Journal of Alloys and Compounds 372 (2004) 180–186
Two-way shape memory effect of a TiNiHf high
temperature shape memory alloy
X.L. Meng∗ , Y.F. Zheng, W. Cai, L.C. Zhao
School of Materials Science and Engineering, Harbin Institute of Technology, P.O. Box 433, Harbin 150001, PR China
Received 8 July 2003; received in revised form 10 October 2003; accepted 10 October 2003
Abstract
The two-way shape memory effect in a Ti36 Ni49 Hf15 high temperature shape memory alloy (SMA) has been systematically studied by
bending tests. In the TiNiHf alloy, the martensite deformation is an effective method to get two-way shape memory effect even with a small
deformation strain. When the TiNiHf alloy is deformed at a full martensite state, the deformation mechanism is the martensite orientation
accompanied by the dislocation slip. The experimental results indicated that the internal stress field formed by the bending deformation is
in the directions of the preferentially oriented martensite variants formed during the bending deformation. Upon cooling the preferentially
oriented martensite variants form under such an oriented stress field, which should be responsible for the generation of the two-way shape
memory effect. Proper training process and aging benefit the formation of the oriented stress field, resulting in the improvement of the two-way
shape memory effect. When the training strain is 7.1% and the training temperature is room temperature, a two-way shape memory strain of
0.88% is obtained in the Ti36 Ni49 Hf15 alloy. This value is the maximum two-way shape memory strain in the TiNi-based high temperature
SMAs so far. In comparison with the TiNi alloy, the TiNiHf high temperature SMA shows the two-way shape memory effect with the relative
poor stability due to the lower strength of martensite.
© 2003 Elsevier B.V. All rights reserved.
Keywords: TiNiHf alloy; High temperature shape memory alloy; Two-way shape memory effect; Training; Stability
1. Introduction
Two-way shape memory effect means the shape memory
alloys (SMAs) are able to “remember” a geometrical shape
at high temperature (above Af ) and another shape at low
temperature (below Mf ), and during the repeated heating and
cooling the SMAs change its shape between these two shapes
without the help of the external stress. It is well known
that in near-equiatomic TiNi SMAs a two-way shape memory effect can be obtained commonly through certain thermomechanical treatments (so-called training), which mainly
include shape memory training [1], stress-induced martensitic transformation training (pseudoelastic training) [2] and
thermal cycling training under a constant stress [3,4]. These
three methods all involve phase transformations between the
parent phase and martensite. Commonly, an anisotropic dislocation structure is introduced into the parent phase ma∗ Corresponding author. Tel.: +86-451-6412163;
fax: +86-451-6413922.
E-mail address: [email protected] (X.L. Meng).
0925-8388/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2003.10.020
trix through training. This dislocation structure creates an
anisotropic stress field, which benefits the formation of preferentially oriented martensite variants adopting the training
procedure, thus resulting in a macroscopic shape change
during subsequent thermal transformation cycles [5].
TiNiHf alloys are considered to be attractive candidates
for SMAs used at high temperatures due to its high transformation temperatures, excellent workability and lower cost
compared with the other high temperature SMAs [6]. Its
constitutional phases [7], microstructure and substructure
[8,9], transformation behaviors [10], mechanical properties
and shape memory effect [11], and precipitation behaviors
during aging [12–14] have been studied. However, except
our previous works [15], seldom results about its two-way
shape memory effect are reported. Dissimilar to TiNi alloys,
a small two-way shape memory effect in the aged TiNiHf
alloys can be obtained directly by the deformation of the
martensite without training [12,15]. Its detailed mechanism
is not clear so far. The finding of the two-way shape memory
effect in the TiNiHf high temperature SMA plays an important role to accelerate its practical applications. The shape
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
memory properties of TiNiHf alloys cannot be improved by
cold deformation and recovery annealing because it is too
brittle to be cold rolled [16]. In the present study, the effects of martensite deformation, training and aging on the
two-way shape memory effects and their stabilities in the
Ti36 Ni49 Hf15 high temperature SMA have been investigated
by bending tests in detail. In addition, the corresponding
mechanisms have also been discussed.
