`Technologies, Kunming, China, P.285-292 (2001).
`
`Fabrication of Nitinol Materials and Components
`
`Ming H. Wu
`Memry Corporation, 3 Berkshire Blvd., Bethel, CT 06801, U.S.A.
`
`Abstract. As Nitinol emerges to find more and more applications in medical devices and industrial
`commercial markets, understanding on the effects of material processing becomes increasingly
`important. More so than other materials, properties of Nitinol are significantly affected by its
`fabrication processes. There are abundant amounts of processing data available in the industry.
`However, they are often kept proprietary and not released in the public domain. This paper reviews
`available publications on Nitinol fabrication processes, such as melting, forming, shaping, treating,
`cutting, joining and finishing, etc., commonly employed or developed for the manufacture of
`Nitinol material and components. Effects of these operations on the material properties are
`discussed.
`
`Introduction
`
`Nitinol is well known for its thermomechanical properties of superelasticity and shape memory
`effect. Along with excellent kink resistance, biocompatibility and MRI compatibility, the alloy has
`emerged as a unique biomaterial finding an increasing number of medical applications such as
`orthodontic wires, orthopedic devices, guide wires, stents, filters and components in minimally
`invasive surgical devices [1]. Development and growth of Nitinol applications in the industrial and
`commercial markets have also been fairly strong in recent years. Eyeglass frames, cellular phone
`antenna, high pressure sealing plug for diesel fuel injectors, over-temperature protection device for
`lithium ion battery are among the applications experiencing significant growth [2].
`
`As more Nitinol products are being developed and produced in mass scale, controlling
`manufacturing processes to deliver consistent quality becomes increasingly important. Nitinol
`properties are extremely sensitive to the initial chemistry as well as the subsequent processing. The
`alloy’s flexibility, significant work-hardening rate and high titanium content also pose additional
`challenges in metal fabrication. Understanding various manufacturing processes and their effects
`on product performance is the key to successful process controls.
`
`Although significant efforts have been devoted to the understanding of Nitinol processing issues,
`large amount of the information has been kept proprietary. Present paper reviews available public
`information related to melting, fabrication, secondary processing methods and finishing techniques
`commonly used for the manufacture of Nitinol products. The review also includes novel processing
`methods currently in development.
`
`Melting
`
`General requirements on Nitinol chemistry and trace elements are defined in an ASTM standard,
`F2063-00 [3]. The Nitinol transformation temperatures are extremely sensitive to a small variation
`in the Ni or Ti concentration. The sensitivity increases with Ni content in the alloy. For alloys
`having greater than 55.0 weight percent Ni, a one weight percent deviation in Ni (or Ti)
`
`1
`
`Medtronic Exhibit 2008
`Edwards v. Medtronic
`IPR2014-00362
`
`
`
`concentration will result in approximately a 100°C shift in transformation temperatures. This
`extreme sensitivity puts a strict requirement on any melting practice to tightly control the Ni and Ti
`ratio in order to meet the required tolerance in transformation temperatures.
`
`Having close to fifty percent Ti, molten Nitinol is highly reactive and must be processed in vacuum.
`Both vacuum induction melting (VIM) and vacuum consumable arc melting (VAR) processes are
`commonly used for production. Other melting processes such as non-consumable arc melting,
`electron beam melting and plasma melting are also used in experimental scales. In the VIM process,
`graphite or calcia (CaO) crucible is preferred. Other crucible material such as alumina or magnesia
`contaminates the molten Nitinol with oxygen. Molten Nitinol does pick up carbon contaminant
`from graphite crucible. By keeping the melting temperature below 1450°C when using a graphite
`crucible, the carbon content of a VIM Nitinol ingot can be controlled between 200 and 500 ppm [4].
