Mechanical Properties of Ultrahigh-Purity Ti 45 mol%al Alloy

Transcription

1 Materials Transactions, Vol. 43, No. 2 (22) pp. 163 to 167 Special Issue on Ultra-High Purity Metals (II) c 22 The Japan Institute of Metals Mechanical Properties of Ultrahigh-Purity Ti 45 mol%al Alloy Chikara Kawarada, Nobuyuki Harima, Seiichi Takaki and Kenji Abiko Institute for Materials Research, Tohoku University, Sendai , Japan An ultrahigh-purity Ti 45 mol%al alloy ingot was isothermally hot-pressed along each of the X, Y and Z coordinates at 1473 K. The hot-pressed ingot consisted mainly of very fine recrystallized grains, with an average size of 5 µm. The mechanical properties of this finegrained Ti Al alloy were examined by tensile test at 293 K, 823 K, and 923 K at strain rate of s 1 in a vacuum of Pa. The ductile-brittle transition temperature (DBTT) was determined by the tensile tests and scanning electron microscope observation of the fracture surfaces of the samples. This ultrahigh-purity Ti 45 mol%al alloy showed the following excellent high-temperature mechanical properties: (1) The.2% proof strength was 486 MPa and elongation at 923 K was 45%. (2) The DBTT was 87 K, which is about 2 K lower than that of conventional binary Ti Al alloys with similar composition. (Received November 15, 21; Accepted January 11, 22) Keywords: intermetallic compound, titanium aluminide, isothermal hot forging, microstructure, recrystallization, tensile properties, strength, ductility, grain size, fracture surface, ductile-brittle transition temperature 1. Introduction Ti Al alloys containing 45 to 5 mol%al are one of the most promising lightweight and heat-resistant alternatives to conventional heat-resistant steels and super-alloys. It is, however, difficult to process these Ti Al alloys because of their low ductility and poor workability. For example, the ductilebrittle transition temperature (DBTT) of binary Ti Al alloys is reported to be about 17 K; that is, ductility of these alloys is very low at temperatures below 17 K. There have been many studies attempting to improve the ductility of Ti Al alloys. Kawabata et al. 1) reported that the ductility at room temperature improved with increasing purity. Lipsitt et al. 2) investigated the DBTT of a Ti 4 mol%al (abbreviated hereafter as Ti 4Al) alloy made fine-grained by extrusion. They reported that the specimen tensile-tested at 173 K showed cleavage and intergranular fracture with some indications of ductility, and the DBTT of the Ti Al alloy was about 173 K. Hosomi and Maeda 3) investigated tensile properties of a Ti 48.5 mol%al (Ti 48.5Al) alloy made finegrained by isothermal hot pressing. These reports showed that grain fining was effective for the improvement of the ductility; however, the Ti Al alloys used in these reports contained a considerable quantity of impurities, and the ductility of these alloys was low. It is, therefore, very important to investigate the mechanical properties of higher-purity Ti Al alloys. The purification of Ti Al alloys is, however, very difficult. Ti Al alloys prepared by conventional methods have typically contained a considerable quantity of gaseous and metallic impurities, particularly oxygen at a level of more than 5 mass ppm. 4, 5) It is, therefore, necessary to prepare higherpurity Ti Al alloys for the clarification of their inherent properties. Concerning our current research series, Nakajima et al. 6) reported that a high-purity Ti Al alloy containing only 13 ppm oxygen was prepared by floating-zone melting under ultra-high vacuum (UHV); however, the volume was too small to measure its mechanical properties. Next, they pro- Present address: Nikko Materials Co., Ltd., Kitaibaraki , Japan. duced an 8 g ingot of Ti Al alloy by cold-crucible induction melting (CCIM) in UHV; 6) however, the ingot contained 85 ppm oxygen, and metallic impurities in the ingot were not analyzed. Recently we reported on the preparation by coldcrucible induction melting in UHV 7) of a 1 kg ultrahigh-purity Ti 45 mol%al (Ti 45Al) alloy of more than mass% purity after analysis of 4 elements. The oxygen content was 35 mass ppm and the total concentration of metallic and metalloid impurities was less than 9 mass ppm. The purpose of the present work is to investigate whether mechanical properties of Ti 45Al alloy with fine grains are improved by ultra-purification. 