CIRP Annals - Manufacturing Technology

CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Contents lists available at ScienceDirect CIRP Annals - Manufacturing Technology journal homepage: http://ees.elsevier.com/cirp/default.asp Severe plastic deformation (SPD) processes for metals A. Azushima (1) a, *, R. Kopp (1) b, A. Korhonen (1) c, D.Y. Yang (1) d, F. Micari (1) e, G.D. Lahoti (1) f, P. Groche (2) g, J. Yanagimoto (2) h, N. Tsuji i, A. Rosochowski j, A. Yanagida a a Department of Mechanical Engineering, Graduate School of Engineering, Yokohama National University, Yokohama, Japan b Institute of Metal Forming, RWTH Aachen University, Aachen, Germany c Department of Materials Science and Engineering, Helsinki University of Technology, Espoo, Finland d Department of Mechanical Engineering, KAIST, Deajeon, Republic of Korea e Department of Manufacturing and Management Engineering, University of Palermo, Palermo, Italy f Timken Research, The Timken Company, Canton, OH, USA g Institute for Production Engineering and Forming Machines, University of Technology, Darmstadt, Germany h Institute of industrial Science, The University of Tokyo, Tokyo, Japan i Department of Adaptive Machine Systems, Graduate School of Engineering, Osaka University, Osaka, Japan j Department of Design, Manufacture and Engineering Management, University of Strathclyde, Glasgow, United Kingdom ARTICLE INFO ABSTRACT Keywords: Forming Metal Strain Processes of severe plastic deformation (SPD) are defined as metal forming processes in which a very large plastic strain is imposed on a bulk process in order to make an ultra-fine grained metal. The objective of the SPD processes for creating ultra-fine grained metal is to produce lightweight parts by using high strength metal for the safety and reliability of micro-parts and for environmental harmony. In this keynote paper, the fabrication process of equal channel angular pressing (ECAP), accumulative rollbonding (ARB), high pressure torsion (HPT), and others are introduced, and the properties of metals processed by the SPD processes are shown. Moreover, the combined processes developed recently are also explained. Finally, the applications of the ultra-fine grained (UFG) metals are discussed. ß 2008 CIRP. 1. Introduction * Corresponding author. Processes with severe plastic deformation (SPD) may be defined as metal forming processes in which an ultra-large plastic strain is introduced into a bulk metal in order to create ultra-fine grained metals [1 7]. The main objective of a SPD process is to produce high strength and lightweight parts with environmental harmony. In the conventional metal forming processes such as rolling, forging and extrusion, the imposed plastic strain is generally less than about 2.0. When multi-pass rolling, drawing and extrusion are carried out up to a plastic strain of greater than 2.0, the thickness and the diameter become very thin and are not suitable to be used for structural parts. In order to impose an extremely large strain on the bulk metal without changing the shape, many SPD processes have been developed. Various SPD processes such as equal channel angular pressing (ECAP) [8 11], accumulative roll-bonding (ARB) [12 14], high pressure torsion (HPT) [15,16], repetitive corrugation and straightening (RCS) [17], cyclic extrusion compression (CEC) [18], torsion extrusion [19], severe torsion straining (STS) [20], cyclic closed-die forging (CCDF) [21], super short multi-pass rolling (SSMR) [22] have been developed. The major SPD processes are summarized in Table 1 with schematic configurations and the attainable plastic strain. ECAP, ARB and HPT processes are well-investigated for producing ultrafine grained metals. It is known that the metals produced by these processes have very small average grain sizes of less than 1 mm, with grain boundaries of mostly high angle mis-orientation. The ultra-fine grained metals created by the SPD processes exhibit high strength [23 25], and thus they may be used as ultrahigh strength metals with environmental harmony. The yield stress of polycrystalline metals is related to the grain diameter d by the following Hall Petch equation: s Y ¼ s 0 þ Ad 1=2 (1) where s 0 is the friction stress and A is a constant. Eq. (1) means that the yield stress increases with decreasing square root of the grain size. The decrease of grain size leads to a higher tensile strength without reducing the toughness, which differs from other strengthening methods such as heat treatment. The relationship between proof stress and grain size of pure iron is shown in Fig. 1 [6]. The proof stress changes inversely with the square root of the grain size, following the Hall Petch relationship. It is seen that the proof stress of the ultra-fine grained irons, with sub-micrometer grains, is five times greater than commercially pure iron. Thus, the conventional structural metals with ultra-fine grains are lighter due to their high strength. Since pure iron does not contain harmful elements, it is in harmony with a clean environment. Moreover, the improvements of the superplasticity, corrosion and fatigue properties of metals processed by SPD are expected. On the other hand, the ultra-fine grained metals are available only for micro-parts [26,27]. 0007-8506/$ see front matter ß 2008 CIRP. doi:10.1016/j.cirp.2008.09.005

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 717 Table 1 Summary of major SPD processes Process name Schematic representation Equivalent plastic strain Equal channel angular extrusion (ECAE) (Segal, 1977) e ¼ n p 2 ffiffi cotð Þ 3 High-pressure torsion (HPT) (Valiev et al., 1989) e ¼ gðrþ p ffiffi, gðrþ ¼n 2pr 3 t Accumulative roll-bonding (ARB) (Saito, Tsuji, Utsunomiya, Sakai, 1998) e ¼ n p 2 ffiffi 3 ln t 0 t In Fig. 2 [28], the mechanical properties of a wire specimen made by SPD is plotted against the ratio of wire diameter D to the grain size d, D/d. The proof stress decreases with decreasing D/d when D/d is less than 100. In particular, when D/d is less than 5, the proof stress decreases abruptly with decreasing D/d. From these observations, the ratio of D/d must be greater than 100 in order to guarantee the safety and the reliability of metals for micro-parts. This paper reviews the severe plastic deformation processes to create metals with ultra-fine grains. In the following, the fabrications of the SPD processes are shown in Section 2. Then, the Fig. 1. Relationship between proof stress and grain size of pure iron [6]. properties of metals processed by SPD processes are shown in Section 3, the combined processes developed recently are explained in Section 4, and the applications of the ultra-fine grained metals are discussed in Section 5. 2. SPD processes 2.1. Equal channel angular press (ECAP) process 2.1.1. Conventional ECAP processes Fig. 3 shows the schematic representation of side extrusion processes, which are a kind of double axis extrusion or side extrusion [29]. Fig. 3(d and e) indicates the process in which pure shear deformation can be repeatedly imposed on materials so that an intense plastic strain is produced with the materials without any change in the cross-sectional dimensions of the workpiece. These processes are named as ECAE (Equal channel angular extrusion) or ECAP. Segal [8,30] proposed this process in 1977 in order to create an ultra-fine grained material. Although ECAP is generally applied to solid metals, it may also be used for consolidation of metallic powder. Kudo and coworkers [31] employed repetitive side extrusion with back pressure to consolidate a pure aluminum powder. In the 1990s, developments of ultra-fine grained materials were carried out with this method by Valiev et al. [9,10,32], Horita and coworkers [33 45] and Azushima et al. [46 48] and others [49 51]. The schematic representation of the ECAE process is shown in Fig. 4. The specimen is side extruded through the shear deformation zone with the dead zone in the outer corner of the channel. When the workpiece is side extruded through the channel, the total strain is e ¼ 1 ffiffiffi 3 p 2cot f 2 þ c 2 þ c cosec f 2 þ c 2 (2) Fig. 2. Material behavior during forming processes of micro-parts of a wire specimen with diameter to grain size D/d [28]. Fig. 3. Schematic illustration of side extrusion process, which are a kind of double axis extrusion or side extrusion [29].

