Martensitic transformation of quaternary
Ti
50.5ÿX
Ni
49.5
Zr
X/2
Hf
X/2
(X = 0±20 at.%)
shape memory alloys
S.F. Hsieh
a, S.K. Wu
b,*
aDepartment of Mold and Die Engineering, National Kaohsiung Institute of Science and Technology, Kaohsiung 800, Taiwan bInstitute of Materials Science and Engineering, National Taiwan University, Taipei 106, Taiwan
Received 9 February 2000; accepted 10 March 2000
Abstract
Martensitic transformation of Ti50.5ÿXNi49.5ZrX/2HfX/2quaternary alloys (X = 0±20 at.%) is studied by different
thermo-mechanical treatments. These alloys have one-stage B2$B190transformation and exhibit 80% shape
memory recovery. Their DSC forward transformation peak M* can be raised from 50°C to 323°C with transformation hysteresis being slightly larger than that of Ti50.5ÿXNi49.5ZrXalloys. In the early 10 cycles, thermal
cycling can depress the M* temperature more significantly in Ti35.5Ni49.5Zr7.5Hf7.5than in Ti35.5Ni49.5Zr15due to
the former alloy having higher hardness in the matrix. Martensite stabilization can be induced by cold rolling at room temperature. The strengthening effects of cold rolling and thermal cycling on Ms temperature are found to follow Ms = ToÿKDsy, in which K values are related to the as-annealed hardness of these alloys.
Ti30.5Ni49.5Zr10Hf10 alloy, aged in martensite phase can cause the phenomenon of thermal-induced martensite
stabilization. D 2000 Elsevier Science Inc. All rights reserved.
Keywords: Martensitic transformation; Quaternary alloys; Shape memory recovery
1. Introduction
Among many shape memory alloys (SMAs), TiNi-based alloys are the most popular due to their superior properties in the shape memory effect (SME) and pseudoelasticity (PE) [1,2]. However, the addi-tion of a third element has a substantial effect on phase transformation behaviors in TiNi alloys. The Ms temperature decreases monotonically following substitution for Ni with Fe, Al, and Co elements [3±
5] but increases remarkably following substitution of Ni with Au, Pd, and Pt in amounts not less than 15 to 20 at.% [6±8], in which they are called ``high-temperature SMAs''. However, the high cost of pre-cious metals limits the practical applications of these high-temperature SMAs. For this reason, other low-cost TiNiX SMAs need to be investigated. Among them, the most significant candidates are TiNiZr and TiNiHf alloys with Zr and Hf being used to replace Ti in these SMAs.
Three phases (Ti,Zr)Ni, (Ti,Zr)2Ni and the l1
phase are observed in Ti-rich Ti53ÿXNi47ZrX alloys
with the Zr content in the range 5±20 at.% at room temperature. Here, the l1phase is a TiNiZr ternary
solid solution and the (Ti,Zr)Ni phase can exhibit the
1044-5803/00/$ ± see front matter D 2000 Elsevier Science Inc. All rights reserved. PII: S1044-5803(00)00068-1
* Corresponding author. Tel.: +886-2-2363-7846; fax: +886-2-2363-4562.
B2$B190one-stage martensitic transformation with
Ms temperature higher than 100°C [9]. Meisner and Sivokha [10] reported that (Ti,Zr)2Ni7, (Ti,Zr)7Ni10,
and NiZr phases can be observed in the Ni-rich Ti50ÿXNi50ZrX alloys with Zr content in the range
30±50 at.% at room temperature. The lattice para-meters of Ti50.5ÿXNi49.5ZrXalloys could be changed
by Zr content in the range 5±20 at.% [11]. Mulder et al. [12] reported that the decrease of transformation temperature in thermally-cycled Ti31.5Ni48.5Zr20alloy
is affected by the precipitates.