2. Experimental Procedure The preparation of the ultrahigh-purity Ti 45Al ingot is described in detail in Refs. 6) and 7). Table 1 shows the composition and total purity of the ultrahigh-purity Ti 45Al alloy. Four non-metallic impurities and thirty-six metallic and metalloid impurities were analyzed by chemical analysis and glow-discharge mass spectrometry (GDMS), respectively. The oxygen analysis of the Ti Al alloy was carefully performed in the same manner as described in Ref. 6); that is, the accuracy was improved by the addition of nickel as an accelerator for complete combustion of Ti Al alloy. If the analysis of oxygen in Ti Al alloys is performed by the generally used method, the value obtained is lower than the true value. 6) The purity of the alloy, which was more than mass% after analysis of 4 elements, includes the values of the detection limit for all elements. The ingot was isothermally hot-pressed at 1473 K with pressings along each of the X, Y, and Z coordinates. The key to obtaining an equiaxial fine grain structure by three-step isothermal hot forging (IHF) is to decompose the microstructure as cast and then produce the fine-grained microstructure through dynamic recrystallization. The forging was carried out in the following three steps: A first forging was performed with a strain rate of s 1 to change the ingot structure into a recrystallization structure. The reduction in height was 51%. A second forging was carried out for grain fining

2 164 C. Kawarada, N. Harima, S. Takaki and K. Abiko Table 1 Composition of ultrahigh-purity Ti 45Al alloy. Element Content C 4.9 N 11 O 35 S.3 Ag <.1 B <.1 Ba <.1 Be <.1 Bi <.1 Ca <1 Cd <.1 Co.3 Cr.8 Cu 5.8 Fe.42 Ga <.5 Ge <.5 Hf <.5 Hg <.5 In.2 K.3 Li <.1 Mg <.5 Mn <.1 Mo <.1 Na.3 Ni.6 P <.1 Pb <.1 Pd <.1 Pt <.1 Sb <.5 Sc <.2 Si.2 Sn <.5 Th.9 U.7 V.2 W.2 Zn <.1 Total <59.67 Purity > mass% (mass ppm) with a strain rate of.5 s 1, and reduction in height was 63%. A final forging with a strain rate of.5 s 1, and a reduction in height of 69% completed the process. These forgings were carried out in a vacuum of Pa. Specimens for tensile tests were cut from the hot-forged sample by electro-discharge machining and then mechanically polished. The gauge size of the specimens was w 1.5 l 7.6 t.7mm 3. The tensile tests were performed at 293 K, 823 K, and 923 K with a strain rate of s 1 in a vacuum of Pa. Table 2 Al-content of α 2 and γ phases quantitatively determined by EDX analysis. α 2 phase 38.7 γ phase Experimental Results and Discussion 3.1 Microstructure of isothermally hot-forged ingot Figure 1 shows the optical microstructure of the ultrahighpurity Ti 45Al alloy observed (a) as cast and (b) after isothermal hot forging. The microstructure as cast is a fully transformed lamellar structure composed of α 2 and γ phases. The lamellar grain sizes average about 5 µm in diameter. On the other hand, the microstructure as deformed consists mainly of very fine recrystallized grains, though a deformed lamellar structure partially remains as shown in this Fig. 1(b). The recrystallized grains are α 2 (Ti 3 Al) and γ (TiAl) phases with equiaxial grains with average diameters of several µm. Deformation twins are observed throughout the specimen. Figure 2 shows a scanning electron image of the ultrahighpurity Ti 45Al alloy observed after isothermal hot forging. Aluminum content of α 2 and γ phases was determined by energy dispersive X-ray spectrometer (EDX) analysis. The positions labeled α 2 and γ in this figure represent the locations of the EDX analysis and the results of phase identification. The results of the EDX analysis are summarized in Table 2. The average aluminum content of α 2 phase was 38.7%. The average aluminum content of γ phase was 47.