The Ti36 Ni49 Hf15 alloy was prepared by the consumable
arc-melting in an argon atmosphere using 99.92 wt.% sponge
Ti, 99.95 wt.% Ni plate and 99.90 wt.% Hf shot. The ingot was then remelted twice to ensure the composition homogeneity by a levitation method in an argon atmosphere,
and the melt was poured into a graphite mold of 35 mm in
diameter. After homogenizing at 1223 K for 1.5 h, the ingot was hot-rolled into plates of 2.1 mm thickness. 1 mm ×
0.5 mm × 50 mm specimens cut along the direction parallel
to the rolling surface were solid solution-treated at 1273 K
for 1 h in vacuum and then quenched into water. The phase
transformation temperatures of as-treated specimen were determined by the differential scanning calorimeter (DSC) to
be Mf = 421 K, Ms = 452 K, As = 489 K, Af = 504 K,
as shown in Fig. 1. Ms , Mf , As and Af denote martensitic
transformation start temperature, martensitic transformation
finish temperature, reverse martensitic transformation start
and finish temperatures, respectively. Some specimens were
aged at 973 K for 5, 10, 20, 40, 80, 120 h in vacuum and
then quenched into the silicon cooling oil, respectively.
The measurement of the two-way shape memory effect
was done by bending tests, as illustrated in Fig. 2. One
training cycle includes: (1) the specimen was bent against
the cylindrical rod to the deformation position at the required temperatures, (2) the specimen was unloaded, and
the specimen sprung back to the spring back position, (3)
Mf = 421 K, Ms = 452 K
Af = 504 K, As = 489 K
Heat Flow (a.u.)
Cooling
Ms
As Af
Heating
300
350
400
Heating position
Spring back position
Cooling position
Deformation position
c
d
R
h
d
Fig. 2. Schematic illustration of two-way shape memory effect measurement.
2. Experimental procedure
Mf
181
450
500
550
600
Temperature (K)
Fig. 1. Transformation behavior of solution-treated Ti36 Ni49 Hf15 specimen.
the specimen was heated up to 600 K to recover to the heating position, (4) and then the specimen was cooled down to
the room temperature, corresponding to the cooling position
shown in Fig. 2. The specimens were trained by repeating
the steps from (1) to (4). During the training process, we
define the bending deformation strain in the first cycle as
training strain, and the bending temperature of the martensite as training temperature. The bending deformation strain
is estimated by εd = d/(d + 2R). The deformation strains
in this paper are 2, 3.3, 4.5, 5.7 and 7.1%, respectively. The
shape recovery strain caused by heating after deformation
in the first cycle is calculated by εre = (180◦ − θh )εd /180◦ .
The two-way shape memory strain is measured by value
of εtw = (θc − θh )εd /180◦ . The permanent strain is εp =
θh εd /180◦ . In addition, some specimens that have obtained
two-way shape memory effect are thermal cycled between
300 and 600 K in order to study the stability of the two-way
shape memory effect of the present alloy.
3. Results and discussion
3.1. Two-way shape memory effect induced only by
martensite deformation
In the Ti36 Ni49 Hf15 alloy, the two-way shape memory effect can be obtained easily by the deformation of the martensite. Fig. 3 shows the variation of the strain with the temperature for the Ti36 Ni49 Hf15 specimens deformed to the
different strains. It can be seen that the recovery of the deformation started from the very beginning of the heating
process with increasing temperature. When the specimens
were heated to a certain temperature above Af , the shape
recovery finished. Moreover, the shape recovery finish temperature increases with increasing deformation strain, indicating that deformation leads to the increase of the reverse
martensitic transformation temperatures of the present alloy
[17]. From these curves shown in Fig. 3, it is observed that
the shape recovery finish temperature of the specimen deformed to 7.1% is higher than that of the specimen deformed
to 2% for about 12 K. This implies that the stabilization of
the martensite is induced by deformation.
Fig. 4 shows the recovery strain (εre ) as a function of the
bending deformation strain in the deformed Ti36 Ni49 Hf15
182
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
2.0
5
7.1%
εp
1.5
5.7%
Strain (%)
4.5%
3
εre
Strain (%)
4
εtw
6
3.3%
2
2.0%
1.0
εtw
0.5
1
0
350
0.0
400
450
500
550
600
2
3
4
5
6
7
8
Bending Deformation Strain (%)
Temperature (K)
Fig. 3. Two-way shape memory effect of the Ti36 Ni49 Hf15 alloy with
various deformation strains.
