Res Chem Intermed (2010) 36:629638 DOI 10.

1007/s11164-010-0198-2

Microstructures and mechanical properties of ultrane grained pure Ti produced by severe plastic deformation
Chang-Young Hyun Jong-Hyun Lee Ho-Kyung Kim

Received: 15 December 2009 / Accepted: 1 February 2010 / Published online: 18 September 2010 Springer Science+Business Media B.V. 2010

Abstract Microstructure evolution and mechanical behavior of ultrane grained (UFG) commercially pure Ti produced by equal channel angular pressing (ECAP) were investigated. Repetitive pressings of the same sample were performed to six passes at 683 K, using the procedure designated as route Bc. After the sixth pass was nished, recrystallized grains were observed as similar as the fourth pass. The average size of the recrystallized grains was approximately 0.3 lm. The hardness value (Hv) continuously increases with decreasing grain size. The Hv values are in good agreement with the other experimental data of Ti produced by severe plastic deformation processes. The similar slop kH suggests that these microstructures have similar density of dislocations in the grains produced by the severe plastic deformation processes such as torsion straining, multiple forging, and ECAP. The grain p size dependence of ky in the present samples is 7.9 MPa m. After six-pass ECAP, the ultimate tensile strength was increased by 60%. This is most likely due to considerable grain renement through severe deformation by ECAP. The standard HallPetch relation for yield strength and hardness in the ECAPed Ti implies that the ECAPed Ti samples have similar texture and that the effect of grain size on strength may prevail over the effect of texture on the strength in Ti. Keywords Ultrane grained Pure titanium Microstructures Grain size Severe plastic deformation

C.-Y. Hyun J.-H. Lee Department of Materials Science and Engineering, Seoul National University of Technology, 172 Kongnung-dong, Nowon-ku, Seoul 139-743, Korea H.-K. Kim (&) Department of Automotive Engineering, Seoul National University of Technology, 172 Kongnung-dong, Nowon-ku, Seoul 139-743, Korea e-mail: [email protected]

123

630

C.-Y. Hyun et al.

Introduction The equal channel angular pressing (ECAP) process is recognized to be quite effective in improving the strength of many metallic alloys through (sub) grain renement [1, 2]. The ECAP process has been applied mainly to metals and alloys having cubic crystal structure, and the number of severe plastic deformation studies on hexagonal materials is limited. It is well known that deformation behaviors of hexagonal metals are signicantly different from those of cubic metals due to the limited number of active slip systems. Pure titanium is important in biomedical applications due to its low weight, excellent corrosion resistance, and high biocompatibility [3]. However, the application of ordinary coarse-grained commercial pure Ti is limited by its relatively low ultimate tensile and fatigue strengths. Utilizing ECAP for pure Ti is great interest for its enhancement in tensile strength through grain renement. However, the mechanical properties of the pure Ti produced by ECAP have not been systematically studied. The purpose of this study is to develop a better understanding of the microstructure evolution during ECAP and tensile behavior of ultrane grained (UFG) pure Ti produced by ECAP. Emphasis is placed on investigating TEM microstructures and the HallPetch relationship for the ECAPed pure Ti.

Experimental A commercially pure titanium (grade 2) was cut to the rods with a diameter of 14.5 mm and a length of 90 mm and then annealed at 1,073 K for 1 h in Ar atmosphere, and then quenched into room-temperature water. Its chemical composition was Ti-0.06Fe-0.01N-0.01O-0.01H (wt%). The average grain size was 105 lm after the heat treatment. ECAP was conducted using a die with an internal angle U of 110 and an outer curvature angle W of 25. The present ECAP die was designed to give an approximate strain e of *0.76 on each pressing according to the following equation [2]; e 2 cotU W W cos ecU W 2 2 2 2 p 3 1

