HIP temperature and properties of a gas-atomized γ-titanium aluminide alloy

HIP temperature and properties of a gas-atomized γ-titanium aluminide alloy

Intermetallics 12 (2004) 63–68 www.elsevier.com/locate/intermet HIP temperature and properties of a gas-atomized g-titanium aluminide alloy Ulrike Ha...

393KB Sizes 0 Downloads 52 Views

Intermetallics 12 (2004) 63–68 www.elsevier.com/locate/intermet

HIP temperature and properties of a gas-atomized g-titanium aluminide alloy Ulrike Habel*, Brian J. McTiernan Crucible Research LLC, Pittsburgh, PA 15205, USA Accepted 26 August 2003

Abstract The effect of HIP-temperature on the tensile properties of a powder metallurgical g-titanium aluminide based alloy has been investigated in this study. Prealloyed, gas-atomized Ti–46Al–2Cr–2Nb powder was HIP’ed at temperatures between 1200 and 1300  C. The observed as-HIP microstructures reflect the path of the HIP-process in the a+g, a2+g and a+b+g phase fields. Yield stresses in the as-HIP conditions range from 460 to 715 MPa. They are accompanied by ductilities of 0.2 –1.7%. After heat treating to achieve fine duplex microstructures, yield stress levels from 515 to 585 MPa and ductilities from 0.4 to 1.6% are observed. Ductility improvements seem to be related to increasing amounts of lamellar grains and b-phase with concomitant lower strain hardening rates. The effect of HIP-temperature on tensile properties is small when microstructural differences are largely eliminated. # 2003 Elsevier Ltd. All rights reserved. Keywords: A. Titanium aluminides, based on TiAl; B. Mechanical properties at ambient temperature, work hardening; C. Powder metallurgy

1. Introduction Processing issues are becoming more and more important as g-titanium aluminide alloys approach the application stage [1,2]. A viable process for producing fully dense titanium aluminide alloys is inert gas atomization followed by hot isostatic pressing (HIP). The inherently high cooling rate of gas atomization suppresses segregation and results in uniform composition. Small compositional variations can have a significant effect on strength and ductility of g-TiAl based alloys, making the uniform composition obtained by PM (powder metallurgy) processing of prealloyed powders particularly attractive [3,4]. Consistent process scale increases and yield improvements have resulted in substantial manufacturing cost reductions making PM titanium aluminides a realistic alternative high temperature material [5]. The HIP process is conducted in the a+g, a+b+g, a2+g and/or a+b+g phase fields. Early work on the effect of HIP temperature on mechanical properties is * Corresponding author at present address: 6003 Campbells Run Road, Pittsburgh, PA 15205, USA. Tel.: +1-412-923-2955x227; Fax: +1-412-788-4665. E-mail address: [email protected] (U. Habel). 0966-9795/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.intermet.2003.08.006

restricted to the investigation of as-HIP microstructure and properties only [6–8]. The as-HIP microstructure depends strongly on the specific path of the HIP-process in the different phase fields, therefore effects of HIP-temperature on microstructure are superimposed on the effects of HIP-temperature itself in addition to its influence on the microstructure. This study has been designed to isolate the effect of HIP-temperature on the mechanical properties from the effect of HIP-temperature on microstructure by investigating materials with comparable microstructures. To this end, prealloyed gas-atomized g-titanium aluminide based powder was HIP’ed at a range of temperatures and then heat treated to achieve the same microstructure for materials HIP’ed at different temperatures.

2.. Experimental Prealloyed Ti–46Al–2Cr–2Nb powder with a small B-addition was inert gas atomized and HIP’ed at temperatures of 1200, 1240, 1270 and 1300  C at a pressure of 100 MPa for 2 h. The analyzed composition is Ti– 45.98Al–2.01Cr–2.03Nb–0.22B; the impurity levels are 300 wppm Fe, 560 wppm O, 60 wppm N, 180 wppm C, and 20 wppm H. All compacts were evaluated in the asHIP and HIP plus heat treated conditions. Blanks for

64

U. Habel, B.J. McTiernan / Intermetallics 12 (2004) 63–68

microscopic investigations and tensile tests were heat treated at 1255, 1275 and 1305  C followed by air cooling (single-step) and 1250 followed by either air cooling or furnace cooling plus 900  C (two-step). The grain size was determined using a linear intercept method according to ASTM E112. Further details of the experimental procedure are given in [5,9].