718 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Fig. 4. Schematic representation of ECAE process. where F is the angle of intersection of two channels and C is the angle subtended by the arc of curvature at the point of intersection. When F =908 and C =08, the total strain from the above equation is e = 1.15. After n passes, it becomes n e. Fig. 5 shows the fundamental process of metal flow during ECAP [6]. The channel is bent through an angle equal to 908 and the specimen is inserted within the channel and it can be pressed through the die using a punch. There are four basic processing routes in ECAP. In route A, the specimen is pressed without rotation, in route B A the specimen is rotated by 908 in an alternate direction between consecutive passes, in route B C the specimen is rotated 908 counterclockwise between each pass, and in route C the specimen is rotated by 1808 between passes. From these macroscopic distortions shown in Fig. 5, the influence of the processing route on the development of an ultra-fine grained microstructure can be considered [33,36]. Horita and coworkers [42] reported that the ultra-fine grained microstructure of pure aluminum after 10 passes in route A was the same as that after 4 passes in route B C. 2.1.2. Developed ECAP processes Azushima et al. [46 48] proposed the repetitive side extrusion process with back pressure. It is a process in which a high back pressure is applied in the process as shown in Fig. 6, in order to produce uniform shear deformation and prevent defects in the workpiece. The specimen is side-extruded between the punches A and B, while the punches C and D are fixed. In this process, the total strain becomes 1.15 after one pass. The punch A, controlled by the function generator, moves at a constant speed and the punch B generates a constant back pressure. Recently, ECAP die-sets have been developed to conduct the ECAP process with a back pressure which is controlled by computers [46,47]. Nishida et al. [52 57] developed a rotary-die ECAP shown in Fig. 7, which consists of a die containing two channels with the same cross-sections intersecting at the center with a right angle in order to remove the limitation in the conventional ECAP, i.e. the sample must be removed from the die and reinserted again in each step. At first, the sample is inserted into the die with the plunger as shown in Fig. 7(a), and after pressing the sample as Fig. 7(b), the die is then rotated by 908, and the sample is pressed again as Fig. 7(c). By using this ECAP apparatus, a sample can be pressed by the punch A with a back pressure from the punch B, similarly to that shown in Fig. 6. Repetitive pressings may be carried out with the rotary ECAP. This process is equivalent to route A in Fig. 5. In the same way as the repetitive ECAP process, a method to reduce the repetitive number by increasing the number of channel turns in the die [58 61] was developed. Using the two-turn channels the strain in one pass becomes double and the productivity of the ECAP process increases. A counterpunch for Fig. 5. Fundamental process of metal flow during ECAP. (a) The deformation of a cubic element on a single pass [33]. (b) Shearing characteristics for four different processing routes [36]. Fig. 6. Schematic representation of repetitive side extrusion process with the back pressure [46].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 719 Fig. 7. The ECAP process using a rotary-die: (a) initial state, (b) after one pass and (c) after 908 die rotation [52]. Fig. 10. The principle of the con-shearing process [64]. Fig. 8. Schematic representation of 2 turn ECAP [61]. providing additional pressure as shown in Fig. 8 may become a viable option available on common hydraulic presses. Presses with two opposite and equally powerful rams could be used for a cyclic process. In this process, the total strain becomes 2.3 after one pass. These ECAP processes have been used only in the laboratory because of their low productivity. For mass production, continuous processing techniques must be developed. First, in order to produce long metal bars and strips, equal channel angular drawing (ECAD) [62] and con-shearing were developed. The principle of ECAD is represented schematically in Fig. 9. In the ECAD process, the material in the form of a bar is drawn through the two channels. The rods are preformed by bending them 1358 to fit to the die, and are drawn through the ECAP die using as Instron tensile testing machine. The principle of the con-shearing process is represented schematically in Fig. 10 [63 65]. An equal-channel die with a channel angle is located at the exit of a satellite mill. Satellite rolls and a central roll are used as feed rolls. All the rolls are driven at an equal peripheral speed to generate large extrusion forces and the strip is extruded through the die continuously. This process uses friction between rolls to push the workpiece through an ECAP die. In this process, the shear deformation is given to the strip continuously, and the total strain after one pass is given by Eq. (2). Recently, equal channel angular rolling (ECAR) [66 68] and ECAP conform [69] were developed. The principle of the ECAR Fig. 11. The principle of the ECAR process for use in continuous production [68]. process is represented schematically in Fig. 11. The strip is fed between two rolls and extruded to reduce the thickness of the strip. Then, the strip flows into the outlet channel. The principle of the ECAP conform process is represented schematically in Fig. 12. The workpiece is driven forward by frictional forces on the three contact interfaces with the groove. The workpiece is constrained to the groove by the stationary constraint die, which restricts the workpiece and forces it to turn by shear deformation similarly to the ECAP process. Fig. 13 shows an Al workpiece at every stage of the ECAP conform process, from the initial feeding stock with a round crosssection to the rectangular rod after the first ECAP pass. For the ECAR process and the ECAE conform process, the total strain after one pass operation is given by Eq. (2), and the accumulated total strain is n e after n passes. The Incremental ECAP (I-ECAP) was developed by Rosochowski et al. [70 72]. Fig. 14 explains the principle of this process; it is based on incremental feeding of the billet by a distance b and using a reciprocating die C whose movement is synchronized with feeding. This enables feeding to take place during the withdrawal phase of die C. When the billet stops at a predetermined position, die C approaches it and deforms a small Fig. 9. Schematic of the equal channel angular drawing process (ECAD) [62]. Fig. 12. Schematic illustration of the ECAP Conform process [69].

