Materials Science and Engineering A 528 (2011) 59275934

Contents lists available at ScienceDirect

Materials Science and Engineering A

journal homepage: www.elsevier.com/locate/msea

Texture and mechanical properties of EUROFER 97 steel processed by ECAP

M. Eddahbi a, , M.A. Monge a , T. Leguey a , P. Fernndez b , R. Pareja a
a b

Departamento de Fsica, Universidad Carlos III, 28911 Legans, Madrid, Spain CIEMAT. Avda. Complutense 22. 28040 Madrid, Spain

a r t i c l e

i n f o

a b s t r a c t
The EUROFER 97 steel was processed by equal channel angular pressing (ECAP) at 550 C for four passes via route C. The starting material consisted of ferritemartensite dual phase composed by small subgrains of about 0.5 m and low angle boundaries less than 5 . The volume fraction of second phase particles was around 10 vol.%, besides a texture formed by several bers orientations belonging to the zone axes 1 1 0 , 1 1 1 and 1 1 2 . Increasing ECAP deformation, this microstructure became into equiaxed grain structures of less than 1 m, and the misorientation between contiguous grains increased. This renement of the microstructure was accompanied by the development of a new texture described by a family of ber orientations related by rotations around axes 1 1 0 and 1 1 1 . Tensile tests have revealed that an ECAP treatment at 550 C for two passes could signicantly strengthen the tempered material still maintaining good ductility. 2011 Elsevier B.V. All rights reserved.

Article history: Received 30 November 2010 Received in revised form 4 April 2011 Accepted 5 April 2011 Available online 12 April 2011 Keywords: RAFM steels EUROFER 97 ECAP processing Texture Microtexture Mechanical properties

1. Introduction Due to its mechanical properties, long-term stability and predictability up to 650 C, the reduced activation ferriticmartensitic EUROFER 97 steel is being considered for structural applications in the breeding blankets of ITER, and for the demonstration fusion reactor DEMO [1,2]. The strength of this steel can be enhanced by precipitation hardening, and strengthening induced by dislocations, particle dispersion or grain boundaries. However, precipitation hardening and strengthening by dislocations or particles dispersion can reduce dramatically ductility and toughness compared with grain renement [3]. Ultrane grained structures can be developed in metals by severe plastic deformation techniques (SPD), among which ECAP is a successful method to develop submicron microstructures in bulk steel [49]. The aim of the present work has been to explore the capability of warm ECAP for developing a stable grain structure and texture in tempered EUROFER 97 that can enhance its mechanical behavior in its operational temperature range, i.e. below 600 C. 2. Experimental procedure The material used for this study was produced by Bhler (Austria) with composition (wt.%): 0.11%C, 8.7%Cr, 1%W, 0.1%Ta, 0.19%V, 0.44%Mn, 0.004%S, with the balance Fe (EUROFER 97). The

Corresponding author. Tel.: +34 91 624 8734; fax: +34 91 624 8749. E-mail address: meddahbi@s.uc3m.es (M. Eddahbi). 0921-5093/$ see front matter 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2011.04.006

as-received plates were normalized at 980 C for 27 min followed by tempering at 760 C for 90 min. 12 mm 12 mm 65 mm billets of the as-tempered material were ECAP processed at 550 C with a velocity of 10 mm min1 through a die with an intersection angle of 105 . The billets were subjected to one, two and four ECAP passes. They were rotated 180 around its longitudinal axis before inserting in the die for the subsequent ECAP pass, what is referred to as route C. Samples for microstructural studies and tensile tests were machined from the billets, polished with alumina and etched by solution of 5 g FeCl3 + 6 ml HCl + 100 ml H2 O. The microstructure was characterized by optical microscopy (OM), Scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Room temperature tensile tests were performed on at tensile specimens at a constant crosshead rate of 0.04 mm min1 (initial strain rate of about 3.3 105 s1 ). The specimens with a gauge length of 20 mm, a width of 3 mm, and a thickness of 1 mm were cut parallel to the ow plane of the billets as shown in Fig. 1. Texture measurements were accomplished by measurement of pole gures using the Schulz reection method, in a Siemens diffractometer equipped with a D5000 goniometer. The measurements were performed through the range of azimuthal angles between 0 and 75 in the step mode with increments of = = 5 . A textureless standard sample of annealed pure iron was used for defocusing correction. Quantitative three-dimensional orientation distribution functions (ODFs) were obtained using a Siemens software [10]. These were represented by iso-intensity lines in equidistant sections, 1 = 10 , through the Euler space dened by 1 , and 2 angles. For the representation