The transformation behavior and shape memory characteristics of Ti-rich TiNiZr alloys have been reported under various thermo-mechanical treat-ments, such as thermal cycling, aging, and cold rolling [13]. However, Ti50.5ÿXNi49.5ZrX alloys with
Ti being partially replaced by Hf, are seldom reported to have their transformation behavior and shape memory characteristics affected by these treatments. The aim of the present work is to investigate sys-tematically the general characteristics of Ti-rich TiNiZrHf quaternary SMAs. The transformation be-havior and shape recovery of these alloys will be compared with those of Ti-rich TiNiZr ternary ones. The effects of aging, cold rolling, and thermal cycling on these alloys will also be discussed in this study.
2. Experimental procedure
A conventional tungsten arc melting technique was employed to prepare Ti50.5ÿXNi49.5ZrX/2HfX/2
(A alloys) alloys with X = 0±20 at.%. These alloys' compositions are based on Ti50.5ÿXNi49.5ZrX alloys
(B alloys) which have been intensively studied in our previous work [13]. Titanium (purity 99.7 wt.%), nickel (purity 99.9 wt.%), zirconium (purity 99.8 wt.%) and hafnium (purity, 99.2 wt.%), totaling about 100 g, were melted and remelted at least six times in an argon atmosphere. A pure titanium button was also melted and used as a getter. The mass loss during melting was negligibly small. The as-melted buttons were homogenized at 950°C for 72 h and then
Fig. 1. DSC curves of as-annealed Ti50.5ÿXNi49.5ZrX/2HfX/2
alloys (X = 0±20 at.%). M* and A* are peak temperatures of forward and reverse martensitic transformation, respectively.
Fig. 2. Transformation temperatures of A* and M* vs. X content for Ti50.5ÿXNi49.5ZrX/2HfX/2 alloys.
Fig. 3. Martensite thermal hysteresis vs. X content for the as-annealed Ti50.5ÿXNi49.5ZrX/2HfX/2alloys (X = 0±20 at.%).
quenched in water. The homogenized buttons were cut into several plates with a low speed diamond saw, and then annealed at 900°C for 2 h and quenched in water. After the annealing treatment, three experi-mental procedures were conducted. First, some plates were sealed in evacuated quartz tubes, aged at 285°C for 2 days to 50 days and then quenched in water. Second, some plates were cold-rolled at room tem-perature to 5%, 10%, 15% and 25% reduction in their thicknesses. Third, other plates were subjected to thermal cycling N times from 0°C to 300°C with N = 1±100 cycles. Specimens for DSC measure-ments, hardness tests, shape recovery tests and micro-structural observations were carefully cut from plates treated by the above three procedures. DSC measure-ments were made with a Dupont 9990 thermal analyzer equipped with a quantitative scanning sys-tem 910 DSC cell for controlled heating and cooling runs on samples encapsulated in an aluminum pan.
Fig. 4. Hardness vs. X content for the as-annealed Ti50.5ÿXNi49.5ZrX/2HfX/2 and Ti50.5ÿXNi49.5ZrX alloys (X =
0±20 at.%).
Fig. 5. (a) TEM bright-field image of as-annealed Ti30.5Ni49.5Zr10Hf10alloy. (b)±(d) Selected area diffraction patterns (SADPs)
The running temperature range was from 0°C to 400°C with a heating and cooling rate of 10°C/min. Specimens for the hardness test were first mechani-cally polished and then subjected to measurement in a microVickers hardness tester with a 500 g load at room temperature. For each specimen, the average hardness value was taken from at least five test readings. The microstructural observations were made by transmission electron microscopy (TEM) with a JOEL-100CXII microscope equipped with a conventional double-tilting stage. The shape recovery measurement was performed as described in an ear-lier report [14]. The chemical composition analysis for each phase was performed using a JOEL JXA-8600SX electron probe microanalyzer (EPMA) equipped with a WDX analysis system.
