Content uploaded by Amanda Skalitzky
Author content
All content in this area was uploaded by Amanda Skalitzky on Apr 22, 2021
Content may be subject to copyright.
1 Copyright © 2020 by ASME
Proceedings of the ASME 2020 Conference on Smart Materials, Adaptive Structures and Intelligent Systems
SMASIS2020
September 14-16, 2020, Irvine, California, USA
SMASIS2020-2290
MACHINE DESIGN FOR MULTISCALE NITINOL ANNEALMENT PROCESS AND
END PRODUCT PERFORMANCE ANALYSIS
Amanda Skalitzky Stuart Coats
Auburn University Auburn University
Auburn, Alabama, USA Auburn, Alabama, USA
Ramsis Farag Austin Gurley David Beale
Auburn University Deft Dynamics Auburn University
Auburn, Alabama, USA Birmingham, Alabama, USA Auburn, Alabama, USA
ABSTRACT
The functional properties of Nitinol (NiTi) are set by
composition, production process, and post-production heat
treatment and cold working. Post-production heat treating is
dependent on two main parameters: anneal temperature and
aging time. Most heat-treating processes performed by
researchers generally consist of simple temperature soaks at
specified aging times. However, there are drawbacks to this
method. More complex heat treatments can result in
performance improvements, but they are difficult to implement
and often proprietary to manufacturers and therefore not widely
used by researchers. By designing a Continuously-Fed heat
treatment System (CFS), this work demystifies this complex
heat-treatment process by rapidly heat-treating NiTi wire
samples across a range of annealing temperatures, soak times,
and tensions with little human intervention. This automated
process ensures samples are created in a consistent manner and
results in a much more consistent end-product when compared
to conventional heat-treating methods. Using the CFS, a gamut
of samples with varying annealing temperatures (400-550°C)
and aging times (1-3 minutes) were created with 0.25mm
diameter high-temperature actuator wire initially in the ‘as-
drawn’ condition. Differential Scanning Calorimetry (DSC)
analysis was performed to determine how the transition
temperature(s) change with the various heat-treating parameters
and the mechanical properties of the wire were determined
utilizing a tensile test. The experimental results demonstrate the
benefits of the CFS and are compared to those of a more
conventional heat treatment process. Experimental results show
that high-performance Nitinol actuator behavior can consistently
be achieved using the CFS. Optimal heat treatment processes
can be determined quickly experimentally.
INTRODUCTION
The term Shape Memory Alloy (SMA) refers to an alloy
that undergoes a microstructural phase change that may be
initiated through the application of mechanical stress or by
varying the material temperature [1]. There are many different
types of SMAs, including iron-manganese-silicon, copper-
aluminum-silicon, copper-zinc-aluminum, and Nitinol (nickel-
titanium). Nitinol is a very popular SMA due to a good balance
of cost, stability, and thermos-mechanical performance [2]. It is
composed of nickel and titanium, with a ratio varying tightly
between 49 and 51 atomic percent nickel [3]. Although the cost
of Nitinol is higher than that of some other SMAs, the benefits
of altering the memory-recovery temperature and overall
functional properties through various methods such as the
production processes, alloy variation, and annealing temperature
and aging time supersede the higher cost. Additionally, Nitinol
has a unique ability to undergo large amounts of stress and strain
during thermo-mechanical loading and has the capability to
recover large inelastic strains near 8% before experiencing any
permanent deformation [4].
Nitinol has two different transformation mechanisms,
shape-memory and super-elastic, which occur during a reversible
phase change from an Austenite (B2) to Martensite (B19’)
microstructure. A chang e in te mperat ure r esults i n the phase
transformation responsible for the shape memory property while
a stress-induced transformation is responsible for the super-
elastic property. There is an additional phase known as the R-
phase, which occurs when the material is annealed at a certain
range of temperatures and appears during the cooling cycle,
generally in between the Austenite and Martensite phases [5]. It
2 Copyright © 2020 by ASME
occurs when Nitinol has a high dislocation density (the material
has been cold worked) and/or when there are Ni-rich precipitates
(the material has undergone a heat treatment process). This
phase does not allow for large strain transformations and causes
a two-stage transformation behavior [6].
The temperature where the various phase changes occur
are often referred to as the transition temperatures and can be
seen on the phase diagram in Figure 1. When no external stress
is applied and the Nitinol is heated, it begins to transition from
the Martensite phase into the Austinite phase. This is known as
the Austenite-start-temperature (
𝐴!
