The Effects of Cold Work and Heat Treatment
`on the Properties of Nitinol Wire
`
`Masao J. Drexel and Guna S. Selvaduray
`San Jose State University, San Jose, CA, USA
`
`Alan R. Pelton
`Nitinol Devices and Components, Fremont, CA, USA
`
`
`
`
`ABSTRACT
`
`The successful medical application of Nitinol requires precise control of its transformational and
`mechanical properties. In this study the effects of heat treatments at 300-550°C for 2-180 minutes
`on Ti-50.8at%Ni wire with 30% and 50% initial cold work were investigated. Transformational
`and mechanical properties were characterized through the BFR technique and tensile testing.
`Thermally activated precipitation and annealing processes were observed. Annealing processes
`tended to increase the maximum slope of the BFR curves. The R-phase was observed with
`greater frequency and prominence in the 50% cold-worked wire after heat treatment. The general
`trends in Af are summarized in two TTT diagrams; both illustrate a maximum precipitation rate
`of Ni4Ti3 at 400-450°C. The trends in tensile properties are outlined for all heat treatment
`conditions. Recovery processes occurred at all temperatures. The onset of recrystallization
`occurred at approximately 450°C for both wires.
`
`INTRODUCTION
`
`Since its discovery in the early 1960s Nitinol has attracted an increasing amount of attention. A
`near room-temperature phase transformation results in shape-memory and superelastic properties,
`which allow Nitinol to provide functionality never possible with conventional engineering alloys
`[1]. Additionally, the good biocompatibility of Nitinol has allowed it to be successfully used in a
`variety of biomedical devices [2]. Devices such as stents, vena cava filters, and endodontic files
`(shown in Fig. 1) all place unique demands on the material. The successful application of Nitinol
`requires precise control of both transformational and mechanical properties. Ultimate tensile
`stress (UTS), upper plateau (UP) stress, lower plateau (LP) stress, and austenite finish
`temperature (Af) are some of the relevant properties. These properties can be adjusted, through
`careful processing, to optimize performance for a given application.
`
`Heat treatment is the most common process used to tailor the properties of Nitinol. During aging,
`the nucleation and growth of Ni-rich precipitates has been well documented and is commonly
`used to increase the Af for Ni-rich compositions [3,4]. These precipitates have also been observed
`to act as effective barriers to dislocation motion, thus strengthening the alloy [4]. Furthermore,
`the strain fields introduced by precipitates can act to stabilize the rhombohedral or ‘R-phase’
`resulting in two-stage transformation [5,6].
`
`Proceedings of the International Conference on Shape Memory and Superelastic Technologies
`May 7–11, 2006, Pacific Grove, California, USA
`Brian Berg, M.R. Mitchell, and Jim Proft, editors, p 447-454
`
`
`
`Copyright © 2008 ASM International®
` All rights reserved.
` DOI: 10.1361/cp2006smst447
`
`
`
`447
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`Figure 1: Bent Nitinol endodontic file.
`
`
`
`
`Heat treatments that provide the thermal energy required for precipitation can also activate the
`processes of annealing, during which the rearrangement of defects and the decrease in defect
`density removes the stored strain energy within the lattice. These processes affect both the
`thermal and mechanical properties. The driving force for annealing is greater in more heavily
`cold-worked metals due to their higher amount of stored internal energy [7]. Therefore, the
`response to heat treatment depends on the processing history, time, temperature, and amount of
`cold work.
`
`OBJECTIVE
`
`This study was undertaken to outline the trends in the transformational and mechanical properties
`of as-drawn Nitinol wire, with 30% and 50% initial cold work, during heat treatment at a range
`of typical processing temperatures where precipitation, recovery, and recrystallization occur.
`
`EXPERIMENTAL METHODOLOGY
`
`Ti-50.8at%Ni wires with 30% and 50% cold work at a final diameter of 1mm were studied. Heat
`treatments were performed in a salt bath followed by a water quench. A summary of the heat
`treatment temperatures and times is given in Table 1.
`
`Table 1: Heat treatment conditions
`Temperature (°C)
`300, 350, 400, 450, 500,
`525, 550
`
`Time (min.)
`2, 5, 10, 20, 60, 180
`
`
`The transformational properties were characterized by the bend and free recovery (BFR)
`technique in accordance with ASTM F 2082-03. This test involves cooling the wires in an
`alcohol (or water) bath to stabilize the martensite phase. A temperature of -70°C was usually
`sufficient in this study. The wires were then strained 2-2.5% and their recovery was monitored
`during heating. Due to the high levels of cold work and the R-phase present in many of the
`specimens tested in this study the Af was not determined by the intersection of tangent lines.