3. Experimental results and discussion
3.1. Transformation behavior in Ti50.5ÿXNi49.5ZrX/2
HfX/2alloys
Fig. 1 shows the experimental results of DSC measurements for homogenized Ti50.5ÿXNi49.5ZrX/2
HfX/2 alloys with X = 0±20 at.% in both forward
and reverse transformations, respectively. The peaks M* and A* (including Ms, Mf, As and Af points) shown in Fig. 1 are associated with the one-stage martensite transformation of B2$B190. The
transfor-mation peaks M* and A* vs. X of Fig. 1 are plotted in Fig. 2. From Fig. 2, the M* temperature increases from 50°C to 323°C with increasing X content. Therefore, based on the results of Fig. 2, a quaternary TiNiZrHf SMA with a desired transformation tem-perature can be obtained by carefully controlling its corresponding X content. For the same X, transforma-tion peak temperatures of A alloys are higher than those of B alloys. Based on the report of Han et al. [15], phase transformation temperatures of Ti
36.5-Ni48.5Hf15 alloy increase largely by using Hf to
substitute Ti. These temperatures are also higher than those of Ti36.5Ni48.5Zr15alloy. Similar characteristic
can be seen in Fig. 2 in which A alloys have Ti being substituted partially by Hf but B alloys do not.
Transformation thermal hysteresis (A*ÿM*) of A and B alloys decreases with increasing X con-tent, as shown in Fig. 3. It is reported that the value of (A*ÿM*) essentially depends on the degree of structural change between parent and martensite in NiTi-based alloys [16,17]. The Ti50Ni50 SMA, NITINOL, has a transformation
hysteresis of about 30 K and a monoclinic marten-site structure. The hysteresis reduction is realized in Ti50Ni50ÿXCuXSMAs, in which the copper addition
changes the martensite structure to an orthorhombic
one and reduces the hysteresis to 10±15 K. Sub-sequently, the R-phase transformation is identified in the binary TiNi alloys under the combination of cold working and heat treatment and its hysteresis is further reduced to 2 K. By contrast, a wide hysteresis up to 100 K is also achieved by a
Fig. 6. Backscattering electron images (BEIs) of as-annealed Ti50.5ÿXNi49.5ZrX/2HfX/2alloys with (a) X = 10, (b) X = 15,
combination of a dispersion of fine niobium parti-cles and cold working. From Fig. 3, the hysteresis effect of A alloys is larger than that of B alloys. We propose that the added Hf in Ti-rich TiNiZrHf alloys may not only increase the transformation hysteresis, but also strengthen the matrix due to solid solution hardening, as shown in Fig. 4.
Fig. 5a shows the TEM bright-field image of martensite in annealed Ti30.5Ni49.5Zr10Hf10 alloy.
Fig. 5b ± d shows the SADPs of Fig. 5a with [100]M, [010]Mand [110]M zones, respectively. Han
et al. [15] found that Ti36.5Ni48.5Hf15 alloy is a
monoclinic B190structure with a = 0.293 nm, b =
0.411 nm, c = 0.473 nm and b = 100.4°. The SADPs of Fig. 5 coincide with the results of Han et al. [15]. Therefore, the structure of martensite in Ti
30.5-Ni49.5Zr10Hf10 alloy is still monoclinic. From Figs.
1 and 5, one can find that the transformation sequence of martensite in Ti-rich TiNiZrHf alloys is the B2$B190one-stage transformation.
Fig. 6a±c shows the EPMA BEIs of 900°C annealed Ti50.5ÿXNi49.5ZrX/2HfX/2 alloys with X =
10, 15, 20 at.%, respectively. The intensity of the backscattered electron image is proportional to the atomic number of the individual element in the illuminated area. A great number of second phase particles are found around the grain boundaries of the matrix. The chemical compositions of the matrix and second phase particles by EPMA ana-lysis are shown in Table 1. The ratio (Ti + Zr + Hf)/Ni of the matrix and that of the second phase particles are also shown in Table 1. According to our previous work [9], the results of Table 1 indicate that the matrix in Fig. 6 is the (Ti,Zr,Hf)Ni phase (Zr and Hf atoms are dissolved in the TiNi phase) and the black particles are the l1phase. The
volume fraction of second phase particles occupies about 4% for these alloys.