5928

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

Die TD 105 PD ND

(b)

Die

(a)

tral region of the specimen surfaces parallel to the ow plane. These surfaces were carefully polished. The EBSD data were analyzed using the in-house software package CHANNEL 5. The OIM images applying the usual coloring scheme for the Euler angles were obtained using the external software package MTEX MatLab Toolbars for quantitative texture analysis [11]. Grain boundaries with misorientation between the contiguous grains of mis 10 have been dened as low angle grain boundaries (LABs), or subgrain boundaries, and the ones with mis > 10 referred to as high angle grain boundaries (HABs).

Fig. 1. Sketches of (a) the ECAP die used, and (b) samples machined from the ECAP processed billets for tensile tests.

3. Results and discussion 3.1. Microstructure

of the ODFs, a monoclinic symmetry was used resulting in a good concordance between the measured ODFs and the calculated ones for all ECAP deformed samples. Grain orientation maps of the material in the as-tempered condition and after ECAP deformation were obtained from the corresponding electron backscattering diffraction (EBSD) patterns recorded using a HKL Nordlys EBSD system interfaced with a eld emission gun scanning electron microscope FEG-SEM JEOL JSM6500F equipped with an energy dispersive spectrometer (EDS). The applied accelerating voltage and the working distance were 25 kV and 1520 mm, respectively. The EBSD scans using step sizes of 0.140.18 m were performed on 54 m 43 m areas in cen-

Fig. 2 shows the effect of the ECAP processing on the EUROFER 97. The microstructure of the material in the as-tempered condition exhibits laths substructure as Fig. 2(a) and (b) reveal. Both the OM observations and the SEM analyses revealed that this microstructure contains small carbide particles as Fig. 2(b) shows. Most of these particles were found at the prior-austenite boundaries and along the lath sub-boundaries. Two types of carbides were identied: Cr-rich M23 C6 with sizes of less than 0.2 m, and ne Ta- and V-rich MX carbides. The volume fraction of second phase particles was estimated of about 10% using SME and EBSD measurements.

Fig. 2. OM and SEM micrographs of EUROFER 97. (a) and (b) microstructures in the as-tempered condition showing martensite laths, carbide particles and prior-austenite boundaries decorated with carbide particles. Microstructure after (c) and (d) one pass ECAP deformation, and (e) and (f) four ECAP passes. The pressing direction (PD) is horizontal.

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

5929

Fig. 3. ECAP deformation effect on the grain structure of the EUROFER 97. (a) and (d) OIM images for as-tempered material and ECAP deformed for four passes, respectively. (b) and (e) grain size distribution, and (c) and (f) misorientation distribution for the material in the as-tempered condition and ECAP deformed for four passes, respectively.

After the rst ECAP pass the microstructure was reoriented and shear strained as shown in Fig. 2(c) and (d) [12]. After four passes, Fig. 2(e) and (f), the microstructure turned out near equiaxed. In addition, the SEM images show that ECAP deformation did not induce agglomeration of second phase particles. 3.2. EBSD observations Fig. 3 shows the effect of four ECAP passes on the grain structure. The indexation of the EBSD patterns also showed that the samples exhibit a dual phase ferritemartensite microstructure. The OIM images and the distributions of grain sizes and misorientation reveal that ECAP deformation under the present conditions