`Instead Af was identified as the point where displacement ended. Three specimens were tested for
`each heat treatment and cold work condition.
`Tensile properties were characterized by ASTM F2516-05, which consists of loading the
`specimen to 6% strain, unloading to a stress of 5MPa, and then loading to failure. To avoid
`
`448
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`temperature effects, a strain rate of 0.06 in/min (~1.5 mm/min) was utilized in this study. Testing
`was performed at room temperature.
`TRANSFORMATIONAL PROPERTIES
`
`BFR testing was attempted with both the as-drawn wires and after heat-treatment at 300°C.
`However, even after cooling to below -100°C the springback of these wires was so significant
`that testing could not be performed.
`
`The increase in Af resulting from the formation of Ni-rich precipitates is observable throughout
`the BFR plots. The time-temperature-transformation (TTT) diagram constructed by Nishida et al.
`had established that at the temperatures of 500-800°C the precipitation sequence of Ni4Ti3 →
`Ni3Ti2 → Ni3Ti occurs [3]. As the Ni concentration in the surrounding matrix is depleted during
`precipitation, the transformation temperature increases with increasing heat treatment time, as
`illustrated in Fig. 2, for both 30% and 50% cold-worked wires. Note that the shape of the low-
`temperature BFR curves do not significantly change with increased heat treatment time.
`However, higher temperature heat treatments do change the shape of the BFR curves, due to
`annealing, and stabilization of the R-phase; these processes tend to interfere with the trend of
`increasing Af, as illustrated in Figs. 3 and 5. The higher level of residual cold work present in the
`50% cold-worked wire was found to flatten the corresponding BFR curves in Fig. 2. A
`comparison of the BFR curves in Figs. 2 through 4 illustrates that with increasing heat treatment
`temperature, and time at the higher temperatures, the maximum slope of the BFR curves
`increases. This indicates that the austenite-martensite transformation proceeds more readily as the
`internal stresses introduced during cold work are released as a result of annealing.
`
`The trend of increasing Af of the 50% cold-worked wire is seen to temporarily pause during heat
`treatment of 2-20 minutes at 450°C, seen in Fig. 3. This pause is the result of the annealing
`processes that cause a shortening of the tail of the BFR curve. The reduction in the tail length
`effectively neutralizes the increase of the Af. These observations illustrate the interplay between
`the precipitation reaction and the annealing processes. This pause in the increase of Af can also
`be observed in Fig. 5.
`
`The presence of the R-phase is clearly observed in Fig. 3. A comparison of the 30% and 50%
`cold-worked BFR curves shows that the R-phase is stabilized after only two minutes of heat
`in
`the 50% cold-worked wire. At
`times of 2-20 min.
`treatment at 450°C
`
`
`
`Figure 2: BFR curves for 30% and 50% cold-worked Nitinol wires after heat treatment at 350°C.
`the R-phase flattens the ‘tail’ of the recovery curve. Longer heat treatments result in a typical
`two-stage transformation. In the 30% cold-worked wire, however, the R-phase is only observed
`after heat treatment for 20 minutes. This supports previous studies, which reported that the R-
`phase is stabilized by cold work [5,6,8].
`
`
`449
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`Following heat-treatment for 2-10 minutes at 550°C the wires are observed to posses lower Af
`values, shown in Fig. 4. The decrease in Af at these high temperatures has been addressed in a
`study by Pelton et al. who explained that 300-500°C heat treatments result in the formation of the
`Ni4Ti3 precipitate; heat treatments at 500-600°C, however, result in the dissolution of the Ni4Ti3
`precipitate followed by the eventual formation of the Ni3Ti2 and Ni3Ti precipitates [4]. The initial
`dissolution process increases the Ni concentration of the matrix thereby decreasing the Af. This is
`observed as the low Af values for heat treatments of 2-20 minutes at 550°C. The rapid increase in
`Af during heat treatments of 20-180 minutes, seen in Figs. 4 and 5, is attributed to the efficient Ni
`depletion resulting from formation of the Ni3Ti2 and Ni3Ti precipitates.
`
`Also interesting to note is the unique shape of the BFR curves obtained after heat treatment at
`550°C for 60 minutes for wires with both levels of cold work. Similar curves were also obtained
`after heat treatments of 180 minutes at 525°C. These curves are distinguished by a sharp
`transformation onset followed by a slow end – a long tail. The cause of these curves shape is not
`known but it is suspected that the shape of these curves may result from a point in the
`precipitation process where the precipitate-matrix coherency strains are at a maximum. This
`unique transformation behavior is also noted to disappear with continued thermal input
`suggesting it results from a transient microstructure obtained during heat treatment.