Fig. 7 shows the shape recovery of A and B alloys. Despite the existence of many second phase particles, these alloys still exhibit good shape recovery up to about 80%. The shape recovery of A alloys is slightly more than that of B alloys after the specimens were heated to 450°C. This feature is closely related to Hf in solid-solution in the matrix of Ti-rich TiNiZrHf alloys. It has been reported that the shape recovery of TiNi alloys can be increased by different strengthening/harden-ing processes [18]. Second phase particles do not transform martensitically when the temperature changes. They are characterized by high brittleness and limited plasticity. However, in the matrix, the hardening effect of A alloys is slightly larger than that of B alloys due to Hf solid solution hardening, as shown in Fig. 4. Therefore, the shape recovery
Table 1
Chemical composition of matrix and second phase particles of as-quenched Ti50.5ÿXNi49.5ZX/2HfX/2alloys
with X = 10, 15, and 20 at.% detected by EPMA 950°C 24 h as-quenched
Composition (at.%)
Phases Ti Ni Zr Hf (Ti + Zr + Hf)/Ni ratio
X = 10 M 41.47 48.92 4.68 4.93 1.04 S 53.04 36.97 6.18 3.81 1.7 X = 15 M 36.60 49.09 6.76 7.55 1.03 S 48.27 37.23 7.36 7.14 1.69 X = 20 M 31.24 49.42 9.36 9.98 1.02 S 44.02 37.94 10.32 7.72 1.64
Note. M: matrix; S: second phase particles.
Fig. 7. Shape recovery vs. heating temperature for the as-annealed Ti50.5ÿXNi49.5ZrX/2HfX/2 and Ti50.5ÿXNi49.5ZrX
of A alloys is also slightly more than that of B alloys, as demonstrated in Fig. 7.
3.2. Cold rolling effect on the Ti35.5Ni49.5Zr7.5Hf7.5
alloy
The effects of cold rolling on martensitic transfor-mation of Ti36.5Ni48.5Zr15and Ti35.5Ni49.5Zr15alloys
have been systematically studied previously [13]. The phenomenon of martensite stabilization is observed in the cold-rolled Ti-rich TiNiZr martensite. In the pre-sent study, the Ti35.5Ni49.5Zr7.5Hf7.5 alloy is
plasti-cally deformed by cold rolling at room temperature. Table 2 shows the detailed results of DSC and hardness measurements for various amounts of cold rolling in the Ti35.5Ni49.5Zr7.5Hf7.5alloy in which the
reverse transformation peaks A1* and A2*,
respec-tively indicate the first and second heating cycle for the specimen just after the cold rolling. In Table 2, A1*
temperatures significantly increase, but A2* decrease,
along with increased cold rolling. Besides, the M* temperature is also found to decrease with increased cold rolling. The above phenomenon is regarded as the mechanically-induced martensite stabilization in Ti35.5Ni49.5Zr7.5Hf7.5alloy. The same behavior is also
reported in Ti50Ni50, Ti51Ni49, and Ti35.5Ni49.5Zr15
alloys [13,19]. After the occurrence of the first reverse transformation of B190!B2, the martensite
stabiliza-tion dies out and A2* temperatures are lower than A1*.
In Fig. 8a, DA* (the difference between A1* and A2*)
stands for the degree of martensite stabilization. From Fig. 8a, the DA* of Ti35.5Ni49.5Zr7.5Hf7.5alloy (DA* =
384°C) is larger than that of Ti35.5Ni49.5Zr15 alloy
(DA* = 351°C) [13] for the same 25% cold-rolled specimen. Fig. 8b shows that the increment of hard-ness under the same cold rolling, is about 188 Hv for Ti35.5Ni49.5Zr7.5Hf7.5 alloy and about 176 Hv for
Ti35.5Ni49.5Zr15alloy [13]. The as-annealed hardness
of the former alloy is slightly greater than that of the latter due to Hf dissolved in the matrix. We suggest that the dislocation movement may be hindered more
in the Ti35.5Ni49.5Zr7.5Hf7.5 alloy than in the
Ti35.5Ni49.5 Zr15 alloy and cause the harder Ti
35.5-Ni49.5Zr7.5 Hf7.5 alloy to have a higher martensite
stabilization under the same level of cold rolling, as demonstrated in Fig. 8.
