Transmission Electron Microscopy Study of the Infrared Brazed High-strength Titanium Alloy

Transcription

1 J. Mater. Sci. Technol., 2010, 26(4), Transmission Electron Microscopy Study of the Infrared Brazed High-strength Titanium Alloy Z.Y. Wu 1), R.K. Shiue 1) and C.S. Chang 2) 1) Department of Materials Science and Engineering, National Taiwan University, Taipei 106, Taiwan, China 2) Engineered Materials Solutions, 39 Perry Avenue, MS 4-1, Attleboro, MA , USA [Manuscript received March 7, 2009, in revised form March 3, 2010] The transmission electron microscopy was employed to investigate the microstructure of infrared brazed highstrength Ti alloy using the Ti-15Cu-15Ni filler metal. Coarse primary Ti 2 Ni and transformed β-ti are observed in the 300 s brazed specimen. Blocky Ti 2 Ni and eutectoid Ti 2 Cu intermetallics are disappeared from the joint with increasing the brazing time to 1800 s. Both acicular α-ti and retained β-ti dominate the entire brazed joint. KEY WORDS: Infrared Brazing; Ti; Ti-Cu-Ni filler; Microstructure 1. Introduction Ti brazing has been carried out using Ag- and Al-based filler metals, but the service temperature of such brazed joints is confined below 400 C [1,2]. Joints applied for elevated temperature service are frequently brazed using Ti-based braze alloys of which Cu and Ni are commonly added as melting point depressants (MPDs) [3,4]. The brazing of Ti alloys using clad Ti-Cu-Ni filler metals has been extensively studied [5 8]. The brazed joint demonstrates excellent bonding strength, so the clad Ti-Cu-Ni filler metal is considered as one of the best choices in brazing highstrength Ti alloys for structural application [5 8]. It is difficult for using the conventional furnace brazing to investigate the microstructural evolution of the brazed joint due to its slow heating rate. On the other hand, infrared brazing is very suitable for studying the microstructural evolution of the brazed joint with the aid of its rapid heating rate up to 50 C/s. Therefore, it has been applied to study the kinetics of brazing in recent years [8,9]. The Ti alloy experiences an α-β phase transformation during the brazing process, so transmission electron microscopy (TEM) Corresponding author. Prof.; Tel.: , Fax: ; address: rkshiue@ntu.edu.tw (R.K. Shiue). study of the brazed joint is necessary in order to unveil the transformation of the joint in greater depth. Both alloying elements of the Ti-based fillers diffusing into the substrate and dissolution of the substrate into the brazed joint complicate the transformation kinetics of the brazed joint. However, detailed TEM study of the brazed joint using clad Ti-Cu-Ni filler metal is still unavailable. Based on the previous study of infrared brazing Ti alloys using Ti-Cu-Ni fillers, the microstructural evolution of the joint is primarily depended on the redistribution of both Cu and Ni contents across the brazed joint [8,9]. There are at least two phases readily distinguished from the brazed joint. One is the Cu-Ni rich Ti phase, the other is the Ti-rich matrix. The amount and morphology of Cu-Ni rich Ti phase are sensitive to filler metal composition, brazing temperature and time [9]. Additionally, the presence of Cu-Ni rich Ti phase deteriorates the shear strength of brazed joint, in particular when the Cu-Ni rich Ti phase forms a continuous layer [8]. It is also noted that the amount of Cu-Ni rich Ti phase is decreased with increasing the brazing temperature and/or time. Accordingly, shear strengths of infrared brazed joints are increased with increasing the brazing temperature and/or time while fracture morphology changes from cleavage type brittle fracture into dimple ductile type

