I. INTRODUCTION II. EXPERIMENTAL PROCEDURE
P
HASE transformations in the Cu-Al-Ni alloys have The alloy, Cu-14.2 wt pct Al-7.8 wt pct Ni (Cu-28.0 at. been studied by many workers.[1–13] These studies havepct Al-7.0 at. pct Ni), was prepared in a vacuum induction shown that when an alloy with a chemical composition in furnace by using 99.9 pct copper, 99.9 pct aluminum, and the range of Cu-(14 to 15.1) wt pct Al-(3.1 to 4.3) wt pct 99.9 pct nickel. The melt was chill cast into a 30⫻ 50 ⫻ Ni was solution-heat-treated at a point in the singlephase 200-mm copper mold. After being homogenized at 1000⬚C (disordered body-centered cubic) region and then quenched for 72 hours, the ingot was sectioned into 2.0-mm-thick into room-temperature water or iced-brine, the microstruc- slices. These slices were subsequently solution-heat-treated ture was single D03 phase,[1–4] or D03 phase containing at 1000 ⬚C for 1 hour and then quenched into iced-brine. extremely fine precipitates.[5,6]By using the electron
diffrac-The aging processes were performed at 500⬚C for various tion method, the crystal structure of the extremely fine pre- times in a vacuum-heat-treated furnace.
cipitates was determined to be of the 2H type.[5,6,7] When
Electron microscopy specimens were prepared by means the as-quenched alloy was aged at temperatures ranging of a double-jet electropolisher with an electrolyte of 70 pct from 450⬚C to 550 ⬚C for moderate times and then quenched, methanol and 30 pct nitric acid. The polishing temperature
␥2 particles started to precipitate within the D03matrix at was kept in the range from ⫺40 ⬚C to ⫺20 ⬚C, and the the aging temperature and the remaining D03matrix would current density was kept in the range from 3.0⫻ 104to 6.0 transform to ␥1⬘ martensite during quenching.[7,8] With ⫻ 104 A/m2. Electron microscopy was performed on a increasing the aging time within this temperature range, fine JEOL* JEM-2000FX scanning transmission electron micro-B2 precipitates were found to occur within the well-grown
*JEOL is a trademark of Japan Electron Optics Ltd., Tokyo. ␥2particles and the remaining D03matrix would completely
transform to␣phase (disordered face-centered cubic) or a scope (TEM) operating at 200 kV. This microscope was mixture of (␣⫹) phases at the aging temperature.[7,8,9]
equipped with a Link** ISIS 300 energy-dispersive X-ray Recently, we have performed transmission electron
**Link is a trademark of Oxford Ltd., UK. microscopy examinations on the phase transformations of
the Cu-14.2 wt pct Al-7.8 wt pct Ni alloy. Consequently, we
spectrometer (EDS) for chemical analysis. Quantitative anal-found that in the as-quenched condition, the microstructure
yses of elemental concentrations for Cu, Al, and Ni were of the alloy was D03 phase containing extremely fine L–J
made with the aid of a Cliff-Lorimer Ratio Thin Section precipitates. The L–J phase has an orthorhombic structure
method. with lattice parameters a⫽ 0.413 nm, b ⫽ 0.254 nm, and
c⫽ 0.728 nm, which was first found and identified by the present workers in a Cu2.2Mn0.8Al alloy.[14] This result is
III. RESULTS
different from that reported by other workers.[5,6]In addition, when the as-quenched alloy was aged at 500⬚C for a long
Figure 1(a) is a bright-field (BF) electron micrograph of time, the B2 precipitates could be observed not only within
the as-quenched alloy, exhibiting the presence of the the␥2particles but also within the␣phase. This feature has
extremely fine precipitates with a mottled structure within never been observed by other workers in the Cu-Al-Ni alloys.
the matrix. This feature is similar to that observed by other workers in the Cu-Al-Ni alloys.[5,6]Figures 1(b) through (d) show three different selected-area diffraction patterns C.H. CHEN, Graduate Student, Department of Materials Science and (SADPs) of the as-quenched alloy. It is seen in these SADPs Engineering, and T.F. LIU, Professor and Dean of the College of
Engi-that in addition to the reflection spots corresponding to the neering, are with the National Chiao Tung University, Hsinchu, Taiwan,
D03 phase,[5–7,15–17] the diffraction patterns also consist of
Republic of China. Contact e-mail address: [email protected]
(a) (b)
(c) (d )
(e) ( f )
Fig. 1—Electron micrographs of the as-quenched alloy: (a) BF and (b) through (d ) three SADPs. The zone axes of the D03phase are (b) [100], (c) [110], and (d) [111] (hkl⫽ D03phase, hkl1or2⫽ L–J phase, 1: variant 1, and 2: variant 2). (e) (111) D03and ( f ) L–J DF.