Table 2
DSC peak temperature A1*, M1*, A2* and the hardness of Ti35.5Ni49.5Zr7.5Hf7.5alloy under various thickness reductions of cold
rolling at room temperature Thickness of cold rolling reduction (%) A1* (°C) DHh1(J/g) M1* (°C) DHC(J/g) A2* (°C) DHh2(J/g) Hardness (Hv) 0 238 26.81 195 24.10 226 23.48 308 5 301 19.33 160 18.32 220 14.50 368 10 341 7.47 132 7.27 190 5.64 407 15 413 6.37 103 6.08 165 5.12 432 25 510 4.12 50 3.91 126 3.53 496
Fig. 8. (a) The degree of martensite stabilization (DA*) and (b) the increment of hardness (DHv) vs. degree of cold rolling for the Ti35.5Ni49.5Zr7.5Hf7.5alloy.
3.3. Thermal cycling effects on the Ti35.5Ni49.5Zr7.5Hf7.5alloy
Fig. 9a and b shows peak temperatures M*, A* and hardness Hv, respectively, vs. thermal cycles N for annealed Ti35.5Ni49.5Zr15 and Ti35.5Ni49.5 Zr7.5Hf7.5
alloys. M* and A* temperatures decrease, but the hardness Hv increases with increasing thermal cy-cling. It has been proposed that this feature comes from the influence of dislocations induced by thermal cycling [20]. A* and M* decrease quickly for the first 10 cycles, with the decrement being about 28°C for Fig. 9b and about 25°C for Fig. 9a. The increase in hardness shown in Fig. 9b (DHv = 48) is greater than that of Fig. 9a (DHv = 43) after N = 10 cycles. This indicates that the Ti35.5Ni49.5Zr7.5Hf7.5alloy can
in-duce more dislocations than the Ti35.5Ni49.5Zr15alloy
in the early thermal cycling. We suggest that the volume change during the martensitic transformation can produce a complex stress field at the interfaces of second phase particles and B2/B190 matrix during
thermal cycling. This complex stress field can en-hance the dislocation multiplication, which increases
the hardness of the alloy and depresses its M* tem-perature. In Fig. 9, after 75 cycles, the M* and A* temperatures reach a constant value. This may indi-cate that the quantities of induced dislocations reach a saturated value after 75 cycles in these alloys. 3.4. Strengthening effects of cold rolling and thermal cycling on martensitic transformation temperatures of Ti35.5Ni49.5Zr7.5Hf7.5alloy
Fig. 10a and b shows the relationship between peak temperature M* and hardness Hv for the cold-rolled and thermally-cycled Ti35.5Ni49.5Zr7.5Hf7.5
al-loy. The results of cold-rolled and thermally-cycled Ti35.5Ni49.5Zr15 alloy are also shown in Fig. 10. It
was pointed out that any strengthening mechanism, which impedes the transformation shear can lower the transformation temperatures because the martensitic transformation involves a shear process [21,22]. This feature can be expressed as
Ms Toÿ KDsy 1
where the constant K contains the factors of proportionality between the critical shear stress and the yield stress Dsy, the equilibrium temperature Tois
a function of the chemical composition, and the yield stress Dsyis regarded as proportional to the hardness.
In this study, both cold rolling and thermal cycling do not change the alloy's composition, hence Tois a
constant. In addition, both cold rolling and thermal cycling can strengthen the alloys by inducing dis-locations, and therefore can raise the yield stress Dsy.