). As the material continues
to be heated, the material eventually transforms entirely into
Austenite phase, which is referred to as the Austenite-finish-
temperature (
𝐴"
). While cooling under a stress-free state, the
temperature at which the phase transformation from Austenite
back to Martensite begins is referred to as the Martensite-start-
temperature (
𝑀!
). The temperature at which this phase
transformation is complete is called the Martensite-finish-
temperature (
𝑀
"
). Above a certain temperature known as the
Martensite-deformation-temperature (
𝑀#)
Martensite can no
longer be induced with stress alone and as a result the material is
permanently deformed [7]. The upper and lower temperature
bounds of the R-phase transformation are known as the R-phase-
start-temperature (
𝑅!
) and the R-phase-finish-temperature (
𝑅"
)
[8].
Figure 1: Nitinol phase diagram.
HEAT-TREATMENT EFFECT
Nitinol can be purchased in many different forms such
as-drawn or straight-annealed. As-drawn Nitinol means the
material has been heavily cold-worked by the manufacturer. The
cold work provides creates an internal strain field in the
crystalline structure resulting in the alloy possessing a higher
yield stress and ultimate tensile strength. However, once a heat-
treatment is performed by the user, the alloy becomes annealed.
This lowers the yield stress and ultimate strength when atomic
diffusion that is accelerated by being held at elevated
temperature releases the internal strain field between the crystals
[9]. If the annealing temperature and aging time is too high, it
will cause the material to recrystallize and therefore have a lower
ultimate tensile strength and extremely high yield strains.
Additionally, in the recrystallized state, it should be noted that
although the material has a high strain to failure, the reversible
recovery strain drops well below 8% [10]. Therefore, it is
important for most applications that take advantage of the shape-
memory or super-elastic effects to ensure recrystallization does
not occur.
There are benefits to working with as-drawn wire such
as easier shape-setting and higher strengths as compared to
straight-annealed Nitinol material. If as-drawn Nitinol is
utilized, the heat-treatment process is the most common process
utilized to alter the thermal and mechanical properties of Nitinol.
Heat-treatment refers to both the annealing temperature and
aging time which crucially effect the behavior of the Nitinol [11].
The temperature at which the reversible phase
transformation from Austinite to Martensite initiates can be
altered through annealing at various temperatures, aging times,
and various quenching profiles after heat treatment [12]. As the
annealing temperature increases, the transformation temperature
does as well, which causes an increase in the endothermic and
exothermic peaks of the DSC curve [13].
SYSTEM DESIGN
The Continuously Fed heat-treatment System (CFS)
comprises four main elements: a temperature-controlled furnace,
a tension-controlled wire feeder, a quenching bath, and a
feedback control computer which can be seen in Figure 2.
During operation, untreated wire is pulled through the furnace at
a continuous slow speed under regulated tension. The soak time
is determined by the speed and oven size and the wire is
quenched immediately upon exiting the furnace. A s equenc e o f
desired sample lengths, temperatures, soak times, and tensions
are provided to the control computer and it autonomously
generates these samples and spools them for the operator to
perform post heat-treatment tests.
Figure 2: Diagram of the Continuously fed heat-treatment system
architecture
The furnace was designed using a modified Neycraft
JFF-2000 Fiber Furnace which was selected due to its large size
which allowed for the heat-treatment of longer sections of wire.
Additionally, the systems’ electrical simplicity allowed for
straightforward modification of the internal temperature
3 Copyright © 2020 by ASME
controller. This furnace was then outfitted with a new
temperature controller to provide more precise temperature
control (within 1°C of the desired temperature) than the original
manufacturer equipment which consisted of a knob that
controlled oven duty cycle and a temperature display. The new
temperature controller was developed by replacing the existing
oven electronics with a high voltage relay connected to the
furnace’s internal resistance heater, a K-type thermocouple and
thermocouple amplifier, and a Teensy 3.2 microcontroller
implementing closed-loop PID control. The Teensy 3.2 was
implemented because of its low cost, high processing speed, ease
of programming, and modularity.
In addition to the temperature controller, three NEMA
17 geared stepper motors, paired with DeftStep stepper-servo
drivers (Figure 3), were attached to an external frame to allow
for the continuous feeding of Nitinol wire through the thermal
chamber of the furnace using spools of 3D-printed polymer
attached to the output shaft of the motors. NEMA 17 motors
offer high torque output per cost when used with a planetary
gearbox as in this application. The DeftStep drivers provide an
easy conversion from a traditional stepper motor into a more
useful stepper-servo motor. The drivers also allow for relatively
simple configuration and communication when using a USB-
based microcontroller such as a Teensy and modification of the
stepper driver firmware allowed for direct speed control.