2 312 Z.Y. Wu et al.: J. Mater. Sci. Technol., 2010, 26(4), Position Al Cu Fe Mo Ni Ti V Phase A 0.8/ / /1.8 0/0 26.0/ / /0.3 Primary Ti 2Ni B 5.6/ / / / / / /1.9 Ti-rich C 5.0/ / / / / / /1.7 Ti-rich D 6.0/ / / / / / /2.0 Ti-rich E 7.1/ / / / / / /2.3 Ti-rich F 7.9/ / / / / / /2.7 Ti-rich(SP-700) Fig. 1 EPMA SEIs and WDS chemical analysis results of SP-700 joint using Ti-15Cu-15Ni filler infrared brazed at 970 C for (a) 300 s, (b) 1800 s. The data in the table are in the format of at. pct/wt pct fracture [8]. SP-700 (4.5 wt pct Al, 3 wt pct V, 2 wt pct Fe and 2 wt pct Mo) is a β-rich α-β Ti alloy developed particularly to yield a superfine microstructure that provides superplastic forming capability at 700 C [10]. Infrared brazing SP-700 using Ti-15Cu-15Ni filler metal is conducted in this research. TEM analysis of the infrared brazed joint is performed in order to unveil the transformation of the brazed zone in greater depth. 2. Experimental Infrared vacuum brazing SP-700 alloy was performed at 970 C for 300 s and 1800 s, respectively. Ti-15Cu-15Ni foil in wt pct with the thickness of 50 µm was selected as the braze alloy. The heating rate was set at 10 C/s, and all samples were preheated at 800 C for 300 s before heating up to the brazing temperature. The cross section of the brazed joint was examined using a JEOL JXA 8600SX electron probe microanalyzer (EPMA) equipped with a wavelength dispersive spectroscope (WDS). The acceleration voltage was 15 kv, and its minimum spot size was 1 µm. For detailed microstructural observations, transmission electron microscopy (TEM) specimens were sectioned in thin slices within the brazed zone of the joint. Thin foils were prepared by a standard jet-polisher using an electrolyte of 6% HClO 4, 30% C 2 H 5 OH and 64% CH 3 COOH at room temperature. The operation voltage is 30 V and the current is ma. Thin foil specimens were examined using a Philips TECNAI G2 TEM operated at 200 kv equipped with an energy dispersive spectroscopy (EDS) for chemical analysis of selected area in the brazed zone. 3. Results and Discussion Figure 1 shows EPMA secondary electron images (SEIs) and WDS chemical analysis results of SP-700 joint using clad Ti-15Cu-15Ni filler infrared brazed at 970 C for 300 s and 1800 s, respectively. It is obvious that microstructures of brazed joints are strongly related to the infrared brazing time. For the 300 s brazed specimen, primary Ti 2 Ni is widely observed in the brazed zone as marked by A in Fig. 1(a), and the Ti-rich matrix is found as marked by B in the figure. In contrast, the blocky Ti 2 Ni is completely disappeared from the joint for the 1800 s brazed specimen and there is only Ti-rich matrix left in the brazed zone as marked by C, D and E in Fig. 1(b). The Ti-rich matrix cannot be accurately identified via EPMA observation even using higher magnification. According to the EPMA chemical analysis results, the Ti-rich matrix is alloyed with Cu and Ni contents greatly exceeding their solubilities in β-ti [11]. It is deduced that the Ti-rich matrix in Fig. 1 may consist of more than one phase. The nominal composition of the brazing foil in wt pct is 70Ti-15Cu-15Ni. Based on the related binary alloy phase diagrams, the maximum solubility of Cu and Ni in the β-ti is 17 wt pct and 12 wt pct, respectively [11]. The disappearance of blocky Ti 2 Ni in the brazed joint is primarily attributed to high solubility of Ni in the β-ti. Both dissolution of SP-700 substrate into the molten braze and diffusion of Ni into SP-700 substrate result in depletion of Ni from the brazed zone during infrared brazing. Similarly, depletion of Cu from the infrared brazed zone also strongly depends on the brazing condition. Longer

3 Z.Y. Wu et al.: J. Mater. Sci. Technol., 2010, 26(4), Position Al Cu Fe Mo Ni Ti V Phase A 5.1/ / / / / / /0.3 Acicular α-ti B 3.1/ / / / / / /3.6 Retained α-ti Fig. 2 TEM micrographs and EDS chemical analysis results of region IV in Fig. 1(b) infrared brazed at 970 C for 1800 s: (a) BF image of the acicular α-ti and retained β-ti, (b) DF image using a (42 2) retained β-ti spot from position C, (c) DF image using a (11 01) acicular α-ti spot from position C, (d) BF image of acicular α-ti and retained β-ti. EDS spectra of positions A and B in Fig. 2(d) are included. The data in the table are in the format of at. pct/wt pct brazing time such as 1800 s favors the infrared brazed joint free of blocky Ti2 Ni, and transformed β-ti dominates the entire brazed joint as illustrated in region IV of Fig. 1(b). The transformation of β-ti in the brazed zone upon cooling cycle of brazing is related to its chemical composition and cooling rate. The average cooling rate between 970 C and 600 C during infrared brazing is 1.5 C/s. TEM examinations of various regions in the brazed joint are necessary in order to unveil microstructures in grater depth. There is no coarse Ti2 Ni in region IV of the brazed zone, and the Ti-rich matrix dominates the entire region as illustrated in Fig. 1(b). Figure 2(a) displays bright field (BF) image of the region IV. Based on the selected area diffraction pattern (SADP) analysis of position C in Fig. 2(a), the Ti-rich matrix mainly consists of the retained βti (Fig. 2(b)) and needle-like phase (Fig. 2(c)) trans-