3. Results and discussion The as-HIP microstructures are shown in Fig. 1. The as-HIP materials are fully dense as typical for the HIP PM route. They have fine and isotropic microstructures without requiring thermo-mechanical treatment; the microstructures are independent of compact sizes [5,10]. After HIP at 1200  C, PM Ti–46Al–2Cr–2Nb has a near-g microstructure with a 5 mm grain size. It contains small amounts of a2- and b-phases, the latter appearing light gray and white, respectively, in backscatter electron imaging (Fig. 1a). The low Al-content and the Cr-

addition place Alloy 46-2-2 in the three-phase region constituted of a2-, b-, and g-phases during HIP at 1000 and 1100  C and of a-, b-, and g-phases during HIP at 1200  C [11,12]. Due to the low HIP-temperature, very fine grains of the constituting phases form. The higher HIP temperatures, 1240–1300  C, correspond to the a+g two-phase field for both alloys, but during cooling in the HIP-vessel, both alloys transverse the three-phase fields, where b-phase is precipitated [13]. While older literature reports only the presence of b-phase with the A2 crystal structure in the Ti–Al–Cr system [14], Jewett and Dahms and Hao et al. observe the presence of bphase with both A2- and ordered B2-structure between 800 and 1000  C [12,15]. Kelly et al. and Inkson et al. describe B2 phase for a TiAl-based alloy heat treated at 1000  C [16]. At SEM resolution, the volume fraction of bphase is fairly constant between 2 and 4% for all as-HIP conditions, too low to distinguish between both structures using XRD, therefore TEM-investigations would be necessary to do this. The designation b-phase is used throughout this paper for the b-phase with a B2 superstructure.

Fig. 1. As-HIP microstructures of PM Ti–46Al–2Cr–2Nb (backscatter electron images; the a2-phase appears light gray, the b-phase white) (a) HIP at 1200  C; (b) HIP at 1240  C; (c) HIP at 1270  C; (d) HIP at 1300  C.

U. Habel, B.J. McTiernan / Intermetallics 12 (2004) 63–68

Increasing the HIP-temperature to 1240  C results in a three-phase microstructure with g as the majority phase; lamellar colonies and b-phase are also present (Fig. 1b). A slightly coarser microstructure dominated by lamellar colonies is obtained by HIP at 1270  C as shown in Fig. 1c. The microstructures obtained for both HIP-temperatures are essentially the duplex microstructure but for the additional presence of the b-phase. Grain size is 6 and 8 mm, respectively, for the materials HIP’ed at 1240 and 1270  C. The highest HIP temperature of 1300  C leads to a near lamellar microstructure with coarse lamellar spacing and a colony size of 35 mm (Fig. 1d). The small colony size can be attributed to pinning of the large agrains by the small g-grains during HIP at 1300  C. In addition to the b-phase on the grain boundaries, bright appearing, rod-shaped Ti- and Nb-rich particles are also visible after HIP at 1200 and 1240  C, but not after HIP at 1270 and 1300  C. Qualitative EDX-analysis indicates that these articles are rich in Ti and Nb. It is unclear at this point whether these are Ti-borides similar to those observed in [16,17] or b-phase. PM Ti–46Al–2Cr–2Nb heat treated at 1255  C and air cooled exhibits a duplex microstructure with a grain size of 7.5 mm and finely spaced lamellae (Fig. 2a). Very fine b-phase is present at colony boundaries following air cooling, while the slower cooled microstructure after HIP at 1240  C exhibits larger, blocky b-grains and also a higher fraction of g-grains (Fig. 1b). Duplex microstructures with a higher volume fraction of lamellar colonies are achieved by heat treating at a temperature of 1275  C and air cooling (Fig. 2b). In the g-phase, microcracks are present. They are sometimes formed during grinding, and their edges are rounded during polishing, which makes them appear wider. The grain sizes are 10 and 12 mm, respectively, for the materials HIP’ed at 1200 and 1240  C and then heat treated. The microstructure following HIP at 1270  C exhibits higher amounts of g- and b-phases than the materials HIP’ed at lower temperatures, heat treated at 1275  C and air cooled again due to the differences in cooling rate. The 1305  C heat treatment of materials HIP’ed at 1200 to 1270  C results in a near-lamellar microstructure with a colony size of 20 mm and fine b-phase (Fig. 2c). Differences in cooling rate account for the observed differences in lamellar spacing and presence of b-phase compared to the as-HIP at 1300  C material. The very small colony size following heat treating at 1305  C can be attributed to a combination of intermediate HIP-temperature and pinning of the large agrains by the small g-grains during the heat treatment. After the 1250 plus 900  C two-step heat treatments, three-phase microstructures are observed consisting of equiaxed g-grains, lamellar grains and b-phase as shown in Fig. 2d and e. The b-phase is preferentially located at the ends of the lamellae. Depending on their thermal