induces grain renement and misorientation increment in the ferrite phase. Moreover, ECAP deformation transformed martensite into ferrite. The martensite fraction of 27 vol.% for the EUROFER 97 in the as-tempered condition was lowered to 5 vol.% after four ECAP passes. The as-tempered state consisted of ne subgrains conned in coarse grains of up to 7 m in size as shown in Fig. 3(a) and (b). Most of the subgrains are of about 0.5 m in size. The misorientation histograms corresponding to both ferrite and martensite phase in as-tempered state revealed the presence of high density of LABs of less than 5 and low density of HABs as shown in Fig. 3(c). After four ECAP passes the subgrain size is less than 1 m and the misorientation angle in the ferrite phase increases, Fig. 3(f).

5930

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

Fig. 4. TEM micrographs showing the effect of the ECAP deformation on the microstructure of as-tempered EUROFER 97. (a) Elongated substructure and (b) details showing the nucleation of subgrains at a grain triple junction for the material ECAP deformed for one pass; the arrows indicate sub-boundaries or dislocation walls. Microstructure in a sample processed for two passes showing: (c) lamellar and (d) equiaxed grain structure. (e) and (f) images showing equiaxed grains with high density of dislocations in a sample ECAP deformed for four passes.

3.3. TEM observations Also, the TEM images shown in Fig. 4 illustrate the ECAP effect on the microstructure of the EUROFER 97. After the rst and second pass, the microstructure turned out highly elongated and shear strained as shown in Fig. 4(a) and (c). The microstructure after a single pass exhibited a low dislocation density and small subgrains, or grain domains, as Fig. 4(a) and (b) reveal. These domains appear to form by the development of a structure of dislocations walls (indicated by arrows) that can transform into sub-boundaries and boundaries during further deformation via dislocation motion and accumulation into the sub-boundaries, i.e. by extended recovery, also called continuous recrystallization [13]. The sub-boundaries are pinned by small carbide particles retaining the microstructure on the succeeding passes. EBSD analysis

and TEM observations indicate that the mechanism of continuous recrystallization appears to be the main responsible for the grain renement induced by ECAP in EUROFER 97. Nevertheless, the formation of new subgrains is also observed to occur at precipitate particles associated to sub-boundaries or grain triple junctions as shown in Fig. 4(b). In particular, the size of the subgrains nucleated at the triple junctions appear to be less than 100 nm, i.e. much smaller than the subgrains produced by the mechanism of extended recovery. Due to the considerable volume fraction of precipitate particles present in this steel, the mechanism of subgrain nucleation at particles associated to sub-boundaries, or triple junctions, could signicantly contribute to the development of a homogeneous ultrane-grained structure. It should be emphasized that the mechanism of subgrain formation apparently induced by particles associated to sub-boundaries should be different from the particle

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

5931

Fig. 5. The ODFs for EUROFER 97, represented through 1 -sections ( 1 = 10 ) and the spatial arrangement of the bers in the Euler space, showing the effect of ECAP deformation. a) and b) as-tempered material; c) and d) after a single pass; e) and f) after two passes and g) and h) after four passes.

stimulated nucleation (PSN) mechanism based on the deformation zones around the second phase particle [14,15]. This mechanism is characteristic of some alloys containing hard particles and requires a critical particle size of 12 m, below which a subgrain cannot develop and become into a recrystallization nucleus [14]. Instead, it is suggested that ECAP deformation in EUROFER 97 contributes to

activate new slip systems and creates a homogenous distribution of strain in the surrounding region of a small carbide particle that favors the dislocation rearrangement into wall and sub-boundaries, i.e. giving rise to subgrains. After two ECAP passes the samples exhibited a microstructure composed of elongated grains containing equiaxed substructure

5932

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

Fig. 6. Orientation density functions, f(g), for ECAP deformed EUROFER 97 as a function of 1 : a) one pass, b) two passes and c) four passes.