4 314 Z.Y. Wu et al.: J. Mater. Sci. Technol., 2010, 26(4), Position Al Cu Ni Ti V Phase A 5.4/ / / /0.3 Primary Ti 2Ni B 21.0/ / /61.2 Primary Ti 2 Cu C 17.6/ / /62.7 Lamellar eutectoid Ti 2 Cu D 5.3/ /97.0 Lamellar eutectoid α-ti Fig. 3 TEM micrographs and EDS chemical analysis results of region I in Fig. 1(a) infrared brazed at 970 C for 300 s: (a) BF image of coarse primary Ti 2 Ni and Ti 2 Cu, (b) BF image of lamellar eutectoid Ti 2 Cu and α-ti. EDS spectra of positions A, B, C and D in the figure are included. The data in the table are in the format of at. pct/wt pct formed from the β-ti during brazing. The needlelike phase has a hexagonal structure, and it is categorized as the acicular β-ti. Figure 2(d) shows the microstructure of higher magnification. The retained β-ti is identified next to the acicular β-ti as marked by B in Fig. 2(d), and it is alloyed with 1.7 wt pct Al, 1.2 wt pct Cu, 5.1 wt pct Fe, 7.5 wt pct Mo, 4.3 wt pct Ni, and 3.6 wt pct V, which are all β stabilizers except for the Al. It is worth mentioned that contents of β stabilizers, Fe, Mo and V, in the acicular α-ti are significantly less than those in the retained β-ti due to partition of these elements in brazing. As described earlier, the depletion of Cu and Ni from the molten braze into SP-700 substrate is not prominent for the brazed joint with a short brazing cycle such as 300 s. Figure 3 displays BF images and EDS chemical analysis results of region I in Fig. 1(a). Coarse primary Ti 2 Ni as marked by A are observed in Fig. 3(a) due to the Ni content (15 wt pct) of the braze alloy exceeding the maximum solubility of Ni

5 Z.Y. Wu et al.: J. Mater. Sci. Technol., 2010, 26(4), Position Al Cu Ni Ti Phase A 5.3/ / / /96.1 Non-lamellar eutectoid α-ti B 20.1/ / /61.7 Non-lamellar eutectoid Ti 2 Cu Fig. 4 The TEM micrographs and EDS chemical analysis results of region II in Fig. 1(a) infrared brazed at 970 C for 300 s: (a) BF image of non-lamellar eutectoid Ti 2 Cu and α-ti, (b) indexed SADP of position B, (c) DF image using a (0 111) α-ti spot, (d) DF image using a (004) Ti 2 Cu spot. EDS spectra of positions A and B are included. The data in the table are in the format of at. pct/wt pct (12 wt pct) in the β-ti [12]. On the other hand, the Cu content (15 wt pct) of the braze alloy is slightly lower than the maximum solubility of Cu (17 wt pct) in the β-ti, so much less amount of Ti 2 Cu as marked by B in Fig. 3(a) is observed in the experiment. Based on the EDS chemical analysis results, Ti 2 Ni dissolves 6.6 wt pct Cu, and Ti 2 Cu dissolves 13.4 wt pct Ni in Fig. 3(a). It is obvious that the amount of Cu dissolved in Ti 2 Ni is much less than that of Ni in Ti 2 Cu. Besides the primary Ti 2 Ni in region I of the brazed zone, the lamellar mixture of Ti 2 Cu and α-ti is also identified in the BF image of Fig. 3(b). The morphology of lamellar Ti 2 Cu (marked by C in Fig. 3(b)) and α-ti (marked by D in Fig. 3(b)) can be categorized as a eutectoid, and the lamellar spacing of eutectoid is below 0.5 µm. The formation of eutectoid is resulted