720 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Table 2 Summarizes the geometrical changes of the specimen during the ARB process where roll-bonded by 50% reduction per cycle [81] Number of Cycles, n 1 2 3 5 10 Number of layers, m 2 4 8 32 1024 Total reduction, r (%) 50 75 87.5 96.9 99.9 Equivalent strain, e 0.80 1.60 2.40 4.00 8.00 Fig. 13. Al workpiece undergoing processing by ECAP Conform: the arrow marks the transition to a rectangular cross-section [69]. [81,82]. Stacking of sheets and conventional roll-bonding are repeated in the process. First, a strip is neatly placed on top of another strip. The interfaces of the two strips are surface-treated in advance in order to enhance the bonding strength. The two layers are joined together by rolling, as in the conventional roll-bonding process. Then, the length of the rolled material is sectioned into two halves. The sectioned strips are again surface-treated, stacked and roll-bonded. These procedures can be repeated limitlessly in principle, so that very large plastic strain can be applied to the material. The strain after n cycles of the ARB process can be expressed as, pffiffiffi 3 t e ¼ lnðrþ; r ¼ 1 ¼ 1 1 2 t 0 2 n (3) Fig. 14. Schematic representation of I-ECAP [70]. volume of the billet. The mode of deformation is that of simple shear and, provided the feeding stroke is not excessive, consecutive shear zones overlap resulting in a uniform strain distribution along the billet. Separation of the feeding and deformation stages reduces or eliminates friction during feeding; this enables processing of infinite billets, both bars and plates/ sheets. On the other hand, the friction-reduced ECAP processes were developed in order to produce long bulk bars with square crosssections [73,74]. The principle of this process is schematically represented in Fig. 15 [73]. By moving the tool, the friction forces over the three contacting interfaces become zero and the extrusion load decreases. The ECAP process may be used for the consolidation of metallic powder [31,75 79]. An aluminum powder and a steel powder at room temperature was pressed using the ECAP facility as shown in Fig. 4. where t 0 is the initial thickness of the stacked sheets, t the thickness after roll-bonding and r the reduction in thickness per cycle. Table 2 summarizes the geometrical changes of the specimen thick sheets are stacked and roll-bonded by a 50% reduction per cycle. The number of the initial sheets included in the specimen processed by n cycles of ARB becomes 2n. After 10 cycles of the ARB process, the number of layers becomes 1024 so that the mean thickness of the initial sheet is smaller than 1 mm. Optical micrographs of the ARB processed IF steel are shown in Fig. 17. In the material processed by two cycles ARB (Fig. 17(c)), the interface introduced in the second cycle is seen clearly. However, it is difficult to find the interface of the first pass at a quarter of the thickness. After five cycles, the whole thickness is covered by very thin and elongated grains. This process has been used by many researchers in order to create ultra-fine grained metals [84 88]. 2.3. High Pressure Torsion (HPT) Process The HPT process was first investigated by Bridgman [89]. In his experiments, attention was not paid to the microstructure change taking place in severely deformed metals. Another implementation of HPT was carried out by Erbel [90]. The specimen was a short ring with conical faces whose virtual extensions met at the axis of the apparatus as shown in Fig. 18. The conical matching faces of the 2.2. Accumulative roll-bonding (ARB) process The ARB process was first developed by Saito et al. [80 83]. The principle of the ARB process is represented systematically in Fig. 16 Fig. 15. The principle of friction-reduced ECAP processes [73]. Fig. 16. Diagrammatic representation of the accumulative roll-bonding (ARB) process [81].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 721 Fig. 17. Longitudinal cross-section of initial and ARB processed IF steel strips [82]. Fig. 20. Schematic illustration of the bulk-hpt process [94]. Fig. 18. Schematic diagram of ring tension device and dimensions of ring specimens [90]. punches have radial teeth to facilitate the application of torque. The ring specimens were constrained from all directions which created a condition closer to hydrostatic pressure. Recently, Valiev et al. conducted the HPT process using devices under high pressure as shown in Fig. 19 [15,16,89,91 104]. The design is a further development of the Bridgman anvil type device. In this device, a very thin disk is compressed in a closed die by a very high pressure. The torque is provided by the punch with contact friction at the interface between the punch and disk. The strain in torsion is given by gðrþ ¼ 2pnr (4) l where r is the distance from the axis of the disk sample, n the number of rotation and l the thickness of the sample. The equivalent strain according to the von Mises yield criterion is given by eðrþ ¼ gðrþ p ffiffiffi ; (5) 3 This method has the disadvantage that it utilizes specimens in the form of relatively small discs and is not available for the production of large bulk materials. Another disadvantage is that the microstructures produced are dependent on the applied pressure and the location within the disc. In order to solve the problem, Horita and coworkers developed an HPT process for use of the bulk sample as shown in Fig. 20 [94]. This process is designated as Bulk-HPT for comparisons with conventional Disk- HPT [95 104]. The severe plastic torsion straining (SPTS) process can be used for the consolidation powders using a similar apparatus as shown in Fig. 19 [105 107]. By using this process at room temperature, the disk type samples with a high density close to 100% were developed. The SPTS consolidation of powders is an effective technique for fabricating metal ceramic nano-composites with a high density, ultra-fine grain size and high strength. 2.4. Other processes The principle of the cyclic extrusion compression (CEC) process developed by Korbel et al. is represented schematically in Fig. 21 [18,108 111]. In the CEC process, a sample is contained within a chamber and then extruded repeatedly backwards and forwards. This process was invented to allow arbitrarily large strain deformation of a sample with preservation of the original sample shape after n passes. The accumulated equivalent strain is approximately given by e ¼ 4nln D d where D is the chamber diameter, d the channel diameter and n the number of deformation cycles. Since the billet in the CEC process is compressed from the both ends, a high hydrostatic pressure is imposed. The extrusion compression load becomes high so that the special pre-stressed tools are required, otherwise the tool life will be short. This process is better suited for processing soft material such as aluminum alloys. However, the strain introduced in the forward extrusion may be cancelled by the strain introduced on the backward extrusion. The principle of the cyclic closed-die forging (CCDF) process developed by Ghosh et al. is represented schematically in Fig. 22 [21,112,113]. A billet is first compressed in the vertical direction and then in the horizontal direction. The equivalent strain per operation is given by (6) e ¼ 2 lnðh=wþ p ffiffiffi (7) 3 Fig. 19. Schematic illustration of the thin disc-hpt process. where W is the width of specimen and H the height of specimen. The strain distribution is not uniform after strain accumulation. The principle of the repetitive corrugated and straightening (RCS) process developed by Huang et al. is represented schema-