Repetitive pressings of the same sample were performed to six passes. During ECAP, all pressings were conducted at 683 K, using the procedure designated as route Bc, in which each sample was rotated 90 around its longitudinal axis between the passages. The details of the ECAP processing have been reported elsewhere [4]. Micro-hardness and tensile tests were carried out to evaluate the strength and ductility of the ECAP processed materials. Vickers micro-hardness (Hv) was measured on the plain perpendicular to the longitudinal axes, by imposing a load of 100 g for 15 s. Tensile properties in the transverse longitudinal directions of the ECAPed and unECAPed Ti rods were measured using the miniature tensile specimens with geometry of 5-mm gauge length, 2-mm width, 1-mm thickness, and 1.5-mm shoulder radius. The tensile samples were extracted from the center portion

123

Microstructures and mechanical properties of ultrane grained pure Ti

631

of the ECAPed and unECAPed materials using electro-discharge machining. A displacement rate of 1 mm/min was used for tensile testing, corresponding to an initial strain rate of 3.3 9 10-3/s. The microstructure of the ECAPed samples was investigated using a transmission electron microscopy (TEM). The TEM specimens were thinned mechanically to a thickness of around 50 lm. Disks of 3-mm diameter were punched out of the thinned sheet, and followed by a jet polishing technique using a solution of 25% HF?75% HNO3. TEM observations were conducted using a Philip EM420 operated at 120 kV.

Results and discussion Prior to the ECAP process, the annealed pure Ti (unECAPed) sample contains a low dislocation density. Figure 1a shows the bright eld TEM image of unECAPed pure Ti sample. In the TEM images, the low number of straight or crooked dislocations was observed in the large grains of about 150-um size. Even though typical twin images were not observed, twin-like straight striations of 5*15-nm interval were observed inside of the grains. The twin-like striations were distinguished with typical twin structures because the boundaries of the striation patterns were not shown clearly, specically in the region exposed to the electron beam. A possible mechanism of the twin-like striation formation is considered as follows: During polishing procedure for the preparation of TEM specimens, the specic surface oxidation layer of several nanometers in thickness might be formed on Ti. The oxidation layer as the straight striations was produced by the interaction with the Ti matrix. However, the extra spots, i.e., the so-called satellite spots in the selected area diffraction (SAD) pattern image of the striations (Fig. 1b), suggested a low possibility for the oxidation layer because the satellite spots related with the severalnanometers-thick oxidation layer might not be observed in the SAD pattern. Therefore, the oxidation effect was not produced in the current microstructure. It was inferred that twin-like striation structure was made by the short-range displacement induced within one lattice spacing. In other words, the boundaries of the patterns indicated the formation of crookedness by elastic deformation, i.e., tweed. Tweed is a modulated structure formed in the elastic deformation regime prior to twin structure. With regard to Ti, tweed is a metastable deformation structure of hexagonal close packed (HCP) lattice by strain or supersaturation. In the tweed microstructure, satellite diffraction spots mean that lattice dispersive wave (LDW) related with deformed lattice or charge density wave (CDW) representing the special periodicity of electron density is built in the matrix, which is corresponding to the formation mechanism of metastable pre-martensite. TEM micrographs of pure Ti observed after the rst ECAP pass are shown in Fig. 2a, b, c. Figure 2 indicates the dislocation density increased by ECAP deformation. The dislocations were mainly distributed at the interface between the columnar or polygonal grains. The straight band structures observed in Fig. 2a were analyzed as the slip bands. Firstly, severe plastic deformation induced during ECAP

123

632

C.-Y. Hyun et al.