Additionally, these drivers employ current chopping, direct
control of current, which not only increases the bandwidth of the
motor, but also allows for the use of currents twice as high as the
motor specification for short periods of time resulting in even
higher torque output.
Figure 3: Neema 17 stepper motor (left) and DeftStep stepper-servo
driver (right)
The first motor was used to control the speed of the wire
through the furnace. The second was used to control the tension
on the wire throughout the annealing process, as measured with
two load cells attached to the base of this motor. The load cells
utilized were straight bar load cells due to the desired mounting
of the motor and the direction of the force due to the wire tension.
The third motor was used for wire ‘take-up’ onto a final spool for
storage while the annealing process was in progress. Upon
exiting the furnace, the wire was immediately quenched in room-
temperature water using a constantly flowing overhead
quenching shower providing continuous quenching of the wire
after the heat-treatment process.
The annealing process was then specified by the
temperature of the furnace and the time that the wire was in the
thermal chamber. Once manufactured, the CFS allowed for
continuous heat treating of long wire lengths. The entire CFS
can be seen in Figure 4.
Figure 4: Render of the Continuously Fed Nitinol Heat Treating
System
EXPERIMENTAL SETUP
The heat-treating process developed is referred to as
Continuously Fed (CF) heat-treating. This process involves
heating the furnace up to the desired annealing temperature, then
commanding the motor speed and tension. The motor speed was
determined using the desired aging time and the tension was
controlled to 1N for all heat treatments presented in this work.
This allowed for thermal expansion to be neglected during this
process. The tension of 1N was chosen for the because during
the initial heat-treating runs, it was determined that tensions
above 5N led to the wire breaking in the thermal chamber of the
furnace. Due to the novelty of this heat-treating scheme it was
decided that 1N tension would be used and further research could
explore higher tensions. As the wire is fed through the oven, it
is heat-treated continuously and then immediately quenched with
room temperature water to complete the heat treatment process.
During this process the wire was kept straight so the desired
trained shape could be achieved. This system allows for much
more ideal control of the wire’s trained shape, annealing
temperature, and aging time when compared to the conventional
method.
The conventional heat-treating process is referred to as
non-Continuously Fed (non-CF) heat-treating, also known as
conventional heat-treatment. This is the traditional process of
placing the wire in a fixture to hold its desired straight shape and
putting the fixture into the furnace at the desired temperature.
After the desired aging time was reached, the fixture and wire
were removed and then quenched in room temperature water to
complete the heat-treating process. However, this conventional
method has several issues. Opening the door of the furnace
4 Copyright © 2020 by ASME
reduces the temperature significantly and causes the resulting
wire to be trained at a lower than desired temperature.
Additionally, the thermal expansion of the wire is not
appropriately accounted for with the fixture. The conventional
method does not allow for sections larger than the width of the
fixture to be trained making it not suitable for applications
requiring long lengths of heat-treated wire.
Differential Scanning Calorimetry (DSC) analysis was
performed on the heat-treated wire to determine the transition
temperatures of the wire and to determine if any R-phase was
present. The transition temperatures were necessary to
determine the affects the annealing temperatures and aging times
had on the Nitinol wire. It was decided that DSC analysis was
only to be performed on the heat-treated CFS wire since the
change in the transition temperature is a function of the aging
temperature and time [12, 14]. The annealed and aged Nitinol
wire was tested in a DSC machine in accordance to ASTM
F2004-17 [15]. This standard calls for the Nitinol wire to be
precisely cut and placed into a DSC pan between 25 and 45 mg
in mass. The testing procedure began with the sample at 0°C,
heated to 160°C, and then cooled back down to 0°C at a ramp
rate of 10°C/min. The temperatures were chosen to ensure the
sample was cooled to 30°C below the
𝑀
"
and heated 30°C above
the
𝐴"
. The data recorded by the DSC Q2000 includes testing
time, temperature (
℃
), heat flow (mW), and heat capacity (mJ).
This test determined the Nitinol wires’
𝑀!
,
𝑀
"
,
𝐴!
, and
𝐴"
temperatures, as well as if there was any
𝑅!
and
𝑅"
present
(Figure 5).