As seen from Eq. (1), this feature should cause the M* and A* temperatures to be lowered by the strengthening effect. This prediction is qualitatively consistent with the results of Fig. 10. In Fig. 10, the slope represents the constant K, which is not the same for different strengthening processes. These K values indicate that cold rolling and thermal cycling can provide different strengthening mechanisms and ex-hibit different effects on transformation temperatures. As mentioned above, strengthening processes can introduce dislocations in these alloys. However, dis-locations induced by cold rolling originate from the plastic deformation of martensite, and those induced by thermal cycling come from the thermal stress and transformation shear associated with B2$B190. A
careful examination of Fig. 10 shows that the con-stant K of the Ti35.5Ni49.5Zr7.5Hf7.5 alloy is larger
than that of the Ti35.5Ni49.5Zr15 alloy for the same
strengthening process. We propose that the K value is closely related to the inherent hardness of annealed TiNiZr or TiNiZrX alloys. The higher the annealed hardness, the larger the K value is. For example, the thermally-cycled Ti35.5Ni49.5Zr7.5Hf7.5 alloy has its
annealed hardness at about 308 Hv and its K
Fig. 9. Peak temperatures A* and M* and hardness Hv, vs. number of thermal cycles, N, for (a) Ti35.5Ni49.5Zr15alloy
value is found to be 0.85°C/Hv, which is larger than the thermally-cycled Ti35.5Ni49.5Zr15 alloy
(289 Hv, K = 0.79°C/Hv). This characteristic is also found in cold-rolled alloys, as shown in Fig. 10a. In other words, the depression of Ms(M*) and As(A*) temperatures by the strengthening mechanism is stronger for the alloys having a higher annealed hardness. As mentioned above, A alloys have a higher annealed hardness than B alloys. This feature can explain why the K value of Ti35.5Ni49.5Zr7.5Hf7.5alloy is higher than that of
Ti35.5Ni49.5Zr15 alloy under the same strengthening
process, as demonstrated in Fig. 10.
3.5. Aging effect on the Ti30.5Ni49.5Zr10Hf10alloy
Fig. 11a and b shows the results of DSC mea-surements in 285°C aged Ti30.5Ni49.5Zr10Hf10 alloy
and 200°C aged Ti30.5Ni49.5Zr20 alloy, respectively.
Both alloys are aged in the martensite phase. In Fig.
11, A1* and A2*, respectively indicate the first and
second heating cycle for the specimen just after aging. From Fig. 11, the A1* temperature is seen to
increase with increasing aging time. This feature exhibits the phenomenon of martensite stabilization, the same behavior as reported in Cu-based SMAs [23±26]. Two mechanisms were proposed to ex-plain the martensite stabilization in Cu-based SMAs: (i) reordering in the martensite, where the atomic rearrangement in martensite results in some change of the relative stability between parent and martensite [23,24]; and (ii) pinning or locking the interfaces of martensite/parent and martensite/mar-tensite by aging-induced defects or precipitates [25,26]. Fig. 12a shows the TEM bright-field image of the martensite in a 285°C for 50 days aged Ti30.5Ni49.5Zr10Hf10 specimen. Fig. 12b±d shows
the SADPs of Fig. 12a, in which the foil is parallel to [100]M, [010]M, and [101]M directions,
respec-tively. No extra reflection spot can be observed in
Fig. 11. Peak temperatures A* and M* and hardness vs. aging time for (a) Ti30.5Ni49.5Zr10Hf10 alloy and (b)
Ti35.5Ni49.5Zr15alloy.
Fig. 10. The temperature M* vs. hardness Hv for (a) cold-rolled and (b) thermally-cycled Ti35.5Ni49.5Zr7.5Hf7.5 and
Fig. 12. This feature implies that the mechanism of the reordering in martensite after the martensite stabilization may not occur. It has been reported that the interfaces between parent and martensite or mar-tensite and marmar-tensite plates in the stabilized Ti26.5
Ni48.5Zr25martensite may be pinned by point defects
such as interstitial atoms (H, O, etc.) and quenched in vacancies [13]. We propose that the same phenom-enon of martensite stabilization may also occur in the aged Ti30.5Ni49.5Zr10Hf10alloy.