Fig. 5. Effect of bending deformation strain on the two-way shape memory
strain.
alloy. Meanwhile, the variation of the shape recovery ratio
(rsr ) with the bending deformation strain is also plotted in
Fig. 4 for comparison. With increasing the bending deformation strain, the recovery strain increases, while the corresponding shape recovery ratio decreases rapidly. This is
similar to the previous results [18].
The variation of the two-way shape memory strain (εtw )
and the residual permanent strain (εp ) with the bending deformation strain is shown in Fig. 5. Clearly, the two-way
shape memory strain almost increases linearly with increasing deformation strain. It is seen that for the experimental alloy the shape cannot be recovered completely upon
heating even when the deformation strain is only 2%. The
corresponding permanent strain is about 0.194%. It is suggested that when the TiNiHf high temperature SMA is deformed at full martensite state, the slip of dislocations is introduced accompanied by the reorientation of the martensite.
When the deformation strain is 7.1%, the two-way shape
memory strain reaches a maximum of 0.45%. This value is
smaller than that obtained in the TiNi binary alloy by the
constant-stress training [19] or constraint aging [20]. For
the solid solution-treated TiNi alloy, its stress-curve shows
a stress plateau with the corresponding straining of about
7% when the alloy is deformed at full martensite state. The
corresponding martensite deformation mechanism is mainly
the reorientation of the martensite variants. However, for the
Ti36 Ni49 Hf15 alloy, no obvious stress plateau is observed
in the stress–strain curves, which are characterized by the
strong work hardening and continuous yielding. The deformation mechanism of martensite in the TiNiHf alloy is the
reorientation of martensite variants accompanied by the slip
of dislocations [15]. In addition, some dislocations are also
introduced to compensate the mismatch of the preferentially
oriented martensite variants in the neighboring grains. All
the dislocations create an oriented stress field, which induces
the two-way shape memory effect.
It should be noted that for the present alloy, the two-way
shape memory effect could be obtained by even 2% deformation strain as shown in Fig. 3. Although the two-way
shape memory strain is only 0.034%, its corresponding angle change during the bending test is about 3◦ , which can
be precisely detected. For the near-equiatomic TiNi alloy
the reorientation of martensite variants occurs when it is deformed to 2% at full martensite state and the two-way shape
memory effect could not be induced by such a small deformation strain. Whereas for the present alloy the critical
stress for the martensite reorientation is almost equal to that
for the dislocation slip [11]. The dislocations are introduced
easily by the martensite deformation to create an internal
stress field, which induces a two-way shape memory effect.
The larger the deformation strain is, the stronger the stress
field is. This explains that the two-way shape memory strain
increases remarkably with increasing the deformation strain.
6
rsr
5
εre
92
Strain (%)
3
84
2
80
Recovery Ratio (%)
88
4
1
76
0
1
2
3
4
5
6
7
Bending Deformation Strain(%)
3.2. Two-way shape memory effect induced by training
8
Fig. 4. Effect of bending deformation strain on the permanent strain and
the shape recovery ratio.
The effect of the number of training cycles on the two-way
shape memory strain for the specimens with the different
training strains is shown in Fig. 6. It is seen that the two-way
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
1.0
εd=7.1%
Two-way Shape Memory Strain (%)
Two-way Shape Memory Strain (%)
1.0
εd=5.7%
0.8
0.6
εd=4.5%
0.4
εd=3.3%
0.2
0.0
εd=2.0%
0
5
10
15
20
25
Number of Training Cycles (N)
183
Td=293K
Td=423K
0.6
0.4
0.2
0.0
30
Td=353K
0.8
With 7.1% training deformation
0
5
10
15
20
25
30
Numberof Training Cycles (N)
Fig. 6. Effect of the number of training cycles on the two-way shape
memory strain of the Ti36 Ni49 Hf15 alloy with different training strains.