extremely fine precipitates. When compared with our previ- In this figure, it is also seen that a high density of the extremely fine L–J precipitates (dark contrast) was present within the ous studies in the Cu-14.6Al-4.3Ni and Cu2.2Mn0.8Al
alloys,[13,14]it is found that the extra spots with streaks should D0
3domains. Figure 1(f), a (1001) L–J DF electron micro-graph, exhibits the presence of the extremely fine L–J precipi-belong to the L–J phase with two variants, rather than the
2H phase, as reported by other workers.[5,6]Figure 1(e) is a tates. Based on the preceding observations, it is concluded that the as-quenched microstructure of the alloy was D03 (111) D03dark-field (DF) electron micrograph of the same
(e) ( f )
Fig. 2—Electron micrographs of the alloy aged at 500⬚C for 5 min: (a) BF and (b) and (c) two SADPs taken from the precipitate marked as “R” in (a). The zone axes of the␥2particle are (b) [001] and (c) [111]. (d ) An SADP taken from the precipitate marked as “M” in (a). The zone axis of the␥1⬘ martensite is [101]. (e) An SADP taken from a␥2particle and its surrounding␥1⬘ martensite. The zone axes of the␥2particle,␥1⬘ martensite, and internal twin are [001], [101], and [101], respectively (hkl⫽␥2phase, hkl⫽␥1⬘ martensite, and hklT⫽ internal twin). ( f ) (121)␥1⬘ DF.
When the as-quenched alloy was aged at 500 ⬚C for 5 “R” in Figure 2(a), indicate that the precipitate has an ordered body-centered cubic structure with lattice parameter a ⫽ minutes and then quenched, both the D03and L–J phases
could not be observed and the other two kinds of phases 0.872 nm, which is consistent with that of the␥2phase.[1,8] Figure 2(d) is an SADP taken from an area marked as “M” started to occur, as illustrated in Figure 2(a). Figures 2(b)
(a) (b)
(c) (d )
Fig. 3—Electron micrographs of the alloy aged at 500⬚C for 10 min: (a) BF and (b) an SADP. The zone axes of the B2 phase,␥1⬘ martensite, and internal twin are [001], [101], and [101], respectively (hkl⫽ B2 phase, hkl ⫽␥1⬘ martensite, and hklT⫽ internal twin). (c) (121)␥1⬘ and (d ) (100) B2 DF, respectively.
the Cu-Al-Ni alloy,[2]it is found that the positions and streak Figure 3(a) shows a BF electron micrograph of the alloy aged at 500⬚C for 10 minutes and then quenched, revealing behaviors of the reflection spots are the same as those of
the ␥1⬘ martensite with internal twins. The ␥1⬘ martensite the presence of the␥2particles and the␥1⬘ lamellar martens-ite. This feature is similar to that observed in Figure 2(a). has an orthorhombic structure with lattice parameters a⫽
0.440 nm, b ⫽ 0.534 nm, and c ⫽ 0.422 nm.[1]Shown in Therefore, it is likely to deduce that the microstructure of the alloy isothermally held at 500 ⬚C for 10 minutes was Figure 2(e) is an SADP taken from a ␥2 particle and its
surrounding ␥1⬘ martensite, indicating that the orientation still the mixture of (D03 ⫹ ␥2) phases. However, electron diffraction examinations of the lamellar structure indicated relationship between the␥2particle and the␥1⬘ martensite
is (110)␥2//(121)␥1⬘and [001]␥2//[101]␥1⬘. Figure 2(f), a DF that in addition to the reflection spots of the␥1⬘ martensite with internal twins, some faint reflection spots corresponding electron micrograph taken with the reflection spot marked
as “1” in Figure 2(d), clearly reveals the presence of the to the B2 phase could be detected. A typical SADP is shown in Figure 3(b). Figures 3(c) and (d), (121)␥1⬘ and (100) B2
␥1⬘ martensite with a lamellar structure. In the previous
studies,[3,7,10] it was reported that when the Cu-14Al-4Ni DF electron micrographs, clearly exhibit the presence of the ␥1⬘ martensite and the extremely fine B2 precipitates, alloy was aged at 450⬚C for short times and then cooled to
160 ⬚C, the D03 → ␥1⬘ martensitic transformation would respectively. In Figure 3(d), it is also seen that when the alloy was aged at 500⬚C for short times, the B2 precipitates occur during cooling, and the increase of either aluminum
or nickel content would lower the martensitic transformation were only formed within the D03matrix and no evidence of the B2 precipitates could be detected within the␥2particles. temperature. In the present alloy, the aluminum content is
similar to that of the Cu-14Al-4Ni alloy and the nickel Figures 4(a) and (b) are BF and (100) B2 DF electron micro-graphs of the alloy aged at 500 ⬚C for 30 minutes and content is obviously higher. Therefore, it is reasonable to
believe that the␥1⬘ martensite present in Figure 2(a) should then quenched, revealing that the size of the B2 precipitates increased with increasing the aging time and the remaining be formed through a D03→␥1⬘ martensitic transformation
during quenching from the aging temperature, rather than D03matrix transformed to the␣phase completely. Figure 4(c), an SADP taken from an area covering the at the aging temperature.