as shown in Fig. 4(c) and (d). The corresponding OIM images obtained from a sample area of 2322 m2 reveal enhancement of the microstructure fragmentation and a shift of the misorientation distribution toward high mis values although the LABs are still the major fraction of boundaries present in the ECAP deformed samples. On the contrary, the equiaxed substructure had a high dislocation density, and boundaries that exhibited a thick and blunt image. After four passes, the elongated structure is no longer visible and the microstructure is roughly equiaxed as shown in Fig. 4(e) and (f). This result is in contrast to that obtained for an IF steel, where the lamellar structure is still retained after four passes via route C [16]. The morphology of the substructure developed by ECAP will depend on the initial grain orientation respects to the shear plane imposed in the ECAP die. This substructure would control the rearrangement of dislocations into dislocation walls contributing to increase the angular misorientation across the cell walls or sub-boundaries. The as-tempered material consists of laths substructure randomly distributed in different orientations into the volume material (Fig. 2(a) and Fig. 3(a)), which would imply different ECAP induced substructures depending on their initial orientation respect the shear plane. ECAP deformation activates dislocation gliding on the slip planes giving rise to an elongated lamellar substructure on these planes. After an ECAP pass, this elongated substructure tends to align almost parallel to the shear plane, thus should conserve the initial lamellar substructure on ECAP deformation, although they become more elongated and its spacing might be reduced. On the contrary, if the lamellar substructure, or the active slip planes, forms a high angle with shear plane, a new substructure would be developed, so that ECAP deformation renes the substructure of these grains. After several ECAP passes, the lamellar structure would be fragmented resulting in a more equiaxed and rened substructure with higher misorientation mis across the subgrain boundaries as Fig. 3(f) shows. The effectiveness of the ECAP treatment for rening the grain structure will depend on the change of strain path induced by the ECAP routes. 3.4. Texture Fig. 5 shows the effect of ECAP deformation via route C on the texture of EUROFER 97 material. The ODFs are represented in the Euler space through 1 -sections from 1 = 0 180 , for the material in the as-tempered condition (Fig. 5(a)), and ECAP processed for

one pass (Fig. 5(c)), two passes (Fig. 5(e)) and four passes via route C (Fig. 5(g)). The corresponding spatial arrangements of the ODFs are represented in Fig. 5(b), (d), (f) and (h) showing the positions of the bers in the Euler space. For more details, Fig. 6 plots the orientation intensity, f(g), as a function of 1 for the ECAP deformed materials. In the present work, the Miller index {h k l} u v w represents an orientation that has an {h k l} parallel to the ow plane of the billet, i.e. plan normal to ND in Fig. 1, and an u v w direction parallel to the longitudinal axis of the billet, i.e. PD in Fig. 1. As observed in Fig. 5(a) and (b), the as-tempered material exhibits numerous weak bers. This texture results from a combination of severe rolling and cross rolling underwent by the material before being tempered. Each ber is an ensemble of orientations related to each other by rotations around the 1 1 0 , 1 1 1 and 1 1 2 axes. It is worth noticing that these axes are either parallel to the ND or located at an azimuth angle. After the rst ECAP pass some weak bers disappear and new ones form, compare Fig. 5(b) and (d). However, the stronger ber components observed in the material ECAP deformed for a single pass, marked as S1 , S2 , S3 , S4 , S5 and S6 , remain after four ECAP passes, although their intensity varies along the skeleton line as shown in Fig. 6(a)(c). The maximum intensities for these bers correspond to {1 1 0} 1 2 2 , {1 1 0} 1 1 1 and {1 1 1} 1 2 3 orientations. The weak bers, labeled W1 , W2 and W3 can be described by (1 0 6)[0 1 0], (0 1 3)[4 3 1] and (1 0 3)[3 8 1] orientations. The present ECAP induced texture in the steel EUROFER 97 contrast with the ber texture 1 1 1 and 1 0 0 and the bers {1 1 0} u v w and {h k l} 1 1 1 reported for low carbon and interstitial-free (IF) steels ECAP deformed at room temperature [9,16]. The discrepancies could be attributed to the differences in the ECAP parameters as well as to the high content of carbide particles in EUROFER 97. The crystallographic texture developed during the rst ECAP pass is retained after the succeeding passes as shown in Fig. 5(c) and (g), albeit the distribution of orientations along the bers changes. After the second ECAP pass the intensity of the main bers decreases and the weak bers vanish or even disappear as occurred for the W1 , W2 and W3 -bers, compare Fig. 6(a) and (b). Then, the second pass does not reverse the rst pass texture to the original one of the billet as the approach assuming simple shear deformation at the intersection plane of the die channels predicts for ECAP processing of bcc materials [17,18]. After the fourth ECAP pass, the intensities of two main bers increase respect to the second pass, and the weak bers, which are removed or reduced by the