Fig. 1 TEM micrographs of pure Ti observed in unECAPed sample: a bright eld and b SAD pattern image

produced the slip bands. After the rst ECAP pass, the microstructures shown in Fig. 2b, c were also observed partially. Depending on the region deformed by ECAP process, relatively larger (Fig. 2b) or smaller polygonal grains (Fig. 2c) were observed. The smaller polygonal grains were formed in the more severely strained regions. Consequently, it was estimated that the columnar slip bands were broken and transformed to the polygonal gains as the induced strain was increased, and the polygonal grains became to the small one and simultaneously rotated during the transformation. As shown in Fig. 2c, new small regions representing the complex pattern were also observed between the polygonal grains. The region was judged as a cluster of low-angle grain boundaries. After the second pass of ECAP, the observed slip band width was thinner than that of the rst pass, and the band interfaces were distinguished more clearly (Fig. 2d). Meanwhile, the straight-shaped bands varied partially to the crooked shape in the severely strained region. Moreover, the discriminable images with average size of about 0.5 um were also observed in the intermittent places of the broken slip bands, indicating the most severely strained region (Fig. 2e). These images represented the clusters of low-angle grain boundaries as mentioned above. Finally, the broken polygonal gains and clusters of low-angle grain boundaries were transformed to the recrystallized grains through the repeated grain rotation and merging (Fig. 2f). The size of the recrystallized grains was approximately 1 um and denitely coarse in comparison with the slip band width in Fig. 2d or the former grains size observed in Fig. 2e. The recrystallized gains were uniformly distributed throughout the specimen after the fourth ECAP pass was completed (Fig. 2g). Phase transformation into the recrystallized grains was considered to be completed throughout the specimen. The average size of the recrystallized grains was approximately 0.4 um. After the sixth pass was nished, the recrystallized grains were observed as similar as the fourth pass (Fig. 2h). However, the size distribution of the grains was more uniform than that of the fourth pass. The accumulated deformation energy by the additional ECAP process might change the sub-micron grains into the ner ones. However,

123

Microstructures and mechanical properties of ultrane grained pure Ti

633

Fig. 2 TEM micrographs of pure Ti observed after ECAP

signicant grain size reduction was not observed and the average size of the recrystallized grains was approximately 0.3 um. According to the result by Valiev et al. [5] on eight passed pure Ti with a total strain of 800%, the grain size was 0.3 lm, suggesting that the grain renement effect with repeated pressing seems to diminish after about 400500% total strain. Therefore, the grain size reduction following the recrystallization was estimated to be minor by the additional ECAP

123

634

C.-Y. Hyun et al.

passes. In summary, the average size of recrystallized grains decreased to 1, 0.4, and 0.3 um with respect to the ECAP pass number of 2, 4, and 6, respectively. Figure 3 shows the relation of Hv against d-1/2 plotted in order to examine the HallPetch relationship for the ECAPed Ti. HV H0 kH d1=2 2

where H0 and kH are the material constants. The Hv value continuously increases with decreasing grain size. Present experimental data point lie on a single straight p line depicted by Eq. (2) with H0 = 156.5 Hv and kH 1:04 HV mm. The present results are in good agreement with the experimental data of Ti produced by high pressure torsion by Sergueeva et al. [6]. However, experimental data of multiple forged Ti by Salishchev et al. [7] have lower H0 (=123.4 Hv) with similar p kH 1:01 HV mm. The similar slop kH suggests that these microstructures have similar density of dislocations in the grains produced by the severe plastic deformation processes such as torsion straining, multiple forging, and ECAP. The p values of H0 and kH can be converted to 1,533.7 MPa and 10:2 MPa mm, p respectively. The kH slope of present pure Ti 10:2 MPa mm was close to that p of UFG Mg alloy 12:7 MPa mm [8] and much higher than that of UFG pure p Al 3:6 MPa mm [9]. In general, severe plastic deformation processes lead to a larger kH slope because the ultrane grains contain a high density of dislocations. This result suggests that the dislocation density of the present ECAPed Ti sample is

Fig. 3 Room temperature hardness against grain size for CP Ti from Refs. [6, 7] and current investigation