Fig. 8. Effect of the number of training cycles on the two-way shape memory strain for the Ti36 Ni49 Hf15 alloy with different training temperatures.
shape memory strain increases rapidly in the first 5–8 training cycles and then almost does not change with further
increasing the cycle number. This result indicates that at
the onset of training procedure the dislocations are introduced continuously to create an oriented stress field, which
is gradually strengthened. Further increasing the number of
training cycles, the stress field created by the beginning several cycles would suppress the introduction of more dislocations during the subsequent training processes. Thus, the
two-way shape memory strain becomes stable rapidly after
the beginning several training cycles. Fig. 7 shows the relation curve between the two-way shape memory strain and
the training strain for the Ti36 Ni49 Hf15 alloy after 30 training cycles. Compared Fig. 7 and Fig. 5, it is found that for
the Ti36 Ni49 Hf15 alloy specimens with the same deformation strain, the two-way shape memory strain obtained by
30 training cycles is about twice more than that obtained
only by martensite deformation. When the training strain is
7.1%, the two-way shape memory strain reaches 0.88% after 30 training cycles. Although the value is very poor in
comparison with that in TiNi binary alloys, it is the maximum two-way shape memory strain obtained in the TiNiHf
high temperature SMAs so far.
During the training process, the deformation temperature
of martensite also affects the two-way shape memory effect seriously. Fig. 8 shows the number of the training cycles dependence of the two-way shape memory strain of the
Ti36 Ni49 Hf15 alloy specimens with the variant training temperatures when the training strain is 7.1%. The shape of the
curves is similar to that shown in Fig. 6. It is seen that he
two-way shape memory strain obtained by 30 training cycles
decreases when the training temperature increases. Fig. 9
depicts the effect of the training temperature on the two-way
shape memory strain in the Ti36 Ni49 Hf15 alloy with the training strain of 7.1%. When the training temperature is below
457 K (near Ms ), the two-way shape memory strain obtained
in the first training cycle almost keeps constant, irrespective of the change of the deformation temperature. While
when the training temperature is over 457 K, the two-way
shape memory strain in the first training cycle drops down
with the training temperature increasing and reaches zero
1.0
After 30 training cycles
Two-way Shape Memory Strain (%)
Two-way Shape Memory Strain (%)
1.0
0.8
0.6
0.4
0.2
0.0
1
2
3
4
5
6
Training Strain (%)
7
8
Fig. 7. Effect of training strain on the two-way shape memory strain.
after 30 training cycles
in the first cycle
0.8
0.6
0.4
0.2
0.0
300
400
500
600
Deformation Temperature (K)
Fig. 9. Effect of training temperature on the two-way shape memory strain.
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
at about 600 K. For the Ti36 Ni49 Hf15 alloy after 30 training
cycles, the two-way shape memory strain decreases continuously when the training temperature increases from 293 to
630 K. This indicates that the serious dislocation slip is accumulated gradually by each training cycle when the training temperature is high enough. The critical stresses for both
the dislocation slip and the martensite reorientation of the
Ti36 Ni49 Hf15 alloy decrease when the deformation temperature increases [11]. It is suggested that the dislocations slip
is very easy to be introduced when the TiNiHf alloy is deformed at high temperature (near Ms ). The higher the training temperature is, the larger the amount of dislocations introduced during training is. The dislocations introduced with
a higher training temperature would damage the repeated
movement of the boundaries between the parent phase and
the martensite to some extent. Therefore, its two-way shape
memory strain decreases continuously with increasing the
training temperature. A better two-way shape memory effect
of Ti36 Ni49 Hf15 alloy must be obtained when the training
temperature is far lower than Ms .
3.3. Effect of aging on two-way shape memory effect
In the present study, we try to get the better two-way
shape memory effect in the aged Ti36 Ni49 Hf15 alloy. The
two-way shape memory effect is obtained directly by deformation at room temperature without training in the aged
Ti36 Ni49 Hf15 alloy. Fig. 10 shows the effect of aging time on
the two-way shape memory strain of the Ti36 Ni49 Hf15 alloy
aged at 973 K with a bending deformation strain of 4.5%. It
is found that for the Ti36 Ni49 Hf15 alloy aged at 973 K for 5 h,
its two-way shape memory strain is about 0.44%, which is
about twice than that in the solid solution-treated specimen
with the same bending deformation strain as shown in Fig. 3.
Further increasing the aging time the two-way shape memory strain decreases. Han et al reported that fine (Ti, Hf)Ni
particles precipitates in the Ti36.5 Ni48.5 Hf15 alloy aged at
873 K for 150 h [13]. However, in our previous TEM obser-
Two-way Shape Memory Strain (%)
0.5
Aging temperature: 973K
4.5% bending deformation strain
0.4
0.3
0.2
0.1
0.0
0
20
40
60
80
100
120
Aging Time (h)
Fig. 10. Effect of aging time on the two-way shape memory strain of
Ti36 Ni49 Hf15 alloy aged at 973 K with a deformation strain of 4.5%.