722 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Fig. 24. Principle of linear flow splitting [114]. Fig. 21. Schematic of cyclic extrusion compression (CEC). Fig. 22. Schematic of cyclic closed-die forging (CCDF). Fig. 25. Principle of spin extrusion [115]. tically in Fig. 23 [17]. The technique consists of bending a straight billet with corrugated tools and then restoring the straight shape of the billet with flat tools. The equivalent strain per one operation is given by ½ðr þ tþ=ðr þ 0:5tÞŠ e ¼ 4ln p ffiffiffi (8) 3 where t is the thickness of sample and r is the curvature of bent zone. By repeating these processes in a cyclic manner, high strains can be introduced in the workpiece. Linear flow splitting developed by Groche et al. is another possibility to obtain ultra-fine grained metal [114]. The principle of this process is shown in Fig. 24. A sheet metal is compressed between the splitting roll and the supporting rolls. Under this state of stress two flanges are formed into the gap between the splitting and the supporting rolls. The material flow is mainly associated by a surface enlargement of the band edge. Several hundred percent of plastic strain occur. As a consequence, the outer surface areas of the flanges consist of ultra-fine grained metal. The properties of the metal in this state can be used for an increase of load bearing capability, e.g. bearings for rollers. The applicability of incremental bulk forming processes with high deformation for grain refinement in the sub-micrometer range was investigated by Neugebauer et al. [115]. A specific aspect of this approach is the opportunity to create a changed structure in the surface region, keeping the lower region or core unchanged. The incremental forming method of the spin extrusion as shown in Fig. 25 is used to create cup shaped or tube shaped parts from solid billets. The hollow shape is created by the concurrent partial pressure of three rolls on the surface of the workpiece and the pressure of the forming mandrel acting in the axial direction. The material flows axially and a cup wall is created between the forming tools [116]. The principle of the severe torsion straining (STS) process developed by Nakamura et al. is represented schematically in Fig. 26 [20]. The process consists of producing a locally heated zone and creating torsion strain in the zone by rotating one end with the other. The rod is moved along the longitudinal axis while creating the local straining. Therefore, a severe plastic strain is produced continuously throughout the rod. In order to create the torsion strain efficiently, the locally heated zone should be narrow and the rotation of the rod should be fast with respect to the moving speed of the rod. Moreover, a modification is made for the cooling system so that the heated zone is more localized to create torsion strain. The principle of the torsion extrusion process developed by Mizunuma et al. is represented schematically in Fig. 27. This process is characterized by rotation of a die or a container during an extrusion process for introducing a very large strain in to the metal. As high hydrostatic pressure involved in the extrusion raises the ductility of the metals, a very large torsion straining can be introduced to the workpiece. The mean value of representative Fig. 23. Principle of repetitive corrugating and straightening [17]. Fig. 26. Principle of the severe torsion straining (STS) process [20].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 723 Fig. 29. TEM Micrograph of ultra-low carbon steel after ECAPed 10 passes by route A [46]. Fig. 27. Principle of the torsion extrusion process [19]. strain on a cross-section of a column can be calculated as below. e ¼ 4pRN p 3 ffiffiffi (9) 3 H where R is radius of column, H is the height of the column, N is the number of rotation. Fig. 28 [19] shows a magnified view of the longitudinal section of the etched aluminum specimen after the torsion extrusion process, compared with that of the conventional extrusion. The torsion extruded part of the specimen is clearly observed to be more severely strained than that of the conventional extrusion. 3. Properties of metals processed by SPD The SPD-processed metals normally have ultra-fine grained structures that cannot be obtained through conventional thermomechanical processing. As a result, the SPD metals exhibit unique and excellent properties such as high strength, compared with the conventional materials having a coarse grain size of over several tens of micrometers. In the optical microstructure of metals over 5 passes of the ECAP processes, it is observed that the strong filamentary microstructure is developed with an increasing number of passes. In these conditions, observed microstructure must use a TEM analyzer. From the TEM microstructure, it is confirmed that many metals with an ultra-fine grain size (under 1 mm) are developed by ECAP processes. The ultra-fine grains of sub-micron size were created by ECAP processes in many of the metals and the grain size of the Al 4%Cu 0.5%Zr alloy became about 200 nm by ECAP with a plastic strain of 7 at 160 8C [2]. Aluminum and aluminum alloys with a sub-micron grain size were developed by ECAP processes [45]. For the ultralow carbon steel an ultra-fine grain size with a major axis length of 0.5 mm and a minor axis length of 0.2 mm was developed by 10 passes of repetitive side extrusion at room temperature as shown in Fig. 29 [46]. At the same time, they showed the relationship between the area fraction and the mis-orientation angle by the EBSP analysis [46]. They reported that most of the boundaries are high-angle grains, so that the processed steel is considered to be a kind of ultra-fine grain structured metal. In the ARB processes, it was noted that the evolution of microstructure and the increase in mis-orientation of boundaries were much faster than those when using conventional rolling [117,118]. A typical TEM micrograph of the ultra-fine structure in the interstitial free (IF) steel ARB processed by 7 cycles at 500 8Cis shown in Fig. 30. From the crystallographic analysis by Kikuchiline analysis, they reported that most of the boundaries were at a high angle. From these TEM microstructures, it is expected that the hardness and the tensile strength of metals with ultra-fine grains become higher. A number of studies have been conducted on the strength and ductility of various kinds of metallic materials processed by various SPD processes. The SPD-processed materials generally have very high strength compared with conventional metals. Fig. 31 illustrates a general tendency of the change in strength and ductility during SPD. The strength of the materials continuously increases with increasing the applied strain and then gradually saturates. On the other hand, the ductility drops greatly with a relatively small strain, and then keeps a nearly constant value or slightly decreases as the strain increases. Fig. 32 shows the relationship between the tensile strength, elongation and number of passes in ECAP for Armco steel [30]. The tensile strength increases with increasing pass number. The tensile strength is increased from 300 to 750 MPa after one pass. The tensile strength is increased by a factor of 2 after one pass in comparison with the specimen before the ECAP process, and increases with increasing pass number up to 8 passes. The tensile strength is higher than 800 MPa after 8 passes. On the other hand, the elongation decreases from 20% for the specimen before the ECAP process to several percents after 8 passes. Fig. 28. Magnified view of a longitudinal section of the etched aluminum specimen [19]. Fig. 30. TEM microstructure of the IF steel ARB processed by 7 cycles (e = 5.6) at 500 C [118].