4. Conclusion
Martensitic transformation of Ti50.5ÿXNi49.5ZrX/2
HfX/2high temperature SMAs have been studied by
DSC measurement, hardness test, and microstructural observation. The important conclusions are as follows. (1) The annealed Ti50.5ÿXNi49.5ZrX/2HfX/2 alloys
undergo a one-stage B2$B190martensitic
transforma-tion in which the transformatransforma-tion peak temperature M*
increases from 50°C to 323°C with increasing X from 0 to 20 at.%. Under the same X, the transformation temperatures and hysteresis of Ti50.5ÿXNi49.5ZrX/2
HfX/2 alloys are both larger than those of Ti50.5ÿX
Ni49.5ZrXones. The former alloys are harder than the
latter due to the solid solution of Hf atoms in the matrix. Many second phase particles are found around the grain boundaries of the matrix and are identified as l1 phase for alloys with X 10 at.%. Despite the
existence of second phase particles, these alloys still exhibit good shape recovery that can reach about 80%. (2) The martensite stabilization can be induced by cold-rolled Ti35.5Ni49.5Zr7.5Hf7.5and Ti35.5Ni49.5Zr15
alloys at room temperature. The hardness increment of the former alloy is larger than that of the latter under the same degree of cold rolling, owing to the former alloy having higher annealed hardness.
(3) A* and M* temperatures decrease and the hardness increases in the first 10 cycles of thermally-cycled Ti35.5Ni49.5Zr7.5Hf7.5and Ti35.5Ni49.5Zr15
al-loys. The decrement of the A* temperature of the
Fig. 12. (a) Bright-field image of martensite in 285°C 50 days aged Ti30.5Ni49.5Zr10Hf10alloy. (b)±(d) SADPs of (a) with (b)
former alloy is larger than that of the latter at the same N due to the former alloy having the harder matrix.
(4) The strengthening effects of cold rolling and thermal cycling on Ms(M*) temperatures of Ti
35.5-Ni49.5Zr7.5Hf7.5and Ti35.5Ni49.5Zr15alloys are found
to follow the equation Ms = ToÿKDsy. Strengthening
processes of cold rolling and thermal cycling have their different K values. The annealed hardness of Ti35.5Ni49.5Zr7.5Hf7.5 alloy is higher than that of
Ti35.5Ni49.5Zr15alloy. This feature causes the former
alloy to have a higher K value than the latter under the same strengthening process.
(5) In the Ti30.5Ni49.5Zr10Hf10alloy, a
thermally-induced martensite stabilization occurs after aging in the martensite phase. This characteristic may be caused from the pinning/locking effect, i.e., the interfaces of martensite/parent or martensite/marten-site are pinned by aging-induced point defects. Acknowledgments
The authors sincerely acknowledge the financial support of this study by the National Science Council (NSC), Republic of China, under the Grant NSC 86-2216-E002-033.
References
[1] Miyazaki S, Kotsuka K, Suzuki Y. Transformation pseudoelasticity and deformation behavior in a Ti ± 50.6% Ni alloy. Scr Metall 1981;15:281±87. [2] Miyazaki S, Imai T, Igo Y, Kotsuka K. Effect of cyclic
deformation on the pseudoelasticity characteristics of Ti±Ni alloys. Metall Trans 1986;17A:115±20. [3] Eckelmeyer KH. The effect of alloying on the shape
memory phenomenon in nitinol. Scr Metall 1976;10: 667±72.
[4] Hwang CM, Wayman CM. Phase transformation in Ti± Ni ± Fe, Ti ± Ni ± Al and TiNi alloys. Scr Metall 1983;17:1345±50.
[5] Hwang CM, Meichle M, Salamon MB, Wayman CM. Transformation behavior of a Ti50Ni47Fe3alloy: III.
Martensitic transformation. Philos Mag 1983;47A: 177±91.
[6] Lindqist PG, Wayman CM. Shape memory and trans-formation behavior of martensitic Ti±Pd±Ni and Ti± Pt±Ni alloys. In: Duering TW, Melton KN, Stockel D, Wayman CM, editors. Engineering Aspects of Shape Memory Alloys. London: Butterworth, 1990. p. 58. [7] Lo YC, Wu SK, Wayman CM. Transformation heat as
a function of ternary Pd additions in Ti50Ni50ÿXPdX
alloys with X = 20±50 at.%. Scr Metall 1990;24: 1571±6.