`The transformation temperature can be controlled within +/- 5°C in a VIM ingot. To achieve a
`more precise control, an in-situ composition control process can be implemented where samples are
`taken from the molten metal and the transformation temperatures are quickly analyzed for instant
`composition adjustments [4]. One also needs to bear in mind that because Nitinol transformation
`temperatures are so sensitive to small variations in chemistry that the analytical methods are not
`sufficiently accurate to predict the transformation temperature. The transformation temperature in
`fact correlates much better with the charge chemistry that is used more reliably for ingot
`formulation [5]. In the VAR process, consumable electrodes of Nitinol are melted and solidify in a
`water-cooled copper mold. Because there is no contamination from the crucible, cleaner materials
`with carbon content less than 200 ppm are produced. However, the molten pool of a VAR process
`is limited to only a small zone. Lacking complete mixing within the entire ingot leads to less
`homogeneous chemistry and the distribution in transformation temperature when compared to the
`VIM ingot. Multiple re-melts are required in the VAR process to achieve acceptable homogeneity.
`Double melting process using VIM primary melting followed by VAR re-melt is used in production
`with good results. Ingots of 1,000 Kg and 14 inch (35.5 cm) diameter are routinely produced using
`the VIM/VAR double melt process [5].
`
`Regardless of the fundamental difference between the two melting processes, wires manufactured
`from VAR and VIM/VAR double melt appear to have similar mechanical and fatigue properties [6].
`
`Metal Fabrication
`
`After melting, the Nitinol ingot is usually forged and rolled into a bar or a slab at elevated
`temperatures. Extrusion of Nitinol billets and tubes at temperatures between 850°C and 950°C has
`been experimented [7]. Such hot working processes break down the cast structure and improve
`mechanical properties. Optimal hot working temperatures appear to be around 800°C where the
`alloy is easily workable and the surface oxidation in air is not too severe [4]. Following hot
`working, Nitinol alloys are cold worked and heat-treated to obtain final dimensions with desired
`physical and mechanical properties.
`
`Cold working of Nitinol is quite challenging because the alloy work-hardens rapidly. It requires
`multiple reductions and frequent inter-pass annealing at 600-800°C until the final dimension is
`obtained. Round wires are produced by die drawing processes. Retaining surface oxide, Nitinol
`wires can be successfully drawn to small sizes. Lubricants successfully used for the drawing of
`Nitinol wires include sodium stearate soap [8], molybdenum disulfide [9], graphite-containing water
`based lubricant [10] and oil based lubricant [11]. Following a similar reduction schedule,
`rectangular wires can be manufactured by drawing round wires while flat wires are typically
`produced by cold rolling. Comparing to those of rolled flat wire, the dimensions of drawn
`2
`
`
`
`rectangular wire are much more tightly controlled. For example, a typical tolerance for the rolled
`wire is +/-10% while +/-0.0005 inch (0.013 mm) or better is typical for drawn wires. Nitinol tubing
`although difficult to fabricate is routinely manufactured in commercial scale. Tubes as small as
`0.010 inch (0.25 mm) outside diameter are produced by drawing over a non-deformable or a
`deformable mandrel [12]. Rolled Nitinol sheets with thickness down to 0.010 inch (0.25 mm) and
`width up to 5 inches (125 mm) are available.
`
`Forming
`
`Because superelastic Nitinol exhibits significant spring-back when deformed in both cold worked
`and heat-treated states, the alloy is difficult to form at ambient temperatures. Over-deformation of
`superelastic Nitinol induces martensite and therefore affects the mechanical and transformation
`properties. Like shape memory Nitinol alloys, if the part is not constrained during heat treatment,
`the shape will recover partly back to the original configuration. Prototype Nitinol components are
`normally fabricated by holding the part in a fixture during heat treatment. This process can be
`scaled up to production quantities by increasing the number of fixture and heat treatment capacity.
`Semi-automation of this process is also possible by cold forming the part using an automatic
`forming machine. The formed part is then placed and constrained in a fixture and subsequently heat
`treated to a desired shape with final properties [4].
`
`Heat Treating
`
`To achieve optimized properties, materials with 30-40% retained cold work before heat treatment
`should be used. Superelastic Nitinol materials are typically heat treated in the vicinity of 500°C.