%. This means that the equiaxial grains consist mainly of γ phase. 3.2 Tensile properties and fracture Figure 3 shows stress-strain curves of the ultrahigh-purity Ti 45Al alloy tensile-tested at 293 K, 823 K, and 923 K. At 293 K, the specimen fractured soon after yield. The ultimate tensile strength was 62 MPa, and the elongation was.5%. At 823 K and 923 K, work-hardening occurred after yield, and then the specimens fractured after uniform elongation. The ultimate tensile strengths were 86 MPa and 75 MPa, and the elongations 13% and 45%, respectively. Figure 4 shows the optical microstructure of fractured ultrahigh-purity Ti 45Al alloy after tensile test at (a) 823 K and (b) 923 K. Both showed uniform elongation without necking. At 823 K, the reduction in area was 13%. Some grains deformed slightly along the tensile direction throughout the specimen. At 923 K, the reduction in area was 4%. All grains were greatly deformed along the tensile direction. Figure 5 shows scanning electron micrographs of fracture surfaces for ultrahigh-purity Ti 45Al alloy after tensile test at (a) 293 K, (b) 823 K, and (c) 923 K. At 293 K, the fracture surface shows a brittle fracture. This is mostly an intergranular fracture, and partly a cleavage fracture. At 823 K, the fracture surface shows mostly intergranular brittle fracture with some indications of ductility. It was observed that at this temperature partial slip deformation takes place on the intergranular grain facets of the fracture surfaces. At 923 K, many dimples are observed throughout the specimen; that is, the fracture surface shows ductile fracture. Mutou et al. 8) investigated the fracture surfaces of conventional binary Ti Al alloys hav-

3 Mechanical Properties of Ultrahigh-Purity Ti 45 mol%al Alloy Fig Optical microstructure of ultrahigh-purity Ti 45Al alloy observed (a) as cast and (b) after isothermal hot forging K 8 /MPa 6 Stress, K 293K stress Fig. 3 Stress-strain curves of ultrahigh-purity Ti 45Al alloy tensile-tested at 293, 823 and 923 K with a strain rate of s 1. Fig. 2 Scanning electron image of ultrahigh-purity Ti 45Al alloy observed after isothermal hot forging. ing similar composition to the present sample. While they reported observing some dimpling at 173 K and dimpling throughout the specimen at 1273 K, in the present ultrahighpurity Ti Al specimen dimple fracture was observed at 923 K. 3.3 DBTT Figure 6 shows (a).2% proof stress and (b) elongation of Ti Al alloys as a function of testing temperature. Circles show the data obtained from the ultrahigh-purity Ti 45Al alloy, with the average grain size of 5 µm. Triangles show the data for the Ti 4Al alloy reported by Lipsitt et al.2) The average grain size was 25 µm. This alloy contained more than ppm impurities. Squares show the values for the Ti 48.5Al alloy reported by Hosomi and Maeda.3) The average grain size was 2 µm. This alloy contained more than 5 ppm impurities. These conventional binary Ti Al alloys with the average grain size of more than 2 µm and containing more than 5 ppm of impurities have been used in the experiments to improve the ductility. The values for.2% proof stress and the elongation of a conventional Ti 45Al alloy are supposed to be located between the values for the Ti 4Al alloy and the Ti 48.5Al alloy. In comparison with these predicted values, those for the present Ti 45Al alloy were remarkably improved. In Fig. 6(a), the.2% proof stress of the present Ti 45Al alloy is higher than that of the conventional binary Ti Al alloys at temperatures up to 923 K. The strength of metallic materials generally decreases after purification. It is, therefore, thought that the tensile strength of the ultrahigh-purity Ti Al alloy was improved by grain fining. In Fig. 6(b), if the temperature corresponding to 2% elongation is defined here to be DBTT, that of the present Ti 45Al alloy is about 87 K, which is lower by about 2 K than that of conventional binary Ti Al alloys with similar composition. It has been suggested that tensile ductility may be imparted to brittle intermetallic polycrystals by fining the grains to