724 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Fig. 31. Illustration showing the general tendency of the change in strength and ductility during SPD. Horita et al. [33 35] reported the same results for the aluminum alloys, and Azushima et al. [46 48,119 121] and Shin et al. [122 129] also reported for the steels. In particular, Aoki and Azushima [121] reported the relationship between nominal stress and nominal strain of specimens of ultra-low carbon steel, 0.15%C steel, 0.25%C steel and 0.50%C steel processed by ECAP of 1, 2, 3, 5, and 10 passes in route A at room temperature. They reported that the as-received material exhibits a stress strain curve that indicates normal strain hardening, while the specimens after ECAP do not exhibit strain hardening. The stress for each specimen increases rapidly with increasing strain and reaches a maximum at lower strain. Fig. 33 shows the relationship between the tensile strength and the pass number for the carbon steels. The tensile strength increases with increasing number of passes of ECAP. The tensile strength of ultra-low carbon steel after 10 passes was greater than 1000 MPa and was increased by a factor of 3 in comparison with the as-received material. The experimental data of the specimen Fig. 32. Relationship between tensile strength, elongation and pass number of ECAP for Armco steel [30]. Fig. 34. Relationship between total elongation and pass number for carbon steels based on ref. [121]. after 10 passes are plotted in the Hall Petch relationship of the yield stress against the root grain size as shown in Fig. 2. In this figure, the results for these specimens show good agreement with the standard Hall Petch relationship of iron obtained by Takaki and coworkers [130]. Fig. 34 shows the relationship between total elongation and the pass number for the carbon steels. For the low carbon steel, the elongation decreases to 20% after 3 passes, and for the other carbon steel, it decreases to 10% after 3 passes. Moreover, Shin and coworkers [128] also reported the stress strain curve of low carbon steel processed by ECAP at elevated temperatures as shown in Fig. 35. The tensile strength decreases with increasing processing temperature of ECAP and the total elongation increases. In the ARB process, Saito, Tsuji et al. [131 142] reported the mechanical properties of many metals processed by ARB. The relationship between the tensile strength, elongation and cycles of a commercially pure aluminum (JIS-1100) SPD processed by the ARB process is shown in Fig. 36 [132]. The tensile strength of the 1100-Al greatly increases to 185 MPa while the total elongation drops down to 13% by the 1 ARB cycle (equivalent strain of 0.8). As the number of the ARB cycles (strain) further increases, the flow stress continuously increases and reaches 340 MPa, which is four times higher than that of the starting material having a conventionally recrystallized microstructure. On the other hand, the elongation of the 1100-Al does not change as much after the second ARB cycle. As was illustrated in Fig. 29, this is the typical change in the mechanical properties during SPD, which seems to occur regardless of the kind of SPD process and material. The decrease in ductility is a general feature of strain-hardened metallic materials. Thus, it is not surprising that the SPD-processed materials, i.e., ultra-high strained materials show limited tensile ductility. It can be expected that the ductility can be recovered by subsequent heat treatment, as is the case with deformed and annealed materials. However, this has proven to be not so simple. Fig. 33. Relationship between tensile strength and pass number for the carbon steels based on ref. [121]. Fig. 35. Stress strain curves of the CS steel after ECA pressing at 350, 480, 540 and 600 8C [128].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 725 Fig. 36. Tensile strength and elongation of the 1100 commercially pure aluminum ARB processed by various cycles at RT [131]. Fig. 38. Yield strength and UTS vs. accumulated strain for AA-6061 SPD processed by ECAP, MAC/F and ARB at room temperature [149]. Fig. 37 shows the stress strain curves of the 1100-Al and ultralow-carbon interstitial free steel SPD processed by the ARB and then annealed at various temperatures for 1.8 ks [132]. In the figures, true stress and true strain are indicated by assuming uniform elongation. Also the mean grain size of the specimens measured from the microstructure observations are superimposed in the figures. The strength of the materials decreases with increasing grain size, i.e., with increasing annealing temperature. However, large elongation can be obtained only after the strength decreases. In particular, the curves clearly show that the flow stress reaches its maximum at an early stage of tensile test and is then necked down to fracture in the UFG specimens. The limited tensile ductility of the ultra-fine grained materials is understood in terms of early plastic instability. As is well-known, the plastic instability condition (i.e., necking condition in tensile test) for strain-rate insensitive materials, for example, is expressed as s ds (10) de where s and e are true stress and strain, respectively. Ultra-grain refinement greatly increases the strength, especially yield strength, of the materials. When the strain-hardening rate coincides with the flow stress, plastic instability, in other words necking, starts in the tensile test, which demonstrates a uniform elongation. The mechanical properties of the metals with ultra-fine grain processed by SPD have been investigated [143 149]. Cherukuri et al. reported a comparison of the properties of SPD-processed AA- 6061 by ECAP, CCDF and ARB as shown in Fig. 38 [149]. Commercially available AA-6061 in the annealed condition was subjected to severe plastic deformation processing by ECAP, CCDF and ARB at room temperature to approximately the same accumulated strain (4). From Fig. 38, it is understood that the SPD technique used did not show much effect on the flow behavior of AA-6061. Besides the mechanical properties, the fatigue property [120,150 154] and superplasticity property [133,155 167] were investigated by many researchers. 4. Combined process and properties 4.1. SPD process and conventional process In order to improve the strength of the ECAP processed metals, cold deformation can be combined with the ECAP process to introduce crystalline defects and refine the grains. Recently, two combined processes, the ECAP process and cold rolling, and the ECAP process and cold extrusion were developed. The principle of the combined process of ECAP and cold rolling is represented schematically in Fig. 39. Azushima et al. carried out experiments in which the specimens of ultra-low carbon steel were processed by ECAP in route A at room temperature and then the specimens processed by ECAP were rolled repetitively at room temperature in order to increase the strength. Fig. 40 shows the tensile strength after the combined process [168]. After 10 passes of ECAP, the tensile strength of ultra-low carbon steel is 1000 MPa and after cold rolling with a reduction in thickness of 95% it becomes 1300 MPa. Next, warm ECAP process was first used to refine the grain size of commercially pure Ti billets and the billets were further processed by repetitive cold rolling. The properties of the pure Ti processed by the two-step method are summarized in Table 3 [169]. ECAP increased the yield and tensile even strength to 640 and 710 MPa, respectively. After a cold reduction of 35%, the yield and tensile strengths increased to 940 and 1040 MPa which are higher than those for the Ti 6Al 4V alloy. Further cold rolling to a reduction of 55% resulted in even higher yield and tensile strengths. The principle of the combined process of ECAP and cold extrusion is represented schematically in Fig. 41. Stolyarov et al. [170] carried out experiments in which the billet of commercially pure Ti were first processed by ECAP in route B C at about 400 8C and then the billets processed by ECAP were further processed by cold extrusion to the accumulative reduction. The properties of the pure Fig. 37. True stress strain curves of (a) the 1100-Al ARB processed by 6 cycles at 200 8C and then annealed at various temperatures ranging from 100 to 400 8C for 1.8 ks and (b) IF steel ARB processed by 5 cycles at 500 8C and then annealed at various temperatures from 200 to 800 8C for 1.8 ks [132]. Fig. 39. Principle of the combined process of ECAP process and cold rolling.