`Lower temperatures in the range between 350°C and 450°C are also suitable for shape memory
`alloys. Alternatively for alloys with greater than 55.5 weight percent Ni, good superelasticity and
`shape memory effect can be obtained by solution treatment at high temperatures between 600°C and
`900°C and subsequent aging at a temperature around 400°C. This aging process induces
`precipitation hardening of Ni-rich phases [13]. The transformation temperatures are elevated
`significantly as the matrix composition adjusts during aging [14].
`
`Machining
`
`Although it is difficult and causes significant tool wear, Nitinol can be machined using conventional
`techniques such as milling, turning and drilling. Shearing and blanking are quite effective with
`proper tool design and maintenance. Carbide tools with chlorinated lubricant are recommended for
`these operations. Abrasive processes such as grinding, sawing and water jet cutting with abrasive
`particles are successfully used for Nitinol. For example, tips of Nitinol guide wires are commonly
`tapered by centerless grinding. Laser machining; electro-discharge machining (EDM) and
`photochemical etching processes are used to fabricate Nitinol components such as stents, baskets
`and filters. In particular, laser machining has become the preferred process for the manufacture of
`Nitinol tubular stents. Modern laser cutting machine using a pulsed Nd:YAG laser and equipped
`with a CNC motion control system offers high speed, high accuracy and the capability for rapid
`prototyping. Drawbacks are the occurrences of heat-affected zone (HAZ) and microcracks.
`Managing heat and shape-stability of Nitinol parts is critical and post-processing is required to
`remove slag, microcracks and HAZ [15]. EDM works well with most Nitinol compositions. A
`recast surface layer consisting of oxides and contaminants from Cu electrode and dissolved
`dielectric medium is present and may need to be removed depending on the application [16].
`
`3
`
`
`
`Joining
`
`Welding Nitinol to itself has been successfully performed using CO2 laser [17], Nd:YAG laser [18],
`tungsten inert gas (TIG) [19] and resistance welding [20, 21] under Ar or He protective atmosphere.
`Superelasticity and shape memory effect are generally well preserved in these welding processes.
`Degradation in tensile strength and the resistance to permanent deformation in fusion zone and
`HAZ however was noted in particular during CO2 laser and TIG welding. Nd:YAG laser process,
`preserving 75% of tensile strength of the base metal and maintaining permanent deformation below
`0.2% after a 7% deformation of a superelastic weld specimen, appears to suffer the least thermal
`degradation. Ti-rich alloys are more susceptible to weld cracking. Using consumable filler metal in
`resistance welding helps to reduce the risk and significantly increases joint strength [21].
`
`Soldering Nitinol alloys using halogen-based fluxes is described in a US patent 5,242,759. Plating
`Ni on Nitinol surface improves surface wetting and allows soldering using mild fluxes [20, 22].
`Plating a secondary layer of noble metal such as Au further enhances solderability [20]. Ultrasonic
`solder joint of Nitinol using Sn based solder has also been experimented with good results. Preheat
`of base metal; ultrasonic power control and proper wetting of the iron tip by the solder are key
`factors for the success [23].
`
`Joining Nitinol to dissimilar metals is significantly more challenging. Welding Nitinol to stainless
`steel is especially difficult due to the formation of brittle intermetallic compounds, TiFe and TiFe2.
`Using a Ta interlayer, Nitinol has been successfully welded to stainless steel by percussive arc
`welding [20]. It is anticipated that other welding techniques will also work well for joining Nitinol
`to stainless steel using proper interlayer materials. With proper design and tooling, mechanical
`joining techniques such as crimping and swaging offer alternatives that can provide reliable joints
`between Nitinol and dissimilar metals.
`
`Finishing
`
`Heat treated Nitinol has a typical oxide finish ranging from straw color to blue color. Black color
`oxide may appear if the processing oxide was not removed during fabrication. These oxide layers
`can be removed by mechanical means such as grit blasting and polishing. By proper selection of
`polishing media, a mirror-like finish can be achieved by mechanical polishing. Chemically etching,
`also effective in removing surface oxide, produces a silver-looking surface. Electro-polishing of
`Nitinol has also been demonstrated to produce a highly smooth finish.