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

5933

second pass, reappear. These results indicate that shear deformation induced by the rst pass, or by an odd-numbered pass, is not reversed by the subsequent pass to the strain state of the billet at the beginning or after a previous even-numbered route C pass. In fact, it has been reported for IF steel that the real ECAP processing conditions alter the cyclic changes of the texture predicted by the simple shear model [16,19]. The inhomogeneous strain across the ECAP processed billet, the complex microstructure, the die geometry and the interaction with neighboring grains appear to constraint the capability of a route C pass to reverse shear deformation induced by the previous ECAP pass. However, the cyclic tendency of the texture for going back to the initial one can be still maintained for an IF steel, in particular if the die angle is changed from 90 to 120 [17,18]. In the present work, the cyclic reversion of the texture by ECAP via route C neither appears to be partially accomplished even though a die with = 105 is used. The results obtained for tempered EUROFER 97 ECAP processed under the present conditions reveal that shear deformation induced by a pass is practically irreversible against a subsequent route C pass. This irreversibility that is much stronger than the reported for IF steel in Refs. [17,18] may be attributed to the complex substructure and high concentration of carbide particles present in the tempered EUROFER 97 steel. Furthermore, the present texture analyses give rise to issues related to deformation mechanism responsible for the evolution of the microstructure during severe plastic deformation of EUROFER 97. Actually, it was found that the orientations developed during a single ECAP pass are connected by rotations in the range 10 40 around the same axes found for the material in the as-tempered condition, i.e. around 1 1 0 and 1 1 1 axes. Nevertheless, these axes are now located between the PD and TD directions. Similar results are found in the material deformed for two and four ECAP passes. To illustrate these rotations qualitatively, Fig. 7 shows the localization of ND of the orientations in the cubic stereographic projection. The differences in the ND distribution with increasing ECAP deformation, and keeping the rotation axes 1 1 0 and 1 1 1 , are clearly evident. For one ECAP pass, the S5 -ber orientations are con nected through 35 rotation around [1 1 1] and those corresponding to the S6 -ber through rotation of 20 and 10 around [0 1 1] and [1 0 1]. However, both bers are closely joined together into rotations around 1 1 0 axes after two passes. Furthermore, the S1 -ber described by a rotation of 30 around [1 0 1] transforms into the s11 and s12 -bers as shown in Fig. 7(a) and (b). The s11 -ber main tains the rotation axis [1 0 1] whereas the rotation axis in s12 -ber is parallel to the ED. These results indicate that route C can produce signicant changes on the distribution of orientations along the main bers during ECAP deformation of EUROFER 97. 3.5. Tensile properties Tensile test true stresstrue strain curves are depicted in Fig. 8. The tensile strength of the samples ECAP deformed for one pass and two passes via route C increased noticeably in comparison with that for the as-tempered material. The yield stress y , ultimate tensile strength (UTS) and the uniform elongation (u ) for the as-tempered material are 549 MPa, 685 MPa and 9.4%, respectively. After two passes u decreased to about 7.5% whilst y and UTS increased to 638 MPa and 800 MPa, respectively. However, ECAP deformation for four passes resulted in a decrease in the mechanical properties compared to two passes ( y = 602 MPa, UTS = 746 MPa and u = 4%). It has been reported that route C beyond two passes is ineffective for reducing the grain size in an IF steel ECAP processed at room temperature [17]. The strength decrease in EUROFER 97 deformed for four passes under the present ECAP conditions cannot be attributed to a limited capability for the grain renement beyond the second pass via route C because grain renement was still achieved during the fourth pass as the TEM observa-