6 316 Z.Y. Wu et al.: J. Mater. Sci. Technol., 2010, 26(4), from the decomposition of β-ti alloyed with high content of Cu upon cooling cycle of brazing. The morphology of eutectoid α-ti is very different from that of acicular α-ti. The eutectoid α-ti is alloyed with 3.0 wt pct Al and free of Fe, Mo, Ni and V, β-ti stabilizers. It is worth mentioning that the eutectoid contains no Ti 2 Ni. For the β eutectoid system, the alloying element stabilizes the β phase, but the β phase can transform to α plus another phase of compound [12]. The β eutectoid alloying elements are of two types: rapid or active eutectoid former e.g., Cu, and slow or sluggish eutectoid former, e.g. Ni [12]. Therefore, rapid decomposition of the β phase to produce Ti 2 Cu and α-ti is observed in the experiment. Additionally, the formation of coarse Ti 2 Ni in region I consumes the Ni content in the rest of braze alloy. The eutectoid Ti 2 Cu is alloyed with 15.8 wt pct Ni as marked by C in Fig. 3(b), and it is consistent with the previous reports that Ti 2 Cu dissolves Ni up to 16.9 wt pct [13,14]. All of these reasons favor the formation of eutectoid Ti 2 Cu instead of Ti 2 Ni. Figure 4(a) shows BF image and EDS chemical analysis of region II in Fig. 1(a) infrared brazed at 970 C for 300 s. The coarse primary Ti 2 Ni is disappeared, and non-lamellar eutectoid dominate the microstructure of region II. Based on the SADP analysis result, non-lamellar eutectoid α-ti (Fig. 4(c)) and Ti 2 Cu (Fig. 4(d)) is identified from this region. The morphology of eutectoid in Fig. 4(a) is different from that in Fig. 3(b) due to different Cu contents in the brazed zone. The eutectoid morphology is nonlamellar in the hypoeutectoid alloy and lamellar in the near eutectoid alloy [15]. The Cu content in region II is lower than that in region I due to depletion of the Cu from central brazed region into SP-700 substrate. Therefore, non-lamellar eutectoid is observed in region II. Based on the current SADP analysis results, there is no orientation relationship between the phases in 300 s and 1800 s brazed specimens. The microstructure of region III in Fig. 1(a) is similar to that of region IV in Fig. 1(b). The decomposition of β-ti upon cooling cycle of brazing plays an important role in microstructural evolution of the brazed joint. Because the cooling rate of infrared brazing is kept the same for all specimens, chemical composition of the β-ti is the crucial factor in determining microstructures of the brazed zone. Accordingly, the transformation of β-ti is strongly related to the redistribution of Cu and Ni across the infrared brazed joint. 4. Conclusions Transmission electron microscopy study of the infrared brazed SP-700 alloy using Ti-15Cu-15Ni has been performed in the experiment. There are at least three zones readily distinguished from the 300 s brazed specimen. For the highest contents of Cu and Ni, coarse primary Ti 2 Ni/Ti 2 Cu and lamellar eutectoid Ti 2 Cu/α-Ti dominate the central region of the brazed joint. Next to the central region of the brazed zone, the coarse Ti 2 Ni/Ti 2 Cu is disappeared, and non-lamellar eutectoid Ti 2 Cu and α-ti are widely observed. As Cu and Ni contents of the brazed zone are further decreased, the β-ti is decomposed into acicular α-ti and retained β-ti. Increasing the brazing time to 1800 s causes depletion of Cu and Ni contents from the braze alloy into SP-700 substrate, so primary Ti 2 Ni/Ti 2 Cu, eutectoid Ti 2 Cu/α-Ti are no longer the dominant phases in the joint. The transformed β-ti primarily consists of acicular α-ti and retained β-ti. It is expected that a reliable SP-700 brazed joint without detrimental Ti 2 Ni/Ti 2 Cu intermetallic compounds is achieved. Acknowledgements The authors gratefully acknowledge the financial support of this research by the National Science Council (NSC), Taiwan, China, under the grant number E REFERENCES [1 ] Y.L. Li, P. He and J.C. Feng: Scripta Mater., 2006, 55, 171. [2 ] A. Shapiro and A. Rabinkin: Welding J., 2003, 82(10), 36. [3 ] M.M. Schwartz: Brazing for the Engineering Technologist, ASM International, Material Park, [4 ] G. Humpston and D.M. Jacobson: Principles of Soldering and Brazing, ASM International, Materials Park, [5 ] A. Rabinkin, H. Liebermann, S. Pounds, T. Taylor, F. Reidinger and S.C. Lui: Scripta Mater., 1991, 25, 399. [6 ] O. Botstein and A. Rabinkin: Mater. Sci. Eng., 1994, A188, 305. [7 ] O. Botstein, A. Schwarzman and A. Rabinkin: Mater. Sci. Eng., 1995, A206, 14. [8 ] C.T. Chang, Y.C. Du, R.K. Shiue and C.S. Chang: Mater. Sci. Eng., 2006, A420, 155. [9 ] C.T. Chang, R.K. Shiue and C.S. Chang: Scripta Mater., 2006, 54, 853. [10] R. Roger, E.W. Collings, and G. Welsch: Materials Properties Handbook: Titanium Alloys, ASM International, Materials Park, [11] T.B. Massalski: Binary Alloy Phase Diagrams, ASM International, Materials Park, [12] W.F. Smith: Structure and Properties of Engineering Alloys, McGraw-Hill Inc., New York, [13] P. Villars, A. Prince and H. Okamoto: Handbook of Ternary Alloy Phase Diagrams, ASM International, Materials Park, [14] K.P. Gupta: Phase Diagrams of Ternary Nickel Alloys, Indian Institute of Metals, Calcutta, India, [15] S. Krishnamurthy and F.H. Froes: Int. Mater. Rev., 1989, 34, 297.