`
`Corrosion resistance of Nitinol is significantly affected by methods of surface preparation. It is a
`general belief that preferential formation of titanium oxide on the surface enhances passivity and
`corrosion resistance. Mechanically polished surfaces, although can be highly smooth, appear to be
`most susceptible to corrosion attack while the chemically etched surfaces appear to be the most
`passive [24]. Electro-polishing alone does not sufficiently enhance the corrosion resistance of
`Nitinol [24]. Passivation following ASTM F86 procedure results in much improved corrosion
`resistance and biocompatibility [25].
`
`Coating
`
`Ni coating can be deposited onto Nitinol by electrolytic or electroless processes [23]. Other metals
`such as Au can also be applied electrolytically [26]. There are risks associated with the electro-
`plating process that requires careful controls. Good adhesion and ductility of coating are needed to
`prevent flaking when deformed to large strains. Damages of coating by either flaking or scratching
`4
`
`
`
`may lead to galvanic corrosion. Hydrogen charged into the alloy during plating also needs to be
`managed to prevent hydrogen embrittlement.
`
`Polymeric coatings are applied by co-extrusion, spray coating or other surface deposition
`techniques. PTFE is spray coated and cured at high temperatures greater than 300°C which can
`affect the thermomechanical properties of Nitinol. PTFE films can also be applied by plasma
`polymerization [27]. Other polymers such as polyurethane and parylene have been applied with
`good results.
`
`Powder Processes
`
`Various powder metallurgy (PM) processes have been developed for Nitinol on experimental basis.
`These efforts carry two primary objectives, fabricating near-net shape components and more
`accurate control of transformation temperatures. Both pre-alloyed powders and elemental powders
`have been used.
`
`Pre-alloyed Nitinol powders can be fabricated by inert gas atomization [28], hydriding and
`pulverization [29] or mechanical alloying [30]. The powders are then blended, compacted and
`sintered. Reported compaction processes include hot and cold isostatic pressing [28], hot and cold
`uniaxial die compaction [31], direct powder rolling [32] and consolidation by atmospheric pressure
`(CAP) [33]. Resulting density varies from process to process. While 68% was reported for CAP,
`95% density can be achieved by hot isostatic pressing (HIP) [31]. Using a PM process, a proper
`ratio of two Nitinol powders having known different transformation temperatures are blended, hot
`isostatically pressed and sintered. A desired intermediate transformation temperature of sintered
`part can be obtained through this approach with good precision [28].
`
`Alternatively, elemental Ti and Ni powders are blended, pressed and sintered. Nitinol fabricated by
`injection molding using elemental powders has also been reported [34]. Because the fusion of Ni
`and Ti is highly exothermic, the heat of fusion can be used to synthesize Nitinol intermetallic.
`Elemental powders of Ni and Ti can be sintered using either combustion or thermal explosion
`process. In the combustion mode, the synthesis is ignited by a high temperature source and then
`propagates through the entire piece. In the thermal explosion mode, compacted powders are heated
`until an ignition temperature is reached. The fusion energy released at ignition heats up and sinter
`the material. If the heating exceeds the melting point, a cast structure can result [35]. Otherwise,
`Nitinol materials sintered from elemental powders are highly porous and may contain other
`intermetallic phases of Ti2Ni and TiNi3. Secondary processing such as remelt or extrusion may be
`required to achieve higher density and improved microstructures [36, 37].
`
`Porous Nitinol has attracted recent attention on its potential as an implant material. Materials with
`porosity in the range of 30-70% and pore size of 60-100 _m have been synthesized by sintering Ni
`and Ti elemental powders at temperatures with the presence of liquid phase. Ignition synthesis of
`porous Nitinol also achieved good results especially with increased ignition temperature [38].
`
`The limitation of Nitinol PM processes appears to be the oxygen content. Sintered Nitinol materials
`with oxygen exceeding 3000 ppm have been reported [39]. Significant improvement has been
`achieved in a recent study of a NiTiHf alloy where an oxygen level in the range of 1190-1550 ppm
`was reported after sintering elemental powders of Ti, Ni and Hf having an oxygen content of 945,
`727-948 and 2200 ppm, respectively [40]. Oxygen at these levels may not significantly affect shape
`memory effect or superelasticity, but may negatively impact ductility and fatigue resistance.