123

Microstructures and mechanical properties of ultrane grained pure Ti

635

relatively high. The difference of H0 (156.5 Hv *123.4 Hv) might be due to a difference in the impurity content in Ti. The engineering stress-engineering strain curves of the unECAPed and ECAPed Ti samples from one pass to six passes are shown in Fig. 4. Their data for yield stress (YS), ultimate tensile strength (UTS), hardness, elongation to failure, and grain size are summarized in Table 1. Similar to the tensile behavior of UFG materials reported previously [1], the ECAPed sample showed no strain hardening behavior. However, the unECAPed sample exhibited the large strain hardening behavior with large elongation. The relationship between tensile strength (YS, UTS) and number of pressings in the ECAPed Ti is shown in Fig. 5. The UTS and YS increase with an increase in the number of pressings up to six in a similar pattern. The ultimate tensile strength (UTS) and yield strength (YS) of the six-pass ECAPed sample were found to be 669 and 635 MPa, respectively. Especially, UTS and YS increase dramatically due to grain renement (from 105 to 0.7 lm) after one pass. In contrast, the UTS and YS of the unECAPed sample were 418 and 248 MPa, respectively. After six-pass ECAP, the ultimate tensile strength was increased by 60%. This is most likely due to considerable grain renement through severe deformation by ECAP. This fact suggests that the UFG structure of pure Ti is very advantageous for improving titaniums strength without alloying, so that it would be suitable for use in medical devices. The tensile elongation was, however, largely decreased by 31%. This is related to the decrease of strain hardening capability after ECAP, which commonly occurs in other alloys after ECAP [1].

Fig. 4 Tensile stressstrain curves for the unECAPed and ECAPed Ti samples

123

636 Table 1 Mechanical properties and grain size of pure Ti

C.-Y. Hyun et al.

Material (unit) unECAPed 1 pass 2 pass 3 pass 4 pass 5 pass 6 pass ECAPed [5] ECAPed [14]

UTS (MPa) YS (MPa) ef (%) Hv 418 524.1 536.1 558.3 583.3 593.7 668.8 810 1,050 248.0 466.6 481.2 523.7 538.7 553.3 635.1 650 970 380 315 248 190 47 27.8 40.0 31.3 33.8 37.5 32.5 15 8 26

d (mm)

160.9 105 193.4 0.7 203.6 0.5 205.2 0.4 208.3 220.3 0.3 0.3 0.15 15 9 32 100

unECAPed [15] 460 unECAPed [16] 440 unECAPed [16] 380 unECAPed [16] 377

Fig. 5 UTS and YS against grain size of the unECAPed and ECAPed Ti samples

The HallPetch relationship correlates the grain size of a material to its yield stress. According the to HallPetch relationship, the yield stress ry of a material can be expressed as: ry r0 ky d1=2 3

where r0 is the lattice friction stress, ky is the stress intensity for plastic yielding across polycrystalline grain boundaries, and d is the grain size in mm. The values of

123

Microstructures and mechanical properties of ultrane grained pure Ti

637

p r0 and ky were determined to be 206.5 MPa and 7:9 MPa mm, respectively, from p Fig. 6. The grain size dependence of kyin the present samples 7:9 MPa mm is p much lower than that for pure Ti 21:2 MPa mm [10] and almost close to that p for a hot extruded AZ31 alloy 9:6 MPa mm [11]. In the absence of appreciable work hardening, the hardness (Hv) of the material is proportional to the yield stress through the expression Hv & 3ry [12]. The slope kyin the present samples p p 7:9 MPa mm can be converted to about 2:4 HV mm, which is twice higher than p that of kH 1:04 HV mm obtained from the hardness tests for the present samples. It is not clear why there is inconsistency between YS and Hv. this difference right now. According to Table 1, YS and hardness results are consistent each other. The yield stresses of the ECAPed Ti are higher than the unECAPed material whereas the hardness of the former is higher than that of the latter. This result is similar to from the report on the ECAPed Al alloys where a good correlation between YS and hardness was observed [13]. Inconsistency between YS and Hv has been also reported in the ECAPed Mg alloys, where hardness continued to increase while YS did to decrease with grain renement [8]. This behavior has been understood that the strengthening effect by grain renement is more pronounced than the softening effect by texture anisotropy in hardness test where deformation occurs locally and non-uniformly, whereas the softening effect by texture anisotropy is more dominant in tensile test where deformation occurs more macroscopically and uniformly. It is well known that the dislocation density in the grain interiors and other microstructural factors, such as solute elements and particle distribution, can also inuence the slope of the Hall Petch equation (ky and kH) in commercial alloys. However, pure Ti has no inuence