4 166 C. Kawarada, N. Harima, S. Takaki and K. Abiko Fig. 4 Optical microstructure of fractured ultrahigh-purity Ti 45Al alloy after tensile test at (a) 823 and (b) 923 K. Fig. 5 Scanning electron micrographs of fracture surfaces for ultrahigh-purity Ti 45Al alloy after tensile test at (a) 293, (b) 823 and (c) 923 K. sizes smaller than a critical value.9) For example, Schulson and Barker1) examined the dependence of tensile elongation on grain size in Ni Al alloy, and reported that the ductility increases dramatically at grain sizes smaller than about 2 µm. Ni Al alloy, thus, exhibits a critical grain size below which polycrystalline aggregates are ductile in tension. However, for all grain sizes from 8 to 125 µm, fracture occurs in a brittle manner through a combination of intergranular de-cohesion and transgranular cleavage. In the present work, at 823 K, the fracture surface showed both ductile and brittle fracture. The deformation activity with slip takes place at this temperature, and the elongation is 13%. If the grain size of the ultrahigh-purity Ti Al alloy were below the critical grain size, its elongation would be much larger. Thus, it was most likely not below the crit- ical grain size, even if this behavior in Ni Al alloy appears in Ti Al alloy. On the other hand, the dimple fracture of the ultrahigh-purity Ti Al alloy was mainly observed at 923 K, though that of the conventional binary Ti Al alloy was observed at above 173 K. Thus we conclude that in the present work the DBTT of ultrahigh-purity Ti Al alloy was improved not by fining the grains but by purification. 4. Conclusions The ultrahigh-purity Ti 45 mol%al alloy with fine grains averaging 5 µm in size showed the following excellent hightemperature mechanical properties: (1) It showed a high.2% proof strength of 486 MPa and a large ductility of 45% at 923 K.

5 testing temperature (K) Mechanical Properties of Ultrahigh-Purity Ti 45 mol%al Alloy 167 /MPa (a) GS 5 pre p present work (Ti-45Al) r t Lipsitt et al. (Ti-4Al) Hosomi et al. (Ti-48.5Al) elongation (%) GS 2 (b) GS GS 2 GS GS Testing Temperature, T /K Testing Temperature, T /K Fig. 6 Temperature dependence of (a).2% proof stress and (b) elongation of Ti Al alloys. (2) Its DBTT was 87 K, which is about 2 K lower than that of conventional binary Ti Al alloy with similar composition. Acknowledgments The present work has been supported by the Japanese Ministry of Education, Culture and Science, and by the Program of Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Co (JST). We are grateful to Mr. K. Fukai, Japan Atomic Energy Research Institute for tensile test and Mr. Y. Nakamura, Nikko Materials Co., Ltd. for forging. REFERENCES 1) T. Kawabata, M. Tadano and O. Izumi: Scr. Matall. 22 (1988) ) H. A. Lipsitt, D. Shechtman and R. E. Schafrik: Metall. Trans. A 6A (1975) ) M. Hosomi and T. Maeda: Tetsu-to-Hagane 82 (1996) (in Japanese). 4) S. C. Huang and P. A. Siemers: Metall. Trans. A 2A (1989) ) C. McCullough, J. J. Valencia, C. G. Levi and R. Mehrabian: Acta Metall. 37 (1989) ) T. Nakajima, Y. Morimoto, S. Takaki and K. Abiko: Mater. Trans., JIM 41 (2) ) C. Kawarada, N. Harima, S. Takaki and K. Abiko: in the 7th international conference on ultrahigh-purity base metals (UHPM-2) held in Helsinki in 2, Phys. Stat. Sol. (a), in press. 8) R. Gnanamoorthy, Y. Mutoh, N. Masahashi and Y. Mizuhara: Metall. Trans. A 26A (1995) 35. 9) E. M. Schulson: Res. Mech. Letters 1 (1981) ) E. M. Schulson and D. R. Barker: Scr. Matall. 17 (1983)