`
`5
`
`
`
`Thermal Spray
`
`Thermal spray is a conventional thick film coating technology. Nitinol coating can be applied by
`plasma spraying or physical vapor deposition (PVD) [41]. The process has also been used to
`fabricate Nitinol foils and thin wall mill products [42]. Nitinol foils of 0.5 mm thickness have been
`successfully fabricated by low-pressure wire plasma spray with subsequent rolling and interpass
`annealing. The same process has also been used to fabricate tubing of 0.2 mm and 0.76 mm wall
`thickness with subsequent HIP. Porosity was commonly present in the as-sprayed structure but was
`eliminated after rolling or HIP. A shift of 13°C in the transformation temperature (Af) was
`measured between the wire feed stock and the sprayed foil. While carbon and hydrogen contents
`slightly decrease during the process, the oxygen content went from 430 ppm for the wire to 1440
`ppm for the sprayed foil [42].
`
`Thin Film Fabrication
`
`Nitinol thin film actuators have attracted significant development efforts in the recent past. Nitinol
`films normally less than 10_m in thickness were deposited on silicon, glass or polymeric substrates
`by sputter deposition. Commercial devices consisting of Nitinol film on silicon substrate are then
`fabricated by photolithographic techniques [43]. Compositions of Nitinol thin film reported in
`various studies include binaries in the range of 48.2-51.9 atomic percent Ni [44, 45], NiTiHf [46]
`and NiTiCu [47] using either pre-alloyed or elemental targets. Deposition using pre-alloyed targets
`produced films having much more uniform compositions consistent with the target chemistry while
`the process using elemental targets results in much more difficult control of composition with
`variations across the film. Films in the as-deposited state are generally amorphous and require
`subsequent heat treatments to re-crystallize [48]. Crystalline films in the as-deposited state can be
`achieved with heated substrates [47].
`
`Miniature pneumatic valves actuated by thin film Nitinol are currently been commercialized.
`Devices such as spacers in flat panel display and other micro electromechanical systems (MEMS)
`actuated by Nitinol thin film actuators are being developed.
`
`Conclusions
`
`With the rapid growth of the Nitinol industry the associated manufacturing technologies have come
`a long way in the last decade, although not mature in any standard of metallurgical process. Recent
`releases of ASTM standards governing material specifications (F2063), standard terminology
`(F2005) and test methods (F2004 and F2082) not only standardize the language of communication
`but also promote general consensus within the industry. Significant resources are continuously
`being invested in process development by the processors as well as the end users. Advancement in
`Nitinol process development allows the material to penetrate into cost sensitive volume commercial
`markets. The medical device market can tolerate higher cost but has a strict demand on better
`quality and consistency. It is anticipated that requirements will continue to pressure tighter
`chemistry control in raw materials to reliably achieve better than +/-5ºC tolerance in transformation
`temperatures. Developments in laser cutting, photochemical etching, forming and surface finishing
`technologies having potentials to improve physical and mechanical properties, especially the fatigue
`life will be on-going. Control of surface finish to deliver unfailing corrosion resistance and
`biocompatibility will receive intense attention for medical implants. While thin film technology is
`beginning to be commercialized, process stability and quality control of thin film properties will
`probably attract more awareness. Other near-net-shape technologies such as powder and thermal
`spray processes may one day find their niches of commercial value.
`6
`
`
`
`References
`
`[3]
`
`[4]
`
`[5]
`
`[6]
`
`[7]
`
`[8]
`
`[9]
`
`[12]
`
`[13]
`
`[14]
`
`[1]
`[2]
`
`D. Stoeckel, Minimally Invasive Therapy & Allied Technologies, Vol.9, No.2 (2000), p.81.
`M.H. Wu and L.McD. Schetky, Proceedings, Third International Conference on Shape
`Memory and Superelasticity Technologies, Pacific Grove, California (2000), in press.
`ASTM Standard Specification F2063-00 (American Society for Testing and Materials,
`2000).