Fig. 6 Hall-Petch behavior of pure Ti

123

638

C.-Y. Hyun et al.

due to these factors. Thus, only dislocation density works as obstacles to the dislocation movement. The standard HallPetch relation for YS and hardness in the ECAPed Ti implies that the ECAPed Ti samples have similar texture. And, the effect of grain size on strength may prevail over the effect of texture on the strength in Ti.

Conclusions Microstructure evolution and the mechanical behavior of UFG commercially pure Ti (grade 2) produced by ECAP were investigated. After the sixth pass was nished, the recrystallized grains were observed as similar as the fourth pass. However, the size distribution of the grains after the sixth pass was more uniform than that of the fourth pass. From the relation of Hv against d-1/2, the Hv values are in good agreement with the other experimental data of Ti produced by severe plastic deformation processes. The grain size dependence of kyin the HallPetch relation p for the present samples was 7:9 MPa mm. Aftersix-pass ECAP, the ultimate tensile strength was increased by 60%. This is most likely due to considerable grain renement through severe deformation by ECAP.

References
1. R.Z. Valiev, E.V. Kozlov, Y.F. Ivanov, J. Lian, A.A. Nazarov, B. Baudelet, Acta Metall. Mat. 42, 24672475 (1994) 2. A. Yamashita, M. Furukawa, Z. Horita, T.G. Langdon, Metall. Mat. Trans. 29A, 22452252 (1998) 3. Y. Zhu, T. Lowe, T.G. Langdon, Scripta Mater. 51, 825830 (2004) 4. W.J. Kim, C.Y. Hyun, H.K. Kim, Scripta Mater. 54, 17451750 (2006) 5. A. Vinogradov, V. Stolyarov, S. Hashimoto, R.Z. Valiev, Mater. Sci. Eng. A318, 163173 (2001) 6. A.V. Sergueeva, V.V. Stolyarov, R.Z. Valiev, A.K. Mukherjee, Scripta Mater. 45, 747752 (2001) 7. G.A. Salishchev, R.M. Galeyev, S.P. Malysheva, M.M. Myshlyaev, NanoStruct. Mater. 11, 407414 (1999) 8. H.K. Kim, W.J. Kim, Mater. Sci. Eng. A385, 300308 (2004) 9. H. Hasegawa, S. Komura, A. Utsunomiya, Z. Horita, M. Furukawa, M. Nemoto, T.G. Langdon, Mater. Sci. Eng. A265, 188196 (1999) 10. A.A. Salem, S.R. Kalidindi, R.D. Doherty, Scripta Mater. 46, 419423 (2002) 11. Y.N. Wang, C.I. Chang, C.J. Lee, H.K. Lin, J.C. Huang, Scripta Mater. 55, 637640 (2006) 12. M.F. Ashby, D.R.H. Jones, Engineering Materials 1 (Pergamon Press, Oxford, 1980), p. 105 13. D.R. Fang, Q.Q. Duan, N.Q. Zhao, J.J. Li, S.D. Wu, Z.F. Zhang, Mater. Sci. Eng. A459, 137144 (2007) 14. V.V. Stolyarov, I.V. Alexandrov, Y.R. Kolobov, M. Zhu, Y. Zhu, T. Lowe, in Proc. of 7th Int. Fatigue, Congress, Beijing, P.R. China (Eds: X.R. Wu, Z.G. Wang), Higher Education Press, Beijing, China, 1345, (1999) 15. A. Vinogradov, S. Hashimoto, Adv. Eng. Mater. 5, 351358 (2003) 16. N.G. Turner, W.T. Roberts, Trans. AIME. 242, 12231230 (1968)

123