Fig. 7. Traces of the ND to the ber orientations in the stereographic projection for EUROFER 97 ECAP deformed via route C for: a) a single pass, b) two passes and c) four passes. Each ber is marked by its corresponding symbol.

tions showed. Even assuming stabilization in the grain renement beyond two ECAP passes, this cannot give account for the decrease in strength and ductility by itself. Nevertheless, the heterogeneous substructures developed in elongated and equiaxed subgrains could inuence the mechanical behavior of the material. In fact, the microstructure after two passes is a combination of lamellar and equiaxed substructures, which evolved into an equiaxed substructure after four passes. Both substructures contribute to strain hardening at room temperature for a given strain rate, although hardening should be higher in the regions with equiaxed substruc-

5934

M. Eddahbi et al. / Materials Science and Engineering A 528 (2011) 59275934

1000

800

600

400
RT - 3x10 s
-5 -1

200

As-tempered One pass ECAP Two passes ECAP Four passes ECAP

0 0

0.05

0.1

0.15

True strain
Fig. 8. Effect of ECAP deformation via route C on the tensile properties of astempered EUROFER 97.

microstructure (volume fraction of martensite was 27 vol.%). The microstructure contains ne subgrains of about 0.5 m in size and LABs less than 5 and second phase particles around 10 vol.%. The texture was composed by several weak bers with their orientations into the zone axes 1 1 0 , 1 1 1 and 1 1 2 . 2) The rst ECAP pass produced a highly elongated and shear strained structure with fragmented grains. After two ECAP passes a grain microstructure composed of lamellar and equiaxed substructure was observed. With further passes the fragmentation of the microstructure progressed giving rise to a practically equiaxed substructure after four passes. 3) The evolution of the microstructure upon ECAP deformation was accompanied by the development of a new texture described by a family of bers retaining the zone axes 1 1 0 and 1 1 1 after the successive passes. The martensite phase changes to ferrite and the fraction of LABs decreases with ECAP deformation. The martensite fraction lowered to 5 vol.% after four ECAP passes. 4) The yield strength and tensile strength increased signicantly after deformation for two ECAP passes but decreased after four passes. The enhanced mechanical behavior of the ECAP deformed steel EUROFER 97 could be attributed to the formation, renement and texture evolution of the lamellar substructure induced by ECAP deformation. Acknowledgements This work has been supported by Madrid Community through the project TECHNOFUSION (S2009/ENE-1679) and Spanish Ministry of Science and Innovation (Contract ENE2008-06403-C06-04). The authors thank the microscopy laboratory of CENIM-CSIC for EBSD measurements. References
[1] R. Lindau, A. Mslang, M. Rieth, M. Klimiankou, E.-M. Morris, A. Alamo, A.A.F. Tavassoli, C. Cayron, A.M. Lancha, P. Fernndez, N. Baluc, R. Schublin, E. Diegele, G. Fulacchioni, J.W. Rensman, B. van der Schaaf, E. Lucon, W. Dietz, Fus. Eng. Des. 7579 (2005) 989. [2] T. Nishitani, H. Tanigawa, S. Jitsukawa, T. Nozawa, K. Hayashi, T. Yamanishi, K. Tsuchiya, A. Mslang, N. Baluc, A. Pizzuto, E.R. Hodgson, R. Laesser, M. Gasparotto, A. Kohyama, R. Kasada, T. Shikama, H. Takatsu, M. Araki, J. Nucl. Mater. 386388 (2009) 405. [3] T. Gladman, F.B. Pickering, The effect of grain size on the mechanical properties of ferrous materials, in: T.N. Baker (Ed.), Yield, Flow and Fracture of Polycrystals, App. Sci. Publ., London, 1983, p. 148. [4] D.H. Shin, I. Kim, J. Kim, K.T. Park, Acta Mater. 49 (2001) 1285. [5] Y. Fukuda, K. Oh-ishi, Z. Horita, T.G. Langdon, Acta Mater. 50 (2002) 1359. [6] K.T. Park, S.Y. Han, B.D. Ahn, D.H. Shin, Y.K. Lee, K.K. Um, Scripta Mater. 51 (2004) 909. [7] J.T. Wang, C. Xu, Z.Z. Du, G.Z. Qu, T.G. Langdon, Mater. Sci. Eng. A 410411 (2005) 312. [8] Y. Iwahashi, M. Furukawa, Z. Horita, M. Nemoto, T.G. Langdon, Metall. Mater. Trans. 29A (1998) 2245. [9] M. Eddahbi, E.F. Rauch, Mater. Sci. Eng. A502 (2009) 13. [10] H.J. Bunge, Texture Analysis in Materials Science, Butterworths, 1982. [11] R. Heilscher, H. Schaeben, J. Appl. Cryst. 41 (2008) 10241037. [12] P. Fernndez, M. Eddahbi, M.A. Auger, T. Leguey, M.A. Monge, R. Pareja, J. Nucl. Mater., doi:10.1016/j.physletb.2003.10.071. [13] R.D. Doherty, D.A. Hughes, F.J. Humphreys, J.J. Jonas, D. Juul Jensen, M.E. Kassner, W.E. King, T.R. McNelley, H.J. McQueen, A.D. Rollett, Mater. Sci. Eng. A238 (1997) 219. [14] F.J. Humphreys, M. Hatherly, Recrystallization and Related Annealing Phenomena, 1st Ed., Pergamon, 2004 (chapter 8). [15] F.J. Humphreys, Acta Metall. 27 (1979) 1801. [16] S. Li, A. Gazder, I.J. Beyerlein, E.V. Pereloma, C.H.J. Davies, Acta Mater. 54 (2006) 1087. [17] S. Li, I.J. Beyerlein, Mater. Sci. Eng. A13 (2005) 509. [18] J. de Messemaeker, B. Verlinden, J. Van Humbeeck, Acta Mater. 53 (2005) 4245. [19] S. Li, A. Gazder, I.J. Beyerlein, C.H.J. Davies, E.V. Pereloma, Acta Mater. 55 (2007) 1017.