0.5
Two-way Shape Memory Strain (%)
184
without training
after 30 training cycles
aged at 973K for 5 hours
without training
0.4
0.3
0.2
0
5
10
15
20
25
Number of Cycles (N)
30
Fig. 11. Stability of the two-way shape memory effect for the Ti36 Ni49 Hf15
alloy.
vations, no new type second particles were observed in the
Ti36 Ni49 Hf15 alloy aged at 973 K but (Ti, Hf)2 Ni particles
[12]. With further increasing the aging time, the (Ti, Hf)2 Ni
particles grow up and coarsen continuously [12], which deteriorates the two-way shape memory effect.
3.4. Stability of two-way shape memory effect
Thermal stability of the two-way shape memory effect of
the TiNiHf high temperature SMA is an important issue on
its way to practical applications because under many conditions TiNiHf alloy will be subjected to many cycles by
repeated heating and cooling. Fig. 11 compared the stabilities of three two-way shape memory effects obtained by
only martensite deformation, 30 training cycles and aging +
martensite deformation in the Ti36 Ni49 Hf15 alloy with the
deformation strain of 4.5%, respectively. It can be seen that
the two-way shape memory strain decreases rapidly in the
first several thermal cycles and then becomes stable with
further increasing the number of thermal cycles. According to the results above, it can be deduced that the dislocations introduced by the thermal cycling relax the stress field
formed by the martensite deformation or the training process, leading to the difficulty to form the preferentially oriented martensite variants and the decrease of the two-way
shape memory strain. After the initial several thermal cycles, the dislocations introduced by the thermal cycles interact with the oriented stress field formed by the deformation to create a stable dislocation configuration resulting
in the stabilization of the two-way shape memory effect.
In addition, it is observed that after 30 thermal cycles the
two-way shape memory strain obtained by 30 training cycles is slightly larger than that obtained by the other two
methods, indicating that the stress filed induced by training
is more stable. Hence, in the Ti36 Ni49 Hf15 alloy a relative
large, stable two-way shape memory effect can be got by
proper training process.
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
185
3.5. Mechanism of two-way shape memory effect in TiNiHf
alloy
Two-way shape memory effect is not the nature of the
TiNi-based alloy. There are two options on the mechanisms
of the two-way shape memory effect until now. One is that
the two-way shape memory effect is derived from the residual martensite variants in the parent phase after training
[21]. Those residual martensite variants do not transform to
parent phase even when the temperature is over Af . Upon
cooling, the residual martensite variants grow and affect the
other variants to form the preferentially oriented martensite variants resulting in the two-way shape memory effect.
However, increasing the heating temperature far above Af
to make the residual martensite variants transform to parent
phase completely, the two-way shape memory effect does
not vanish indicating that the residual martensite variants
is not the main reason for the two-way shape memory effect. The other is that two-way shape memory effect is due
to the effect of dislocations introduced by training [22]. It
is analyzed that the energy of the dislocations is lowest in
the martensite variants after training than that in the other
type variants. So upon cooling, the martensite variants with
the same orientation with the variants after training would
form preferentially. Thus the two-way shape memory effect
is observed.
In the TiNiHf high temperature SMAs the two-way shape
memory effect mainly arises from the stress field created
by the dislocations introduced by deformation or training
process. This stress field is created in the directions of the
preferentially oriented martensite variants formed during
the deformation or training process. Thus the preferentially
oriented martensite variants would form upon cooling and
then recover upon heating. On the one hand, the dislocations introduced by martensite deformation/training process
contribute to the formation of the oriented stress field. On
the other hand, those dislocations prohibit the movement
of the boundaries between the parent phase and martensite
to some extent. For example, for the Ti36 Ni49 Hf15 alloy
specimen with a high training temperature, the two-way
shape memory strain decrease due to the serious dislocation
slip. However, as a kind of high temperature SMA, TiNiHf
ternary alloy has its own two characteristics about the
two-way shape memory effect compared with TiNi alloys:
(1) Two-way shape memory effect is easy to induce in
the Ti36 Ni49 Hf15 alloy, but its value is small. As mentioned above, it can be got even when the deformation
strain is only 2%. Fig. 12 shows the schematic diagram of the stress–strain curves of TiNiHf and TiNi
SMAs at full martensite state. For the TiNi alloy, when
it is deformed to a deformation strain less than about
7% (within the plateau stage shown in Fig. 12), the
corresponding deformation mechanism is the reorientation of the martensite variants. Upon heating, the
deformation almost completely recovers. Whereas for
II
II
TiNiHf alloy
Stress
I
TiNi alloy
Both alloys at full martensite state
Strain
Fig. 12. Schematic diagram of stress–strain curves for TiNi alloy and
TiNiHf alloy.