Figure 5:Example DSC results with construction lines to determine
phase transitions
The tensile testing machine used was an Instron-5560
equipped with a 1kN load cell. The tensile tests were performed
with a grip separation of 150mm and an extension rate of
5mm/min. The data recorded consisted of force and extension
from the initial position using the load cell and crosshead motion
respectively. This was then converted to a stress-strain curve
using the geometry of the treated wire. ASTM F-2516 states that
a class C extensometer should be used for strain measurement
when testing Nitinol wires thicker than 0.2mm; however, at the
time of this testing, this capability was not available [16].
EXPERIMENTS
A tensil e test was performed for the 0.25mm high-
temperature actuator wire at the various annealing temperatures,
aging times, and heat-treatment methods for a total of 30 sample
variations (Tabl e 1).
Ta bl e 1: Annealing temperatures and aging times
Tem p (°C)
Aging Times
(minutes)
Heat-Trea tmen t Sy ste m
400
1, 2, 3
CFS
Conventional
425
1, 2, 3
CFS
Conventional
450
1, 2, 3
CFS
Conventional
475
1, 2, 3
CFS
Conventional
500
1, 2, 3
CFS
Conventional
The range of tested annealing temperatures were
determined after performing an initial tensile test on wire heat-
treated over a range of 350-550°C. Annealing temperature above
500°C (for this specific wire and aging times) resulted in
extremely high yield strains. Wire annealed at 550°C
demonstrated a 40%-50% yield strain and decrease in UTS,
which only occurs if the Nitinol recrystallizes, rendering it
useless for most actuator applications. The annealed wire would
be useful in applications where actuation stroke is more
important than blocking force, however this would be examined
in future work. An example can be seen in Figure 6 for the aging
time of 1 minute; it was determined annealing the wire at
temperatures higher than 500°C would not be included in the
analysis.
Figure 6: Tensile testing results for various annealing temperatures to
include 550C
5 Copyright © 2020 by ASME
Five tensile tests were conducted for each sample
variation, then averaged to determine the stress-strain behavior.
The 400°C annealed wire using the CFS system at the various
aging times were compared to those of the conventional system
(Figure 7). The CFS heat-treated wire demonstrates that as the
aging time is increased, the yield strain and the UTS decreases.
For the conventional heat-treatment system, the yield strain
increases as the aging time increases, but the UTS is random.
However, there is not a drastic difference and therefore, the aging
times at 400°C should be increased.
Figure 7: Tensile testing results for 400C annealed wire at various
aging times
The 425°C annealed wire using the CFS system at the
various aging times were compared to those of the conventional
system (Figure 8). The CFS heat-treated wire demonstrates that
as the aging time is increased, the yield strain and the UTS
decreases. For the conventional heat-treatment system, the yield
strain is almost 3% higher than the CFS wire. The reason for the
higher strain is a result of having to open the door of the furnace
which reduces the temperature significantly and causes the
resulting wire to be trained at a lower than desired temperature
and the thermal expansion of the wire is not appropriately
accounted for with the fixture.
Figure 8: Tensile testing results for 425C annealed wire at various
aging times
The tensile testing results of the CFS heat-treated wire
were then compiled and interpolated using a cubic interpolation
scheme to allow for simple analysis. The yield stress results
show that increasing the aging temperature has an appreciable
effect on the yield strength of the Nitinol wire as seen in Figure
10. As the temperature increases, the tensile strength decreases.
Aging time has a similar effect, albeit less sensitive than the
aging temperature. These results have an intuitive relationship
which shows that the higher the temperature of the Nitinol wire
at the quenching point the lower the yield stress.
Figure 9: Surface showing interpolated yield stress of the aged Nitinol
wires
The yield strain results follow a generally similar trend
to that of the yield strength with respect to the aging temperature
as seen in Figure 10 (i.e. higher temp, lower yield strain).
However, the aging time results are much different. The longer
the wire remains in the thermal chamber before quenching, the
higher the yield strain. These results can be used to “design” the
desired Nitinol properties at yield which would allow easier
implementation of Nitinol in various applications.
Figure 10: Surface showing interpolated yield strain of the aged
Nitinol wires
6 Copyright © 2020 by ASME
RESULTS AND DISCUSSION
The one-minute aging time at various temperatures
using the CFS demonstrate that as the annealing temperature
increases, the strain failure rate and UTS decrease (Figure 11).
This data is extremely consistent when compared to that of the
conventional heat-treatment system.
Figure 11: Tensile testing data for various annealing temperatures at
one-minute aging
The two-minute aging time at various temperatures
using the CFS demonstrate the strain failure rate is slightly lower
at all temperatures than the one-minute aging. Additionally, as
the annealing temperature increases, the strain failure rate and
UTS decrease (Figure 12). This data is extremely consistent
when compared to that of the conventional heat-treatment
system.