`
`
`
`Figure 3: Representative BFR curves for 30% and 50% cold-worked Nitinol wires after heat-
`treatment at 450°C.
`
`
`
`Figure 4: Representative BFR curves for 30% and 50% cold-worked Nitinol wires after heat-
`treatment at 550°C.
`
`Summaries of the trends in Af for all heat treatment temperatures, and for each level of cold
`work, are shown in Fig. 5. A comparison of the two Af versus heat treatment time plots shows
`that the general trends in Af are similar for the samples of both levels of cold work. At short
`times the intermediate temperatures of 400-450°C are the most effective at increasing the Af of
`the wires, due to the maximum precipitation rate of Ni4Ti3. At longer times the highest heat
`treatment temperature of 550°C yields the highest Af values. This is due to the efficient removal
`
`450
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`of Ni from the matrix during formation of the higher Ni concentration Ni3Ti2 and Ni3Ti
`precipitates.
`
`The only significant effects of cold work on the Af occur at the intermediate temperature range of
`400-450°C. Here the more heavily cold-worked wire has a higher Af at short heat treatment times
`as a result of the long BFR tails. Interestingly, this temperature range falls very close to the
`temperatures commonly utilized in Af tuning processes and therefore these differences are
`relevant when processing medical devices.
`
`The general trends in Af during heat treatment are illustrated in Fig. 6 as two TTT diagrams.
`These two diagrams show the increase in Af at 400 and 450°C as a shift in the nose to the left.
`This shift corresponds directly to the long tails of the 50% cold-worked BFR curves seen in Fig.
`3. It should be noted that these diagrams were constructed from a small number of points and are
`therefore only intended to illustrate the general trends in Af increase – hence the dotted lines. As
`previously noted by Pelton the shape of the TTT diagrams results from the balance between the
`nucleation and growth phenomena [4]. At high temperatures diffusion rates are high but the
`driving force for nucleation is low while at lower temperatures the reverse is true. Therefore, it is
`at intermediate temperatures where the combination of the nucleation and growth rates are
`maximized, resulting in a maximum rate of Af increase.
`
`
`
`Fig. 5: Trends in Af of 30% and 50% cold-worked Nitinol wires during heat treatment at a range
`of temperatures.
`
`
`
`
`Figure 6: Trends in Af of 30% and 50% cold-worked Nitinol wires during heat treatment at a
`range of temperatures.
`
`451
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`MECHANICAL PROPERTIES
`
`The UTS of the as-drawn 30% and 50% cold-worked wires were 1450 and 1900 MPa
`respectively. The high level of stored strain energy existing within the as-drawn wires inhibited
`stress inducement of the martensite phase and no ‘superelastic flags’ were therefore observed
`during tensile testing. Heat treatment at 300°C for only two minutes can be seen to restore the
`superelastic flag of both wires, as shown in Fig. 7. The higher level of residual cold work after
`heat treatment at 300°C results in the steeper slope of the UP and LP, i.e., strain hardening, as
`well as the higher UTS of the 50% cold-worked wire. Similar to the relatively flat BFR curves
`shown in Fig. 2, the forward phase transformation is impeded.
`
`
`
`Figure 7: Tensile responses of 30% and 50% cold-worked Nitinol wires after heat treatment at
`300°C.
`
`In Figs. 7 and 8 heat-treatment at 450°C and below can be seen to result in an increase in UTS of
`the 30% cold-worked wire. This strengthening is the result of the precipitates acting as effective
`barriers to dislocation motion. During heat treatment at 500°C and above the processes of
`annealing dominate the precipitation strengthening effect and a drop in UTS results. A decrease
`in UTS of the 50% cold-worked wire can be seen for all heat treatment conditions in Fig. 8. This
`decrease is the result of the higher driving force for annealing resulting from the higher stored
`internal energy within the 50% cold-worked wire. Precipitation strengthening, however, does still
`occur in competition with the recovery processes. This can be seen in Fig. 9 where both 30% and
`50% cold-worked wires posses a ‘bump’ in UTS from 300-450°C. This bump is the
`superposition of the precipitation strengthening and recovery effects. Also observed in Fig. 9 is
`the sharp drop in UTS after heat treatments at 450°C and above, which corresponds to the onset
`of recrystallization (marked by
`the dashed
`line). The decrease
`in UTS corresponds
`
`
`
`Figure 8: Trends in UTS of 30% and 50% cold-worked Nitinol wires during heat treatment at a
`range of temperatures.