the TiNiHf alloy, its stress–strain curve does not show
a stress plateau and can be divided into three stages
corresponding to different deformation mechanisms,
as indicated in Fig. 12: the initial elastic deformation
of thermal martensite (stage I), the reorientation of
the martensite variants with dislocation slip (stage II),
plastic deformation of martensite (stage III). It is quite
different from that for TiNi alloy that in the stage II the
reorientation of martensite and dislocation slip coexist
for the TiNiHf alloy. This can be supported by that a
permanent deformation of 0.194% was measured even
in the specimen with deformation of a small deformation strain of 2% shown in Fig. 4. Therefore, the stress
field easily induces the two-way shape memory effect.
However, two-way shape memory strain in the experimental alloy is small compared with that obtained in
the TiNi alloy by the conventional training methods.
The maximum two-way shape memory strain obtained
in the present work is only 0.88%, while with the same
training method about 3% two-way shape memory
strain can be obtained in the aged Ti–51.0 at.% Ni alloy
aged at 773 K for 1 h [23]. It is suggested that the dislocation slip also damages the two-way shape memory
effect to some extent due to its obstruction to the movement of the boundaries between the parent phase and
martensite. The increase in the two-way shape memory
strain with increasing the bending deformation strain is
closely related to the extent of the dislocation slip.
(2) The stabilization of the two-way shape memory effect
of the Ti36 Ni49 Hf15 alloy is worse than that of the TiNi
alloys. It is shown in Fig. 11 that the two-way shape
memory strain decreases to 0.2% after 30 thermal cycles, corresponding to the 46.5% of the initial value before the thermal cycling. However, in the Ti–51.0 at.%
Ni alloy aged at 773 K for 1 h, its two-way shape memory strain obtained by the shape memory training with
8% training strain decreases slightly from 2.86 to 2.73%
after 100 thermal cycles [23]. The poor stability of the
two-way shape memory effect in the Ti36 Ni49 Hf15 alloy
is attributed to the introduction of the dislocations during
the thermal cycling that relax the original oriented stress
186
X.L. Meng et al. / Journal of Alloys and Compounds 372 (2004) 180–186
field. TEM observation has found that the substructure
of the martensite of Ti36 Ni49 Hf15 alloy is mainly (0 0 1)
compound twins [9]. In accordance to the theory of Kudoh et al. [24], the slip of a/2 on (0 0 1) is involved during
the formation of the (0 0 1) compound twin. As a result,
it is understandable that a lot of dislocations are introduced during the repeated phase transformations. These
dislocations destroy the original dislocation configuration created by martensite deformation or training process and deteriorate the two-way shape memory effect of
the Ti36 Ni49 Hf15 alloy. According to the above discussion, the easy introduction of the dislocation slip should
be responsible for the easy creation and the instability of
the two-way shape memory effect in the Ti36 Ni49 Hf15
alloy. The experimental results indicate that the internal stress field is created by the dislocation slip is
directional. Hence the preferentially oriented martensite variants are formed in the directions of the internal
stress field upon cooling. In the subsequent thermal
transformation cycles, the newly generated dislocations
change the oriented stress field leading to the decrease
of the two-way shape memory effect. So any methods
that can prohibit the slip of the dislocations in the matrix
would benefit the increase of the two-way shape memory effect and the improvement of its stability. Adding
the fourth element (such as Cu [25], Re [26], etc.) into
the TiNiHf alloy, proper heat-mechanical treatment and
aging maybe the other effective methods to achieve
this, which study is being undergone by the present
authors.