65

history, the microstructures reveal differences: The material, which was air cooled after the first heat treatment step, contains more g- and b-phase than the material, which was furnace cooled after the first heat treatment step. During air cooling, high temperature agrains are transformed into lamellar colonies, and only a very small amount of b-phase forms, while during furnace cooling, diffusion takes place, which allows for the formation of larger g- and b-grains (Fig. 2a and c). Furthermore, there is a difference resulting from the difference in HIP-temperature: the material HIP’ed at 1240  C has a slightly finer microstructure than the material HIP’ed at 1270  C. During the second heat treatment step at 900  C, additional b-phase forms, preferentially at the ends of the lamellae (Fig. 2e). The distribution of b-phase plays an important role for the ductility of PM Ti–46Al–2Cr–2Nb as will be discussed below. Cr is known as a b-stabilizer in g-TiAl based alloys [12,14]. Very fine b-phase precipitates are present after air cooling from the first heat treatment step as seen in Fig. 2a and b. The amount of b-phase is larger for the conditions cooled in the HIP-unit than for the air cooled conditions, and still larger following the two-step heat treatment (Figs. 1 and 2d, e). This is consistent with previous observations that b-precipitates grow in size and in volume fraction during heat treatments at temperatures below the a+g two phase field [12,18]. This refers to the low-temperatures b-phase field present in some ternary TiAl-based alloys [11–13], not to the high-temperature b-phase field present in the binary Ti–Al system as well as in ternary TiAl-based alloys. The tensile properties in the as-HIP conditions vary considerably as a function of the HIP temperature (Table 1). The ductility increases with increasing HIPtemperatures from 0.2% after HIP at 1200  C to 1.7% after HIP at 1300  C. This improvement in ductility is accompanied by a drop in yield stress from 715 to 445 MPa. The tensile properties in the as-HIP conditions are in good agreement with the literature in that higher HIP-temperatures lead to higher ductilities and lower yield stresses [7–9]. The fine grain size, high Al-content, and near-g microstructure of PM Ti–46Al–2Cr–2Nb HIP-ed at 1200  C result in extremely high strength and very low ductility as can be expected for this microstructure and composition [4,19]. Higher HIP-temperatures increase the content of lamellar colonies at the expense of equiaxed g-grains and coarsen of the microstructures, which, consistent with basic microstructural observations, reduce yield stress [20]. The reduction in equiaxed g-grains accompanied by the formation of fine lamellar colonies, fine both in size and lamellar spacing, improves ductility [20]. The conventional wisdom that lamellar microstructures show a reduced ductility holds only for coarse lamellar colonies. Heat treating 1 h at 1255 or 1275  C leads to similar properties for both conditions submitted to the same

66

U. Habel, B.J. McTiernan / Intermetallics 12 (2004) 63–68

heat treatments despite differences in HIP-temperatures. The 1255  C heat treatment decreases the fraction of equiaxed g-grains compared to the as-HIP conditions, which leads to a significant decrease in yield stress. This is a well established relationship [20]. For the material HIP’ed at 1200  C, this heat treatment increases the

ductility while for the material HIP’ed at 1240  C, it decreases the ductility compared to the respective as-HIP condition. While the lower yield stress in the heat treated condition can be expected from the reduced amount of g-phase, it is initially surprising to have this accompanied by a lower ductility. Comparing both

Fig. 2. Microstructures of PM Ti–46Al–2Cr–2Nb after heat treating (backscatter electron images; the a2-phase appears light gray, the b-phase white). (a) 1 h at 1255  C, AC (HIP at 1240  C); (b) 1 h at 1275  C, AC (HIP at 1270  C); (c) 20 min at 1305  C, AC (HIP at 1270  C); (d) 2 h at 1250  C, FC plus 4 h at 900  C (HIP at 1240  C); (e) 2 h at 1250  C, AC plus 4 h at 900  C (HIP at 1270  C).