726 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Fig. 40. Transition of tensile strength after combined process [168]. Fig. 42. Elongation to failure of the ECAP (4 passes) and ECAP (4 passes) + CR(70%) samples as a function of initial strain rate at 450 8C [172]. Fig. 41. Principle of the combined process of ECAP process and conventional extrusion. Ti processed by warm ECAP and cold extrusion are summarized in Table 4. After cold extrusion to 47% reduction in cross-section, the yield and tensile strengths were increased to 910 and 930 MPa respectively, which are higher than those of Ti 6Al 4V. Further cold extrusion in cross-section of 75% yielded even higher yield and tensile strengths. Next, in order to increase the strength of aluminum alloy (AA-6101) this combined process was conducted. The experimental results show that improved properties after cold extrusion are heavily dependent upon the prior ECAP processing routes. On the other hand, in order to improve the superplastic properties of metals processed by ECAP, cold deformation can be combined with ECAP to refine the grains. Park et al. [171,172] examined the superplastic properties at 450 8C of Al Mg alloy (A5154) processed by ECAP to 4 passes at 200 8C and cold rolling at a reduction in thickness of 70%. The comparison of the dependence of elongation on strain rat e between ECAP and ECAP + cold rolling (70%) samples is shown in Fig. 42. The elongations of the ECAP + cold rolling samples were higher than that of the ECAP sample at all strain rates. The maximum elongation was 812% at 5 10 3 s 1 and it was much higher than that of the eight passes ECAPed sample (595%). For the purpose of comparison, an appearance of the ECAP and ECAP + cold rolling (70%) samples elongated to failure is shown in Fig. 43. Similarly, the superplastic properties of 7075 aluminum alloy processed by ECAP of one pass and isothermal rolling at 250 8C were examined and a the alloy processed exhibited a maximum elongation of 820% at a temperature of 450 8C and an initial strain rate of 5.6 10 3 s 1. 4.2. SPD process and annealing The high strength of metal processed by SPD is obtained, but the ductility of the metals after SPD becomes very low. In order to improve the ductility of the metal processed by SPD, the metals were annealed after SPD process. Table 3 Properties of pure Ti processed by two-step [169] Processing state s 0.2 (MPa) s u (MPa) d (%) Coarse grain (10 mm) 380 460 27 ECAP(8) a 640 710 14 ECAP(8) + CR(35%) 940 1040 7 ECAP(8) + CR(55%) 1020 1050 6 ECAP(12) + CR(35%) 920 955 15 CR(35%) b 660 670 16 a ECAP route B C was used for all samples. b The value inside parentheses is cross-section reduction. Table 4 Properties of pure Ti processed by warm ECAP and cold extrusion [170] Processing state s 0.2 (MPa) s u (MPa) d (%) RA (%) Coarse grain 380 460 27 69 ECAP(8) a 640 710 14 61 ECAP(8) + Cold extrusion(47%) b 910 930 55 ECAP(8) + Cold extrusion(75%) 1020 1050 6 42 Ti 6Al 4V c 920 955 10 25 a ECAP route B C was used for all samples. b Reduction in cross-section area from cold extrusion. c From ASTM F 136-96. Fig. 43. Appearance of (a) ECAP sample and (b) ECAP + CR(70%) sample tested up to failure at 450 8C [172].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 727 Fig. 45. Tensile strength and elongation of ECA pressed low carbon steel annealed at 480 8C for various times [173]. Fig. 44. Relationship between tensile strength and hardness of specimens of ultralow carbon steel, 0.15%C, 0.25%C and 0.50%C steel after ECAP of 3 passes and then heat treatments of annealing [121]. Aoki and Azushima [121] carried out experiments in which the carbon steel samples were first processed by ECAP of 3 passes in route A at room temperature and then the samples processed by ECAP were further processed by annealing at a temperature of 600 8C and changing annealing times. Fig. 44 shows the relationship between nominal stress and nominal strain of specimens of ultra-low carbon steel, 0.15%C steel, 0.25%C steel and 0.50%C steel after the ECAP of 3 passes and after annealing. The tensile strengths become lower and the total elongations increase with decreasing tensile strength. The uniform elongations increase with decreasing tensile strength for all carbon steel samples. For example, a 0.5%C steel sample with a tensile strength 900 MPa and a total elongation of over 20% is obtained. Shin et al. [129,173,174] investigated static annealing after warm ECAP with a view to thermal stability. The low carbon steel: 0.15C 0.25Si 1.1Mn (in wt.%) (hereafter CS steel) was used. ECAP was carried out on the samples at 350 8C up to 4 passes by Route C, then samples for subsequent annealing were encapsulated in a glass tube with Ar atmosphere in order to minimize the possible decarburization. The static annealing treatment was conducted at 480 8C up to 72 h. Stress strain curves of the as-received, aspressed and annealed samples are shown in Fig. 45. The as-pressed and annealed samples exhibited no strain hardening behavior. It is of interest to note that stress strain curves of the samples annealed for 24 and 72 h were almost identical. This observation implies that the sample annealed for 24 h was mechanically stable although the microstructural examination revealed that recovery was in progress after 24 h annealing. In order to improve the ductility of metals processed by ARB, an annealing process was conducted. Tsuji et al. carried out experiments in which the aluminum and iron samples were first processed by ARB at a warm temperature and then the samples processed by ARB were further processed by annealing for 600 s or 1.8 ks from 200 to 800 8C. Fig. 37 shows the stress strain curve of commercially pure aluminum (1100-Al) and ultra-low carbon interstitial free steel specimens processed by various annealing conditions. The mean grain size is also indicated in this figure. The flow stress of both metals increases with decreasing mean grain size. Once the mean grain size becomes smaller than 1 mm, elongation of both Al and Fe suddenly reduced, though the strength still increased with decreasing grain size. On the other hand, as the grain size became larger than 1 mm, typical strain-hardening was observed and the elongation increased with increasing grain size. Stolyarov et al. [169] carried out experiments in which the commercially pure Ti billets were first processed by warm ECAP and repetitive cold rolling and further the billets processed by annealing at temperatures of 200 and 300 8C. The properties of the pure Ti billets processed are summarized in Table 5. Annealing pure Ti processed by SPD at temperatures below 300 8C generally improves the ductility without decreasing the strength. 4.3. SPD process and cooling Fig. 46 shows the strip thermo-mechanical control process (TMCP), or combined strip fabrication process, to manufacture fine grained plain carbon steel with a ferrite grain size of 3 mm [175]. TMCP and micro-alloying technology is widely used to manufacture precipitation-hardened high-strength steel sheets. The same process may be used in the rolling of strip, but precipitates are not easily controllable during rolling because strip rolling mills are arranged in tandem to gain higher productivity and constant quality rather than flexibility. To manufacture fine-grained steel strips, the rolling temperature must lie just above the transformation temperature or in the supercooled austenite state to accelerate transformation in the run-out table. This type of combined process can be used to manufacture plain carbon fine-grained steel sheets since the Fig. 46. Super short interval multi-pass rolling process.