The stability of the two-way shape memory effect in the
Ti36 Ni49 Hf15 high temperature SMA is poor due to its lower
strength of martensite. During the thermal phase transformation cycles, the dislocations are easy to be introduced and
damage the orientated stress field.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
4. Conclusions
[17]
The martensite deformation is an effective method to
introduce a two-way shape memory effect in the TiNiHf
alloy. The two-way shape memory strain increases with
increasing deformation strain. When the TiNiHf alloy is
deformed at full martensite state, the martensite reorientation is accompanied by the dislocation slip. The oriented
stress field formed by the dislocation slip is considered to
be responsible for the generation of two-way shape memory
effect.
Training process and aging increase the two-way shape
memory effect of the Ti36 Ni49 Hf15 high temperature SMA.
This is attributed to that the training and aging are beneficial
to the formation of the oriented stress field. The two-way
shape memory strain decrease continuously when increasing
the deformation temperature during training.
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
J. Perkins, R.O. Sponholz, Metall. Trans. A15 (1984) 313.
L. Contardo, G. Guenin, Acta Metall. Mater. 38 (1990) 1267.
Y. Liu, P.G. McCormick, Acta Metall. Mater. 38 (1990) 1321.
L.M. Wang, Y.F. Zheng, W. Cai, X.L. Meng, L.C. Zhao, J. Mater.
Sci. Technol. 16 (2000).
Y. Liu, Y. Liu, J. Van Humbeeck, Acta Mater. 47 (1999) 199.
S.M. Tuominen, in: Proceedings of the First International Conference
on Shape Memory, Superelastic Technologies, CA, USA, 1994, p.
49.
P. Olier, J.C. Brachet, J.L. Bechade, C. Foucher, G. Guenin, J. Phys.
IV C8 (1995) 741.
D.R. Angst, P.E. Thoma, M.Y. Kao, J. Phys. IV C8 (1995) 747.
Y.F. Zheng, W. Cai, J.X. Zhang, Y.Q. Wang, L.C. Zhao, H.Q. Ye,
Mater. Lett. 36 (1998) 142.
Y.F. Zheng, L.C. Zhao, H.Q. Ye, Scripta Mater. 38 (1998) 1249.
Y.Q. Wang, Y.F. Zheng, W. Cai, L.C. Zhao, Scripta Mater. 400 (1999)
1327.
X.L. Meng, Y.F. Zheng, Z. Wang, L.C. Zhao, Scripta Mater. 42
(2000) 341.
X.D. Han, R. Wang, Z. Zhang, D.Z. Yang, Acta Mater. 46 (1998)
273.
S. Besseghini, E. Villa, A. Tuissi, Mater. Sci. Eng. A273–275 (1999)
390.
X.L. Meng, Y.F. Zheng, Z. Wang, L.C. Zhao, Mater. Lett. 45 (2000)
128.
S.M. Russell, F. Sczerzenie, in: Proceedings of the International Conference on Shape Memory and Superelastic Technologies (SMST-94),
CA, USA, 1994, p. 43.
X.L. Meng, W. Cai, L.C. Zhao, Rare Metal Mater. Eng., 2004, in
press.
Y.Q. Wang, W. Cai, W.H. Zhu, L.C. Zhao, in: Proceedings of the
Second International Conference on Shape Memory and Superelastic
Technologies, Pacific Grove, CA, USA, 1997, p. 77.
Y. Liu, P.G. Mccormick, Factors influencing the development of
two-way shape memory in NiTi, Acta Metall. Mater. 38 (1990) 1321.
S.K. Wu, H.C. Lin, T.S. Chou, Acta Matell. Mater. 38 (1990) 95.
J. Perkins, Metall. Trans. A 15a (1984) 313.
R. Stalmans, J. Van Humbeeck, L. Delaey, in: Proceedings
of the International Conference on Martensitic Transformations
(ICOMAT-92), Monterey, USA, 1992, p. 1065.
L.M. Wang, X.L. Meng, W. Cai, Z. Wang, L.C. Zhao, J. Mater. Sci.
Technol. 8 (2000) 73.
Y. Kudoh, M. Tokonami, S. Miyazaki, K. Otsuka, Acta Metall. 33
(1985) 2049.
X.L. Liang, Y. Cheng, H.M. Shen, Z.F. Zhang, Y.M. Wang, Solid
State Commun. 119 (2001) 381.
F. Dalle, A. Pasko, P. Vermaut, V. Kolomytsev, P. Ochin, R. Portier,
Scripta Mater. 40 (2000) 331.