(c) (c)
Fig. 4—Electron micrographs of the alloy aged at 500⬚C for 30 min: (a) Fig. 5—Electron micrographs of the alloy aged at 500⬚C for 3 h: (a) BF, BF, (b) (100) B2 DF, and (c) an SADP taken from an area covering the (b) (100) B2 DF, and (c) an SADP taken from an area covering the B2 B2 precipitates and their surrounding␣matrix. The zone axes of the B2 precipitates and their surrounding ␥2matrix. The zone axes of the B2 precipitate and␣phase are [011] and [111], respectively (hkl⫽ B2 phase, precipitate and␥2matrix are [001] and [001], respectively (hkl⫽ B2 phase,
and hkl⫽␣phase). and hkl⫽␥2phase).
ship. Transmission electron microscopy examinations B2 precipitates and their surrounding ␣ matrix, indicates
that the orientation relationship between the B2 precipitate revealed that when the alloy was aged at 500 ⬚C for less than 60 minutes, the B2 precipitates were only formed within and the␣phase is (111)B2//(101)␣and [011]B2//[111]␣, which
pro-(a)
(b)
(d ) (c)
Fig. 6—Four typical EDS spectra obtained from (a) as-quenched alloy, and (b) a␥2particle, (c) a B2 precipitate, as well as (d )␣phase in the alloy aged at 500⬚C for 30 min.
longed aging at the same temperature caused the B2 precipi- EDS spectra taken from the as-quenched alloy, a␥2particle, tates started to appear within the ␥2 particles. A typical a B2 precipitate (within the␣phase), as well as the␣phase microstructure is shown in Figure 5(a). Figure 5(b) is a (100) in the alloy aged at 500 ⬚C for 30 minutes, respectively. B2 DF electron micrograph of the same area as in Figure The average weight percentages of the alloying elements 5(a), showing that the size of the B2 precipitates at the examined by analyzing at least ten different EDS spectra of
␣/␥2 interface and within the␣ phase is greater than that each phase are listed in Table I. It is clearly seen in Figure within the ␥2 particles. Figure 5(c), an SADP taken from 6 and Table I that the concentration of aluminum in the␥2 an area covering the B2 precipitates and their surrounding particle is much greater than that in the as-quenched alloy,
␥2matrix, indicates that the orientation relationship between and the reverse result is obtained for the concentration of the B2 precipitate and the␥2matrix is cubic to cubic. This nickel. This result is similar to that examined by other work-result is similar to that observed by other workers in the ers in the Cu-14Al-4Ni alloy.[7,8,9]In their studies, they pro-aged Cu-Al-Ni alloys.[11]
posed that compared with the as-quenched alloy, the ␥2 particle was comparatively richer in aluminum atoms;
there-IV. DISCUSSION fore, the precipitation of the ␥2 particles had caused the
martensitic transformation temperature of the remaining D03 On the basis of the preceding results, it is obvious that
matrix to be higher, which induced the D03→␥1⬘ martensitic when the present alloy was aged at 500⬚C for longer times,
transformation would occur during quenching from the aging the remaining D03matrix would transform to the mixture
temperature. Obviously, their proposition is consistent with of (B2⫹␣) phases. This result is quite different from the
the observation in Figure 2(a). In addition, along with the mixture of (⫹ ␣) phases found by other workers in the
growth of the ␥2particles, the nickel concentration in the Cu-14Al-4Ni alloy aged at 500 ⬚C or 550 ⬚C.[7,9] In order
remaining D03matrix would be increased with increasing to clarify the apparent difference, an STEM-EDS study was
enhance the formation of the B2 phase. Similarly, it is plausi- of this research by the National Science Council, Republic ble to suggest that in the present alloy, the higher nickel of China, under Grant No. NSC90-2216-E-009-044. He is concentration in the remaining D03matrix should be favor- also grateful to M.H. Lin for typing.