[8] Wu SK, Wayman CM. Martensitic transformation and the shape memory effect in Ti50Ni10Au40and Ti50Au50
alloys. Metall 1987;20:359±76.
[9] Hsieh SF, Wu SK. Room-temperature phase observed in Ti53ÿXNi47ZrXhigh-temperature shape memory
al-loys. J Alloys Compd 1998;266:276±82.
[10] Meisner L, Sivokha V. Deformation of crystal lattice in the process of martensitic transformation in alloys of Ni50Ti50ÿXZrX. J Phys IV 1995;C8:765±9.
[11] Hsieh SF, Wu SK. A study on lattice parameters of martensite in Ti50.5ÿXNi49.5ZrXshape memory alloys.
J Alloys Compd 1998;270:237±41.
[12] Mulder JH, Mass JH, Beyer J. Martensitic transforma-tions and shape memory effects in Ti±Ni±Zr alloys. Proc. Int. Conf. on Martensitic Transformations, Mon-terey, California (July 20±24, 1992).
[13] Hsieh SF, Wu SK. A study on Ti-rich TiNiZr shape memory alloys. Mater Charact 1998;41:151±62. [14] Lin HC, Wu SK. Strengthening effect on shape
recov-ery characteristic of the equiatomic TiNi alloy. Scr Metall 1992;26:59±61.
[15] Han XD, Zou WH, Wang R, Zhang Z, Yang DZ, Wu KH. The martensite structure and aging precipitates of a TiNiHf high temperature shape memory alloys. J Phys IV 1995;C8:753±8.
[16] Humbeck JV, Stalmans R, Chanadrasekaran M, Delay L. On the stability of shape memory alloys. In: Duer-ing TW, Melton KN, Stockel D, Wayman CM, edi-tors. Engineering Aspects of Shape Memory Alloys. London: Butterworth, 1990. p. 96.
[17] Suzuki Y, Horikawa H. Thermal hysteresis in Ni±Ti and Ni ± Ti± X alloys and their applications. In: Liu CT, Kunsimann H, Otsuka K, Wuttig M, editors. Shape-Memory Materials and Phenomena-Fundamental Aspects and Applications. Pittsburgh, PA: Materials Research Society, 1992. p. 389. [18] Lin HC, Wu SK, Chou TS, Kao HP. The effects of cold
rolling on the martensitic transformation of an equia-tomic TiNi alloy. Acta Metall 1991;39:2069±80. [19] Lin HC, Wu SK, Lin JC. The martensitic
transforma-tion in Ti-rich TiNi shape memory alloys. Mater Chem Phys 1994;37:184±90.
[20] Miyazaki S, Igo Y, Otsuka K. Effect of thermal cycling on the transformation temperatures of TiNi alloys. Acta Metall 1986;34:2045±51.
[21] Cohen M, Machlin ES, Paranjpe VG. Thermody-namics in Physical Metallurgy. Metals Park, OH: ASM, 1950. p. 242.
[22] Hornbogen E. The effect of variables on martensitic transformation temperatures. Acta Metall 1991;33: 595±601.
[23] Scarsbrook G, Stobbs WM. The martensitic transfor-mation behavior and stabilization of rapidly quenched Cu±Zn±Al ribbons. Acta Metall 1987;35:47±56. [24] Arab AA, Chanadrasekaran M, Alers M. Martensite
aging and its stabilization in Cu±Zn±Al shape mem-ory alloys. Scr Metall 1984;18:709±14.
[25] Ahlers M. The stabilization of martensite in Cu±Zn± Al. Proc. Int. Conf. on Martensitic Transformation, Nara, Japan (August 26±30, 1986).
[26] Wu MH, Wayman CM. Isothermal aging of martensite in a Cu±Zn±Al shape memory alloy. Mater Sci Forum 1990;56±58:553±8.