Figure 12: Ten s il e t es t in g d at a f or v a ri o us a n ne al i ng t e mp er a tu re s a t
two-minute aging
The three-minute aging time at various temperatures
using the CFS demonstrate the strain failure rate is slightly lower
at all temperatures than the one-minute and two-minute aging.
Additionally, as the annealing temperature increases, the strain
failure rate and UTS decrease (Figure 13). This data is extremely
consistent when compared to that of the conventional heat-
treatment system.
Figure 13: Ten s il e t es t in g d at a f or v a ri o us annealing temperatures at
three-minute aging
Ultimately, it can be concluded that the CFS heat-
treatment system produces extremely consistent stress-strain
curves and follows a pattern throughout. The conventional
system that is used by most produces random data that various at
each temperature and aging time.
DSC Analysis
DSC analysis was conducted on the following
annealing temperatures: 400°C, 425°C, 450°C, 475°C, and
500°C. Each of these temperatures were aged for 1 minute, 2
minutes and 3 minutes and the transition temperatures were
determined. DSC analysis was also performed on the wire with
the 550°C annealing temperature. These results showed that
with an aging time of 3 minutes, the R-phase present in some of
the higher temperatures was no longer present due to the
recrystallization that occurred with these heat-treating
parameters. It can be seen after performing DSC analysis on
the various annealed and aged wire, the transition temperatures
at multiple iterations of tests on the various annealed and aged
wire, the transition temperatures were determined. DSC results
for a 1-minute aging time (Figure 14) show that as the annealing
temperature increased, so did the
𝐴!
and
𝐴"
temperatures. The
𝑀!
and
𝑀
"
temperatures did change; however, this change was
marginal at most and was considered to be insignificant. As
expected, the R-phase was present in the higher temperature
samples beginning at 500°C.
7 Copyright © 2020 by ASME
Figure 14: DSC results for various annealing temperatures with a
1-minute aging time
The DSC results for an aging time of 2 minutes (Figure
15) were broadly similar to the 1-minute aging time. However,
the overall location of the transition temperatures was higher due
to the longer aging time leading to the Nitinol reaching a higher
temperature.
Figure 15: DSC results for various annealing temperatures with a
2-minute aging time
DSC results for a 3-minute age time (Figure 16) while
generally consistent with the previous two age times, the R-
phase is present at lower annealing temperatures. Additionally,
the higher age time resulted in the
𝐴!
and
𝐴"
temperatures having
a larger overall range than the other two age times.
Figure 16: DSC results for various annealing temperatures with a
3-minute aging time
The results of the DSC testing were then compiled and
interpolated using the same scheme as the tensile testing data.
This allows for quick analysis of the whole test matrix (Figure
17). These results show that as the aging temperature increases,
so does
𝐴!
. The same is true for the aging time. This leads to the
conclusion that the higher the temperature at the quenching time
the higher
𝐴!
is.
Figure 17: Surface showing interpolated Austenitic start temperature
of the aged Nitinol wires
Similarly, to
𝐴!
,
𝐴"
also increases as the annealing
temperature and time in the thermal chamber increases (Figure
18). This is because the transition temperatures generally remain
quite close to each other regardless of the aging process. These
results, like the tensile testing results, allow for the “design” of
the properties of the Nitinol wires allowing for easy
implementation of specific properties.
8 Copyright © 2020 by ASME
Figure 18: Surface showing interpolated Austenitic finish temperature
of the aged Nitinol wires
CONCLUSION
In conclusion, the Continuously Fed heat-treatment
System resulted in rapid sample creation at a higher quality heat
treatment than the conventional method: The annealing
temperature was much more consistent throughout the heat-
treating process. Precise speed control of the wire allowed for
the annealing time to be more accurate for each sample. Ten sion
control of the wire allowed for the thermal expansion of the
Nitinol to be all but neglected as the system actively changes the
wire tension as the material expands. Additionally, the tension
control removed the need for the manufacturing of a fixture
which, if poorly manufactured, could lead to incorrectly trained
shapes after heat treating. The continuous quenching system
placed after the exit of the furnace made the transition between
the heated wire and the cooled wire much faster than the
conventional method. All of these contributed to improved
sample consistency. The tensile testing data shows that the CFS
had overall much lower strain at failure that is more consistent
with trained wire produced by manufacturers (with the exception
of the lower annealing temperatures, which if aged for longer
would likely produce results more similar to the higher annealing
temperatures). DSC results showed that the novel heat treatment
process was consistent with those of previous researchers. Being
able to create long sections of straight trained wire allows for
Nitinol wires to be used on much larger applications for a much
lower cost as the user is no longer limited by the size of their
furnace. In future work, different tension on the heat-treated
wire will be utilized to determine if the tension affects the overall
functional properties of the wire. Additionally, different
diameter wire will be utilized, and multiple stage heat-treatments
will be performed. These results will then be compared to the
conventional heat-treatment wire as well as the single stage
continuously fed wire.