tures. Then, the expected ow stress increase in these zones should be relieved by assuming propagation of deformation into the lamellar substructure preventing the localization of high stresses in the equiaxed substructures. This stress release mechanism could be responsible for the improvement of tensile properties in the material ECAP deformed for one or two passes. So, it appears that the mechanical behavior of the ECAP deformed steel EUROFER 97 could be explained in terms of a combination of lamellar and equiaxed substructures where the morphology, renement, fraction and orientation of the lamellar substructure would determine the strength and ductility. Further assessment is needed for a comprehensible understanding of the strengthening mechanisms competing in EUROFER 97 taking into account dislocation density, volume fraction of martensite phase and carbide particles. Also, in the material ECAP deformed for two passes it should be noted that the 1 1 0 directions are found at 30 40 from the ED of the billet in the S5 and S6 -bers and at 30 in the s11 -ber. This means that many more subgrains in this case are favorably oriented to be plastically deformed in comparison with the materials deformed for one or four ECAP passes, which accounts for the better tensile properties in the steel EUROFER 97 deformed for two ECAP passes. According to the results obtained in EUROFER 97 ECAP processed via route C, it is expected that a selected combination of routes and annealing treatments should produce a nanostructured microstructure with enhanced mechanical properties. 4. Conclusions The EUROFER 97 steel was processed by ECAP at 550 C for up to four passes via route C, and the effects on the grain structure, crystallographic texture and tensile properties were investigated. The main results are as follows: 1) The initial microstructure of the material in the as-tempered condition consisted of a ferrite-martensite dual phase

True stress, MPa