67

U. Habel, B.J. McTiernan / Intermetallics 12 (2004) 63–68

Heat treating PM Ti–46Al–2Cr–2Nb at 1275  C increases the fraction of lamellar colonies with a concomitant improvement in both ductility and strength relative to the 1255  C heat treatment (Table 2). It is surprising that the 1275  C heat treatments lead to similar or even slightly higher yield strengths than the 1255  C heat treatments following the same HIP-temperature, since the higher temperature increases the fraction of lamellar colonies and slightly coarsens the grains, therefore one would expect the lower temperature heat treatment to result in a higher yield strength. Again, the tensile properties following heat treatment are very similar despite the differences in HIP-temperature. The materials heat treated at 1275  C exhibit higher yield stress levels than the as-HIP at 1270  C condition, which is consistent with the observed microstructural differences. Fine lamellar colonies are achieved by air cooling compared to coarser lamellae formed during the slower cooling such as inside a HIP-vessel, and the yield

microstructures, it is apparent that the slower cooled asHIP condition contains a larger amount of b-phase (Figs. 1d and 2c). The as-HIP condition also shows a strain hardening rate of 7 GPa compared to 18 and 24 GPa for the heat treated conditions (Tables 1 and 2; strain hardening rates were measured at 0.5% plastic deformation). These observations suggest that the differences in ductility and strain hardening rate are related to the amount of b-phase present as discussed below.

Table 1 As-HIP tensile properties of PM Ti–46Al–2Cr–2Nb HIP Temperature

0.2% YS (MPa)

UTS (MPa)

Tensile Elongation (%)

Hardening Rate (GPa)

1200  C 1240  C 1270  C 1300  C

715 610 525 445

715 665 660 590

0.2 1.0 1.6 1.7

N/A 7 11 11

Table 2 Tensile properties of HIP and heat treated PM Ti–46Al–2Cr–2Nb HIP Temperature

Heat Treatment

0.2% YS (MPa)

UTS (MPa)

Tensile elongation (%)

Hardening rate (GPa)

1200  C 1240  C 1200  C 1240  C 1240  C 1270  C 1270  C

1 h 1255  C, AC 1 h 1255  C, AC 1 h 1275  C, AC 1 h 1275  C, AC 2 h 1250  C, FC+4 h 900  C 2 h 1250  C, AC+4 h 900  C 20 min 1305  C, AC

560 540 570 555 585 545 515

650 605 690 715 670 710 675

0.5 0.4 0.9 1.1 1.6 1.4 1.1

18 24 18 19 7 15 17

Fig. 3. Ductility of Ti–46Al–2Cr–2Nb in as-HIP and HIP plus heat treated conditions.

68

U. Habel, B.J. McTiernan / Intermetallics 12 (2004) 63–68

strength is known to increase according to a Hall-Petch relationship with lamellar spacing as determining parameter [21]. For the HIP and heat treated conditions, the yield stress decreases somewhat with increasing HIPtemperature, which can likely be attributed to the finer grain size following the lower HIP-temperature. The 1250 plus 900  C heat treatment leads to a significant increase in ductility compared to the 1255  C single-step heat treatment despite the presence of a larger amount of equiaxed g-grains after the two-step heat treatment (Table 2). The higher ductility is again accompanied by a larger amount of b-phase and lower strain hardening rates. It is unclear whether the increase in amount of b-phase accounts for the higher yield stress following the two-step heat treatment. In the tensile tests reported here, improvements in ductilities are associated with increasing amounts of bphase. This agrees with results by Morris and Li, who found that the b-phase in a Ti–44Al–2Mo alloy had a fairly good ductility [22]. Further analysis of the tensile properties obtained in this study shows that the ductility is inversely related to the hardening rate (Fig. 3). This observation may infer that the b-phase strain hardens less than the g-phase or lamellar a2+g-colonies and therefore may accommodate more deformation, but it cannot definitely be concluded at the current stage of investigations.