REFERENCES
[1]
S. e. Cai, "Effect of heat treatment temperature on nitinol
wire," Applied Physics Letters, vol. 105, pp. 105-109,
2014.
[2]
A. Skalitzky, A. Gurley, D. Beale and K. Kubik, "Design
and Analysis of SMA Woven Fabric," Smart Marterials,
Adaptive Structures, and Intelligent Systems, vol. 8206,
pp. 1-8, 2018.
[3]
J. Eaton-Evans, J. M. Dulieu-Barton, E. Little and I. A.
Brown, "Observations during mechanical testing of
Nitinol," J. Mechanical Engineering Science, vol. 222, no.
Part C, pp. 97-105, 2008.
[4]
M. Drexel, G. Selvaduray and A. Pelton, "The Effects of
Cold Work and Heat Treatment on the Properties of Nitinol
Wire," in International Conference on Shape Memory and
Superelastic Technologies, Pacific Grove, 2006.
[5]
J. Shaw, C. Churchill and M. Iadicola, "Tips and Tricks for
Characterizing Shape Memory Alloy Wire - Part 1 -
Differential Scanning Calorimetry and Basic Phenomena,"
Experimental Characterization of Active Materials Series,
pp. 55-62, 2008.
[6]
S. Yoon and D. Yeo, "Phase Transfromations of Nitinol
Shape Memory Alloy by Varying with Annealing Heat
Treatment Conditions," in Smart Materials, Nano-, and
Micro-Smart Systems, Sydney, 2004.
[7]
T. Duer ig an d A. Pelton, "Ti -Ni Shape Memory Alloys,"
Materials Properties Handbook: Titanium Alloys, vol. 1,
pp. 1035-1048, 1994.
[8]
W. H u an g a n d W. T oh , " Tr a in in g tw o-way shape memory
alloy by reheat treatment," Journal of Materials Science
Letters, vol. 19, pp. 1549-1550, 2000.
[9]
J. Fuentes, G. P and J. Strittmatter, "Phase change behavior
of nitinol shape memory alloys: Influence of heat and
thermomechnical treatments," Advanced Engineering
Materials, vol. 4, pp. 437-451, 2002.
[10]
K. e. a. Gall, "The influence of aging on critical
transformaton stress levels and martensite start
temperatures in NiTi: part II," Engineering Material
Te ch no l og y, vol. 121, pp. 28-37, 1999.
[11]
D. Miller and D. Lagoudas, "Influence of cold work and
heat treatment on the shape memory effect and plastic
strain development of NiTi," Materials Science and
Engineering, vol. A308, pp. 161-175, 2001.
[12]
M. Losertova, M. Stencek, D. Matysek, O. Stefek and J.
Drapala, "Microstructure evolution of heat treated NiTi
alloys," in 27th Joint Seminar Development of Materials
Science in Research and Education, Ostrava-Poruba,
2017.
[13]
H. Sadiq, M. Wong, R. Al-Mahaidi and X. Zhao, "The
effects of heat treatment on the recovery stresses of shape
9 Copyright © 2020 by ASME
memory alloys," Smart Materials and Structures,
vol. 19,
pp. 1-7, 2010.
[14]
G. e. a. Eggeler, "On the effect of aging on martensite
transformations in Ni-ri
ch NiTi shape memory alloys,"
Smart Marterials and Structures, vol. 14, pp. 186-
191,
2005.
[15]
A. F2004-17, "Standard Test Method for Transformation
Tem pe rat ur e of Nickel-Titanium Alloys by Thermal
Analysis," 2017. [Online]. Available:
https://doi.org/10.1520/F2004-17. [Accessed 1 September
2018].
[16]
A. F2516-18, "Standard Test Method for Tension Testing
of Nickel-Titanium Superelastic Materials," 2018.
[Online]. Available: https://compass-astm-
org.spot.lib.auburn.edu/EDIT/html_annot.cgi?F2516+18.
[Accessed 5 March 2020].