able for the formation of the B2 precipitates, instead of the
 phase observed by other workers in the Cu-14Al-4Ni
REFERENCES alloy.[7,9]Furthermore, since the aluminum and nickel
con-centrations in the B2 precipitates are very high, it is thus 1. M.A. Dvorack, N. Kuwano, S. Polat, H. Chen and C.M. Wayman:
Scripta Metall., 1983, vol. 17, pp. 1333-36.
expected that along with the growth of the B2 precipitates,
2. N. Kuwano and C.M. Wayman: Metall. Trans. A, 1984, vol. 15A, pp. the surrounding D03matrix would be depleted in both
alumi-621-26. num and nickel. The lower concentration of aluminum would
3. N.F. Kennon, D.P. Dunne, and L. Middleton: Metall. Trans. A, 1982, cause the D03matrix to become unstable.[7,19]Consequently, vol. 13A, pp. 551-55.
the remaining D03matrix would transform to the Cu-rich␣ 4. N. Za´rubova´, A. Gemperle, and V. Nova´k: Mater. Sci. Eng. A, 1997, vol. A222, pp. 166-74.
phase.[19] This result is consistent with the observation in
5. K. Otsuka, H. Sakamoto, and K. Shimizu: Trans. JIM, 1979, vol. 20, Figure 4(a).
pp. 244-54.
Finally, one more feature is worthy of note. In the previous 6. K. Otsuka, H. Kubo, and C.M. Wayman: Metall. Trans. A, 1981, vol. study,[8]it was found that the B2 precipitates formed within
12A, pp. 595-605.
7. J. Singh, H. Chen, and C.M. Wayman: Metall. Trans. A, 1986, vol. the␥2particles had a platelike shape. In the present study,
17A, pp. 65-72. however, the B2 precipitates had a granular shape. The
rea-8. J. Singh, H. Chen, and C.M. Wayman: Scripta Metall., 1985, vol. 19, son why the morphology of the B2 precipitates had this
pp. 887-90.
difference is unclear. 9. J. Singh, H. Chen, and C.M. Wayman: Scripta Metall., 1985, vol. 19,
pp. 231-34.
10. V. Agafonov, P. Naudot, A. Dubertret, and B. Dubois: Scripta Metall.,
V. CONCLUSIONS 1988, vol. 22, pp. 489-94.
11. Y.S. Sun, G.W. Lorimer, and N. Ridley: Metall. Trans. A, 1990, vol. The as-quenched microstructure of the Cu-14.2Al-7.8Ni
21A, pp. 575-88.
alloy was D03phase containing extremely fine L–J precipi- 12. J. Tan and T.F. Liu: Scripta Mater., 2000, vol. 43, pp. 1083-88. tates. When the alloy was aged at 500⬚C for a short time 13. J. Tan and T.F. Liu: Mater. Chem. Phys., 2001, vol. 70, pp. 49-53.
14. S.C. Jeng and T.F. Liu: Metall. Mater. Trans. A, 1995, vol. 26A, pp. and then quenched,␥2particles were formed within the D03
1353-65. matrix at the aging temperature and the remaining D03matrix
15. T.F. Liu, G.C. Uen, C.Y. Chao, Y.L. Lin, and C.C. Wu: Metall. Trans. underwent a D03 →␥1⬘ martensitic transformation during A, 1991, vol. 22A, pp. 1407-15.
quenching. With slightly increasing aging time at the same 16. C.C. Wu, J.S. Chou, and T.F. Liu: Metall. Trans. A, 1990, vol. 22A, temperature, extremely fine B2 precipitates started to occur pp. 2265-76.
17. K.C. Chu, S.C. Jeng, and T.F. Liu: Scripta Metall., 1996, vol. 34, pp. within the remaining D03matrix. The size of the B2
precipi-83-87. tates increased with increasing the aging time. Along with
18. T.F. Liu, S.C. Jeng, and C.C. Wu: Metall. Trans. A, 1992, vol. 23A, the growth of the B2 precipitates, the remaining D03matrix pp. 1395-1401.
would transform to the␣ phase. When the alloy was aged 19. X.J. Liu, I. Ohnuma, R. Kainuma, and K. Ishida: J. Alloys Compounds, 1998, vol. 264, pp. 201-08.
