morris: Science of Metastable and Nanocrystalline Alloys, ed. by A. R. Dinesen, M. Eldrup, D. Juul Jensen, S. Linderoth, T. B. Pedersen, N. H. Pryds, A. Schroder Pedersen and J. A. Wert, (Riso National Laboratory, Roskilde, Denmark, 2001) pp ) Y. Wang, M. Chen, F. Zhou and E. Ma: Nature 419 (2002) ) V. M. Segal, V. I. Reznikov, A. E. Drobyshevskiy and V. I. Kopylov: Russian Metallurgy 1 (1981) ) I. Saunders and J. Nutting: Metall. Sci. 18 (1984) ) J. Richert and M. Richert: Aluminium 62 (1986) ) Y. Kimura and S. Takaki: Mater. Trans., JIM 36 (1995) ) C. C. Koch: NanoStruct. Mater. 9 (1997) ) Y. Saito, N. Tsuji, H. Utsunomiya, T. Sakai and R. G. Hong: Scr. Mater. 39 (1998) ) A. Belyakov, W. Gao, H. Miura and T. Sakai: Metall. Mater. Trans. A 29A (1998) ) D.-W. Suh, S. Torizuka, A. Ohmori, T. Inoue and K. Nagai: Iron Steel Inst. Japan, Int. 42 (2002) ) D. Kuhlmann-Wilsdorf and N. Hansen: Scr. Metall. Mater. 25 (1991) ) R. Kaibyshev and O. Sitdikov: Z. Metallkd. 85 (1994) ) D. A. Hughes and N. Hansen: Acta Mater. 45 (1997) ) Y. Iwahashi, Z. Horita, M. Nemoto and T. G. Langdon: Acta Mater. 45 (1997) ) A. Belyakov, T. Sakai and H. Miura: Mater. Trans., JIM 41 (2000) ) A. Belyakov, T. Sakai, H. Miura and K. Tsuzaki: Philos. Mag. A 81 (2001) ) Yu. Ivanisenko, W. Lojkowski, R. Z. Valiev and H.-J. Fecht: Acta Mater. 51 (2003) ) J. D. Embury, A. S. Keh and R. M. Fisher: Trans. AIME 236 (1966) ) G. Langford and M. Cohen: Trans. ASM 62 (1969) ) A. Ball, F. P. Bullen, F. Henderson and H. L. Wain: Philos. Mag. 21 (1970) ) J. Gill Sevillano, P. Van Houtte and E. Aernoudt: Prog. Mater. Sci. 25 (1981) ) Q. Liu, X. Huang, D. J. Lloyd and N. Hansen: Acta Mater. 50 (2002) ) T. Maki and T. Furuhara: Recrystallization Fundamental Aspects and Relations to Deformation Microstructure, ed. by N. Hansen, X. Huang, D. Juul Jensen, E. M. Lauridsen, T. Leffers, W. Pantleon, T. J. Sabin and J. A. Wert, (Riso National Laboratory, Roskilde, Denmark, 2000) pp ) J. E. Hockett and O. D. Sherby: J. Mech. Phys. Solids 23 (1975) ) Y. Sakai and H.-J. Schneider-Muntau: Acta Mater. 45 (1997) ) D. A. Hughes and W. D. Nix: Metall. Trans. A 19A (1988) ) A. Belyakov, K. Tsuzaki, H. Miura and T. Sakai: Acta Mater. 51 (2003) ) D. B. Williams and C. B. Carter: Transmission Electron Microscopy, (Plenum Press, New York, 1996) pp ) G. K. Williamson and W. H. Hall: Acta Metall. 1 (1953) ) J. K. Mackenzie: Biometrika 45 (1958) ) A. Belyakov, T. Sakai, H. Miura and R. Kaibyshev: Philos. Mag. Lett. 80 (2000) ) Q. Liu, D. Juul Jensen and N. Hansen: Acta Mater. 46 (1998) ) A. E. Romanov and V. I. Vladimirov: Dislocation in Solids, Vol. 6, ed. by F. R. N. Nabarro, (Elsevier Science, 1992) pp ) M. Murayama, J. M. Howe, H. Hidaka and S. Takaki: Science 295 (2002) ) A. C. C. Reis, L. Kestens and Y. Houbaert: Science of Metastable and Nanocrystalline Alloys, ed. by A. R. Dinesen, M. Eldrup, D. Juul Jensen, S. Linderoth, T. B. Pedersen, N. H. Pryds, A. Schroder Pedersen and J. A. Wert, (Riso National Laboratory, Roskilde, Denmark, 2001) pp ) C. M. Young and O. D. Sherby: J. Iron Steel Inst. 211 (1973) ) Y. Kimura, H. Hidaka and S. Takaki: Mater. Trans., JIM 40 (1999) ) B. L. Li, W. Q. Cao, Q. Liu and W. Liu: Mater. Sci. Eng. A 356 (2003)