728 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Table 5 Properties of pure Ti processed by warm ECAP and cold rolling with subsequent annealing [169] Processing state s 0.2 (MPa) s u (MPa) d (%) Coarse grain 380 460 27 ECAP(8) a 640 710 14 ECAP(8) + Cold rolling(35%) b + Annealing 200 8C, 0.5 h 985 990 8 ECAP(8) + Cold rolling(73%) + Annealing 300 8C, 1 h 942 1037 12.5 ECAP(12) + Cold rolling(35%) + Annealing 300 8C, 0.5 h 920 1000 14 Cold rolling(35%) 660 670 16 a ECAP route B C was used for all samples. Reduction in cross-section area by cold rolling. micro-alloying technology is difficult to be applied to this process. The temperature of the strip is controlled throughout the combined process to accumulate the dislocations in the grains before accelerated transformation. An example of such combined processes is shown in Fig. 46. This process can be used to produce ultra-fine grained C Si Mn steel with a grain size of 1 mm [175]. Astripwithawidthof 300 mm was successfully produced by this process. Fig. 47 shows an example of ultra-fine grained C Si Mn steel obtained by hot extrusion. Fig. 48 shows the yield strength of the steel sheet produced by the SSMR process as a function of ferrite grain size [176]. Some previous studies and Hall Petch equation are also shown in the figure as a comparison. The yield strength is increased from 350 to over 700 MPa with in decreasing grain size 4.5 1 mm, which is in good agreement with the Hall Petch relationship. It is also confirmed that the uniform elongation deceases. Another example of the combined process for producing the ultra-fine grained steel is warm rolling and cooling, which uses ferrite recrystallization during warm rolling [177 183]. Torizuka et al. [177,180 183] carried out multi-pass warm caliber rolling of two low carbon steel (SM490) specimens with a microstructure of ferrite and Pearlite. The specimen of the square bar with a side width of 80 mm was used. The warm caliber rolling schedule is summarized in Fig. 49. The caliber rolling at 500 8C was conducted in five stages to obtain specimens of different cumulative strains for different microstructure and mechanical properties. The cumulative reduction and the cumulative strain at each stage of rolling are also shown in Fig. 49. Fig. 50 shows the relationship between nominal stress and nominal strain of specimens subjected to different cumulative strains. The yield and tensile strengths of the caliber rolled specimen increase monotonically with increasing cumulative strain. There is a reduction in the elongation to failure of the caliber rolled specimens compared to the undeformed specimen, but there is almost no change among the specimens with different accumulative strains. Fig. 47. Ultra-fine grained steels obtained by SPD process. Fig. 49. Caliber rolling schedule with cumulative reduction and cumulative strain at each stage [180]. Fig. 48. Yield strength as a function of ferrite grain size [176]. Fig. 50. Nominal stress strain curves of undeformed specimen (e = 0) as well as caliber rolled specimens to various cumulative plastic strains (e = 0.7 3.8) [180].

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 729 Fig. 51. High strength thread articles out of SPD Ti alloy [184]. 5. Applications The properties of the metals processed by SPD exhibit high strength, ductility and fatigue characteristics. UFG metals are used as a structural material due to these properties. Bolts are also manufactured with titanium alloys processed by ECAP as shown in Fig. 51 [184] and are widely used in the automobile and aircraft industries. Micro bolts using the UFG carbon steel processed by cold ECAP have also been manufactured as shown in Fig. 52 [185]. Long carbon steel bars, of over several kilometers, with ultrafine grains are manufactured by the warm continuous caliber rolling and cooling process, from which the micro bolts are manufactured. Recently, in a Japanese National Project, sheets of low carbon steel of 2 mm thickness with ultra-fine grains were manufactured by the TMCP process. The deep drawing ratio of each sheet was 1.9 and the parts were used in sheet metal forming as shown in Fig. 53 [186]. It is well known [187,188] that superplastic forming is a highly efficient method of processing complex shape articles. An example of a possible practical application for nanostructured Al alloys is shown in Fig. 54 [2]. It presents a complex shape article of Piston type which was fabricated from the nanostructured Al1420 alloy by superplastic forming using the high strain rate superplasticity. In practice, despite a range of improved mechanical and physical properties of bulk UFG metals produced by SPD, the uptake of these materials by industry has been very slow so far. There are several reasons for this; one is the lack of industrial awareness of UFG metals. This is despite a large number of academics being engaged in research on SPD and UFG metals. Another reason is the scarcity of appropriately sized UFG samples for industrial trials; those produced by laboratories are usually too small because they are intended for metallurgical observations or basic mechanical testing. Finally, it is still not clear which of the Fig. 52. Overview and cross-section of micro bolts manufactured UFG Carbon steel processed by cold ECAP [185]. Fig. 54. View of article of Piston type fabricated from nanostructured Al1420 [2]. numerous laboratory-based SPD methods will emerge as the most appropriate for industrial implementation. As a result, potential producers of UFG metals hesitate to commit themselves to any particular method. They are also concerned about the commercial viability of UFG metals, which depends on the demand from potential markets and the cost of production. Both of them are difficult to assess because of the low availability of UFG metals and uncertainty regarding the SPD technology. There is also a lack of knowledge regarding post-spd processing or shaping of UFG metals. Nevertheless, there are some applications which, with a high degree of probability, will be leading the introduction of UFG metals into commercial markets. Initially, those applications are likely to be in the niche markets producing low volume specialty products (e.g. sputtering targets). The next step will be the medium volume markets with the emphasis put on product s performance rather than price (medical implants, defense applications, aerospace components, sports equipment). Eventually, the mass production of components may be undertaken by the automotive and construction industries. With the exception of sputtering targets, the examples presented below refer to potential applications rather than the current ones. Despite the focus of this paper on SPD-produced UFG metals, applications using UFG consolidated powders and nanostructured electrodeposited metals will also be considered as these are indicative of what can be achieved with all types of UFG metals. The first commercial application of bulk UFG metals was in sputtering targets for physical vapour deposition (Fig. 55). Honeywell Electronic Materials, a division of Honeywell International Inc., offers UFG Al and Cu sputtering targets up to 300 mm in diameter which are produced from plates by ECAP [189,190]. They are used for metallization of silicone wafers in the production of semiconductor devices. The main advantages of UFG sputtering targets, compared to their coarse grained (CG) counterparts, are: (1) the life span increased by 30% due to stronger material which allows the use of monolithic targets and (2) a more uniform deposited coating which results from reduced arcing. Another company offering UFG Cu targets is Praxair Electronics, which claims better sputter performance and 75% reduction in the ownership cost of such targets. Fig. 53. Examples of ultra-fine-grained C-Mn steel sheet forming [186]. Fig. 55. Worn out UFG 300 mm sputtering target [189].