4. Conclusions Microstructural differences as a result of different HIP temperatures play a significant role for the as-HIP properties. When microstructural differences are largely eliminated, the effect of HIP-temperature on mechanical properties of PM Ti-46Al-2Cr-2Nb is small in the temperature range explored in this study. The better as-HIP ductilities observed for higher HIP-temperatures are largely a consequence of higher fractions of lamellar colonies and the presence of b-phase. On the other hand, higher HIP-temperatures are not necessarily an advantage as they result in somewhat coarser microstructures and concomitant lower yield stress. In general, ductility improvements are accompanied by lower strain hardening rates.

Acknowledgements The authors thank J.E. McCalla, J.M. Connors, C.F.Yolton, and C.E. Rader for experimental assistance and technical discussions.

References [1] LeHolm R, Clemens H, Kestler H. In: Kim Y-W, Dimiduk DM, Loretto MH, editors. Gamma titanium aluminides. Warrendale (USA): TMS; 1999. p. 25. [2] Bartolotta PL, Krause DL. In: ibid., p. 3. [3] Huang SC, Hall EL. Metal Trans A 1991;22A:427. [4] Habel U, Yolton CF, Moll JH. Metal Trans A 1991;22A:301. [5] Moll JH, McTiernan BJ. Metal Powder Report 2000;55(1):18. [6] Beddoes JC, Wallace W, de Malherbe WC. International Journal of Powder Metallurgy 1992;28:313. [7] Zhao L, Beddoes J, Morphy D, Wallace W. Materials and Manufacturing Processes 1994;9:695. [8] Zhang G, Blenkinsop PA, Wise MLH. In: Blenkinsop PA, Evans WJ, Flower HM, editors. Titanium 95. The Institute of Material, 1995. p. 542. [9] Yolton CF, Moll JH. In: ASM Handbook; vol. 7. ASM Internationals, 1998. p. 164. [10] Clemens H, Glatz W, Schretter P, Yolton CF, Jones PE, Eylon D. In: Kim Y-W, Wagner R, Yamaguchi M, editors. Gamma titanium aluminides. Warrendale (USA): TMS; 1995. p. 555. [11] Kainuma R, Fujita Y, Mitsui H, Ohnuma I, Ishida K. Intermetallics 2000;8:855. [12] Hao YL, Yang R, Cui YY, Li D. Acta Mater 2000;48:1313. [13] Ohnuma I, Fujita Y, Mitsui H, Ishikawa K, Kainuma R, Ishida K. Acta Mater 2000;48:3113. [14] Petzow G, Effenberg G. Ternary alloys, vol. 5. Weinheim, Germany: VCH; 1988. [15] Jewett TJ, Dahms M. Sc Metal Mater 1995;32:1533. [16] Inkson BJ, Clemens H. In: George EP, et al., editors. High temperature ordered intermetallics VIII, (p.KK3.12.1). Boston (USA): MRS; 1999. [17] De Graef M, Hardwick DA, Martin PL. In: Nathal MV, Darolia R, Liu CT, Martin PL, Miracle DB, Wagner R, Yamaguchi M, editors. Structural intermetallics. Warrendale (USA): TMS; 1997. p. 185. [18] Habel U, Yolton CF, McTiernan BJ. In: Proc. Euromat 1999, vol. 10, Frankfurt/M, Germany. DGM, p. 373. [19] Shih DS, Huang S-C, Scarr GK, Jang H, Chesnutt JC. Microstructure/property relationships. In: Kim Y-W, Boyer RR, editors. Titanium aluminides and alloys. Warrendale (USA): TMS; 1991. p. 135. [20] Dimiduk DM, Hazzledine PM, Parasarathy TA, Seshagiri S, Mendiratta MG. Metal Mater Trans A 1998;29A:37. [21] Liu CT, Schneibel H, Masziaz PJ, Wright JL, Easton DS. Intermetallics 1996;4:429. [22] Morris MA, Li YG. In: Kim Y-W, Wagner R, Yamaguchi M, editors. Gamma titanium aluminides. Warrendale (USA): TMS; 1995. p. 353.