`
`452
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`to the increase in the slope of the BFR plots previously noted. Although microstructural
`investigation was not performed here, a previous study by Miyazaki has shown the
`recrystallization of cold-worked Nitinol after a one-hour heat treatment at 500°C [8].
`
`An additional heat treatment was performed at 650°C in an attempt to capture the end of
`recrystallization – the onset of grain growth. However, this did not occur as the rapid UTS
`decrease continued, as shown in Fig. 9. The 650°C heat treatment was performed in an air
`furnace due to the temperature limitations of the salt pots. The relatively long heat treatment time
`in the air furnace makes the heat-up time negligible.
`
`
`
`Figure 9: UTS of 30% and 50% cold-worked Nitinol wires after 60-minute heat treatments at a
`range of temperatures.
`
`The net result of the annealing and precipitation processes on the plateau stresses is outlined in
`Figs. 10 and 11. Annealing is responsible for the evolution of the superelastic flag observed in
`Fig. 7, which is primarily responsible for the rapid decrease in UP and LP stresses during heat
`treatment at 300°C. Heat treatments at 400°C and above results in recovery of a relatively flat UP
`and LP after only two minutes. The continued decrease in UP and LP stresses results from the
`increasing Af due to precipitation. As Af increases the stress required to induce the transformation
`to martensite is reduced, as is the stress driving the reverse transformation – the LP stress. The
`rapid decrease in UP stress during heat treatment for 60 and 180 minutes at 550°C corresponds
`directly to the rapid increase in Af occurring at these conditions.
`
`The range of plateau stresses obtained during heat treatment of the 50% cold-worked wire is
`broader than those of the 30% cold-worked wire. The spread of the plateau values may or may
`not be preferred depending on the target UP and LP stresses for a specific application.
`
`
`
`Figure 10: Trends in UP stress of 30% and 50% cold-worked Nitinol wires after heat treatment
`at a range of temperatures.
`
`453
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner
`
`

`
`
`Figure 11: Trends in LP stress of 30% and 50% cold-worked Nitinol wire after heat treatment at
`a range of temperatures.
`
`CONCLUSIONS
`
`This investigation has outlined the trends in several of the pertinent properties of cold-worked Ti-
`50.8at%Ni wire during heat treatment. The thermally activated processes of precipitation and
`annealing have been seen to occur simultaneously and result in substantial changes of the thermal
`and mechanical properties of Nitinol. Temperature, time, and the amount of prior cold work all
`influence the materials net response to heat treatment. To obtain the ideal combination of
`properties for a given application these factors must be considered during device processing.
`
`ACKNOWLEDGMENTS
`
`This research was made possible through the support of Nitinol Devices and Components,
`Fremont, Ca.
`
`REFERENCES
`
`[1] Otsuka, K., Kakeshita, T., “Science and Technology of Shape-Memory Alloys: New
`Developements,” MRS Bulletin, Vol. 27, No. 2 (2002), pp. 91-97.
`[2] Pelton, A., Stockel, D., Duerig, T., “Medical Uses of Nitinol,” Materials Science Forum,
`Vol. 327 (2000), pp. 63-70.
`[3] Nishida, K., Wayman, C., Honma, T., “Precipitation Processes in Near-Equiatomic TiNi
`Shape Memory Alloys,” Metallurgical Transactions A, Vol. 17A (1986), pp. 1505-1515.
`[4] Pelton, A., DiCello, J., Miyazaki, S., “Optimization of Processing and Properties of Medical-
`Grade Nitinol Wire,” SMST-2006 Conference Proceedings, (2000), pp. 361-374.
`[5] Ren, X., Miura, N., Zhang, J., Otsuka, K., Tanaka, K., Koiwa, M., Suzuki, T., Chumlykov,
`Y., Asai, M., “A Comparitive Study of Elastic Constants of Ti-Ni-based Alloys Prior to
`Martensitic Transformation,” Materials Science and Engineering A, Vol. 312 (2001), pp.
`196-206.
`[6] Otsuka, K., Engineering Aspects of Shape Memory Alloys, eds. T. Duerig, et al.,
`Buttersworth-Heinemann (London, 1990), 36-45.
`[7] Reed-Hill, R., Abbaschian, R., Physical Metallurgy Principles, PWS-KENT Publishing Co
`(Boston, 1992), pp. 361-374.
`[8] Miyazaki, S., Engineering Aspects of Shape Memory Alloys, eds. T. Duerig, et al.,
`Buttersworth-Heinemann (London, 1990), 369-413.
`
`454
`
`PGR2015-00019 - Ex. 1036
`US ENDODONTICS, LLC., Petitioner