730 A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 Fig. 56. Plate implants made of nanostructured titanium [192]. The next highly anticipated application is in the area of medical implants. These include hip, knee and dental implants as well as various screws, plates (Fig. 56) and meshes used in orthopaedic applications. Popular materials usually used in these applications are cobalt-chrome alloys, stainless steel and titanium alloys. Titanium alloys are used for implants because of their strength, low modulus of elasticity (better matching that of bones), corrosion resistance and good biocompatibility. Commercial pure (CP) titanium has better compatibility than titanium alloys but it is not used for load bearing implants because it is not strong enough. However, when nanostructured by SPD and subjected to further thermo-mechanical treatment, CP titanium can be strengthened to achieve the yield stress of 1100 MPa, which is comparable with the yield strength of titanium alloys [191]. Traditional titanium implants do not perform well with respect to wear resistance and fatigue life. Therefore, improvements in these properties, reported for UFG titanium, will be appreciated. Some Russian [192] and USA laboratories report that the UFG CP titanium implants are being already tried. The defense industry could benefit from two large scale applications of UFG metals, which are armor plates and armor penetrators. Lighter armor for military vehicles (Fig. 57) is crucial for the reduction of fuel consumption, higher speed, better maneuverability, longer operation range and air-transport of vehicles to remote locations. At the same time the ballistic performance must not be reduced. This can be achieved by the nanostructuring of aluminium or titanium alloys traditionally used for light armored vehicles. A good example is a UFG Al 5083 plate, which was obtained by cryogenic ball milling, consolidation by HIP, forging or extrusion and finally rolling [193]. With the yield strength of 600 700 MPa and elongation of 11%, the material exhibited a 33% improvement in the ballistic performance or a similar mass reduction compared to the standard plate. Improvements in ballistic performance are also reported for the electrodeposited nanocrystalline nickel-iron alloys produced by Integran Technologies. Armor structures are usually fabricated by welding of plates. However, traditional welding based on melting is destructive to the UFG material. An alternative technique is a solid state process of friction stir welding, which has the ability to refine grain structure. This results in the weld hardness being only marginally reduced compared to the initial hardness of a UFG material [194]. Health issues surrounding the use of depleted uranium for armor penetrators resulted in a search for alternative materials with similar performance characteristics. One of those characteristics is an inherent ability of depleted uranium to self-sharpen on impact which is due to the generation of adiabatic shear bands. Tungsten, sometimes considered as a replacement for depleted uranium because of its high density, does not have this ability; thus penetrators made of tungsten undergo mushrooming on impact, which results in less penetration. UFG metals are known to have reduced strain hardening capacity, which promotes localized plastic deformation; at high deformation rates this leads to adiabatic shear banding. This was confirmed by producing UFG tungsten (by ECAP with a die angle of 1208 at 1100 1000 8C and subsequent rolling at 600 700 8C), which exhibited adiabatic shear banding when subjected to a dynamic load [195]. The aerospace industry values even small weight reductions which might be achieved by the introduction of new materials or technologies. However, this industry is very cautious because of the safety concerns, and slow in implementing any changes. Introducing a new material may take 10 20 years, which results from the requirements of the well established technology and a fully developed supply chain. UFG metals are not chemically different from their CG precursors, so there should be no fundamental obstacles to their use. On the other hand the new properties of UFG metals have to be well documented with respect to aerospace applications and the SPD and post-spd processes have to be commercially available. All these requirements mean that we will have to wait a few more years for the first aerospace applications. These, most likely, will be associated with light UFG metals used for structural components of the fuselage and wings. Regarding the engines, some external elements (e.g. shields) and less thermally demanding internal elements (e.g. titanium blades for the low pressure compressor section) can also be considered. There has only been limited information published so far on the potential use of UFG metals by the aerospace industry; Boeing, filed a few patents on friction stir welding used as a means of nanostructuring metals for fasteners and other parts [196] while EADS is interested in UFG aluminum sheets. Users of sports equipment will also benefit from UFG metals, particularly where high strength and low weight is required. UFG metals could find applications in high performance bicycles, sailing equipment, mountaineering equipment, golf, tennis, hockey, etc. One example is NanoDynamics high performance (NDMX) golf balls, which have a hollow nanostructured titanium core (Fig. 58). The core material is manufactured using the UFG chip machining Fig. 57. AAV7A1 Amphibious Assault Vehicle (image courtesy of BAE Systems). Fig. 58. NDMX golf ball (image courtesy of NanoDynamics).

A. Azushima et al. / CIRP Annals - Manufacturing Technology 57 (2008) 716 735 731 Fig. 59. Metallix racquet (image courtesy of HEAD). technology licensed from Purdue University. Another example of using UFG metals in sporting goods is the commercial activity of PowerMetal Technologies, a company with an exclusive license to use Integran s electrodeposition technology in consumer products. They cooperate with HEAD in the production of their new Metallix (Fig. 59) and Airflow racquets, which use a composite of carbon fibres and nanocrystalline metal. UFG metals can be beneficial to some products through improvements in their manufacturing processes. The most promising one is superplastic forming which is currently confined to a low volume production because of a very low process speed, necessary when forming classical superplastic metals. UFG metals possess better superplastic properties, which allow a tenfold increase of the forming speed, and some temperature reduction [197]. Superplastic UFG metals exhibit higher ductility which makes them suitable for forming more complex components. Despite large volume of research on SPF of UFG metals, practical applications are still a matter for the future. One possible application has been presented by the Institute for Metals Superplasticity Problems, Ufa, Russia. They made models of hollow blades by diffusion bonding (DB) and SPF using UFG Ti 6Al 4V sheets (Fig. 60). By using sheets with the grain size down to 0.2 mm they were able to decrease the temperature of the process from 900 to 800 8C for DB and to 700 8C for SPF [198]. The temperature reductions observed will improve technical feasibility and the economics of the process. Among many interesting properties of UFG metals is their ability to flow easier and at lower temperatures when forged into complex shapes. It is claimed that energy savings up to 30% could be achieved due to: lower forging temperature, shorter heat-up time, smaller forging stock size, fewer number of hits and lower forging load [199]. A very small grain size can be a virtue of its own. This is the case with metal micro-parts having geometrical sizes comparable with coarse grains of classical materials. Using UFG metals in microforming allows micro-billets to behave as polycrystalline billets. This refers to both the inner body and the surface of the billet. The latter is illustrated in Fig. 61 as a Fig. 61. SEM pictures of a micro-bulged sheet made of CG and UFG Al 1070 [200]. substantial reduction of the orange peel effect [200]. Another advantage of using UFG metals is better surface finish resulting from micro-milling [201], micro-edm [202] and diamond turning [203]. The above applications of UFG metals are only a fraction of the possible uses. Since the SPD technology can convert all CG metals into UFG metals, it is only a matter of time when new, sometimes unexpected, applications will be discovered. For this to happen, information dissemination among industrial engineers, transfer of reliable SPD technologies to industry and commercialization effort is required [204]. 6. Conclusion Processes of severe plastic deformation, defined as metal forming processes in which an ultra-large plastic strain was imposed on a bulk material in order to make ultra-fine grained metals, were reviewed in this keynote paper. As processes used for this purpose, various methods such as, ARB, HPT, RCS, CEC, STS, CCDF, etc. were developed, and combined SPD processes with conventional processes were also proposed. The properties of the metals processed by SPD are also reviewed. The SPD-processed metals have very high strength, and in order to increase the strength further, conventional cold forming processes are combined with SPD processes. Since the ductility of metals is reduced by relatively low strain, the heat treatment of annealing is conducted after the SPD process in order to improve the ductility. The properties of the metals processed by the SPD processes exhibit high strength and ductility that lead to good fatigue characteristics. The UFG metals could be used as structural materials due to these properties, but the area of application is limited at the moment because the available size of billet is small. Since SPD technology can convert all metals into UFG metals, it is expected that new methods of producing larger billets will enlarge the area of applications. Acknowledgment The authors wish to thank Prof. K. Osakada and Dr. J. Allwood for checking the manuscript of keynote paper. References Fig. 60. Models of hollow blades made of UFG Ti 6Al 4V sheet (image courtesy of Institute for Metals Superplasticity) [198]. [1] Valiev RZ, Islamgaliev RK, Alexandrov IV (2000) Bulk Nanostructured Materials from Severe Plastic Deformation. Progress in Materials Science 45(2):103 189.