`Y. Suzuki, Fabrication of Shape Memory Alloys (Shape Memory Materials, ed., K. Otsuka
`and C.M. Wayman, Cambridge University Press 1998).
`F. Sczerzenie, Proceedings: Shape Memory Alloys for Power Systems, EPRI TR-105072,
`Electric Power Research Institute, p.5-1 (1995).
`M. Reinoehl, D. Bradley, R. Bouthot and J. Proft, Proceedings, International Conference on
`Shape Memory and Superelasticity Technologies, Pacific Grove, California (2000), in press.
`K. Muller, Proceedings, 8th International Conference on Metal Forming, Krakow, Poland
`(2000), p. 657.
`K. Yoshida, J. the Japan Society for Technology of Plasticity, Vol.31, No.355 (1990),
`p.1015.
`S.K. Wu, H.C. Lin, and Y.C. Yen, Materials Science & Engineering A, Vol.215, No.1-2
`(1996), p.113.
`[10] A. Matsuo, Y. Ito and N. Gotou, NiTi Based Alloy Wire And Its Production, Japanese Patent
`Publication JP08117835A (1996).
`[11] A. Matsuo, Y. Ito and N. Gotou, Method For Thinning Wire Of NiTi Series Alloy, Japanese
`Patent Publication JP08112614A (1996).
`J.A. DiCello, B.S. Minhas, J.W. Simpson, R.S. Cornelius and J.D. Harrison, Process for The
`Manufacture of Metal Tubes, United State Patent 5,709,021 (1998).
`S. Miyazaki, S. Kimura, F. Takei, T. Miura, K. Otsuka and Y. Suzuki, Scripta Metall.,
`Vol.17 (1983), p.1057.
`J.H. Lee, H.W. Lee, J.G. Ahn, H.K. Chung and K. Song, Proceedings, International
`Conference on Shape Memory and Superelasticity Technologies, Antwerp, Belgium (1999),
`p.33.
`[15] N.A. Toyama, T. Dickson and B. Moore, Proceedings, International Conference on Shape
`Memory and Superelasticity Technologies, Antwerp, Belgium (1999), p. 215.
`[16] H.C. Lin, K.M. Lin and I.S. Cheng, J. Materials Science, Vol. 36, No. 2 (2001), p.399.
`[17] A. Hirose, M. Uchihara, T. Araki, K. Honda and M. Kondoh, J. Japan Inst. Metals, Vol.54,
`No.3 (1990), p.262.
`T. Haas and A. Schuessler, Proceedings, International Conference on Shape Memory and
`Superelasticity Technologies, Antwerp, Belgium (1999), p. 103.
`[19] A. Ika, J. of Intelligent Material Systems and Structures, Vol.7 (1996), p.646.
`[20] G. Wang, Proceedings, International Conference on Shape Memory and Superelasticity
`Technologies, Pacific Grove, California (1997), p.131.
`P. Hall, Proceedings, International Conference on Shape Memory and Superelasticity
`Technologies, Pacific Grove, California (2000), in press.
`[22] M. Murai, Metallic Eyeglass Frame And Method Of Making The Same, United State Patent
`4,952,044 (1990).
`P. Hall, Proceedings, International Conference on Shape Memory and Superelasticity
`Technologies, Pacific Grove, California (1997), p.125.
`S. Trigwell and G. Selvaduray, Proceedings, International Conference on Shape Memory
`and Superelasticity Technologies, Pacific Grove, California (1997), p. 383.
`
`[18]
`
`[21]
`
`[23]
`
`[24]
`
`7
`
`
`
`[27]
`
`[25] R. Vanugopalan, C. Trepanier and A.R. Pelton, Proceedings, Second International
`Symposium on Advanced Biomaterials (2000), p.77.
`[26] C.G. Rice and P. Burt, Proceedings, International Conference on Shape Memory and
`Superelasticity Technologies, Pacific Grove, California (2000), in press.
`S. Lambardi, L’H. Yahia, J.E. Klemberg-Sapieha, D.L. Piron, A. Selmani, C.H. Rivard and
`G. Drouin, Proceedings, International Conference on Shape Memory and Superelasticity
`Technologies, Pacific Grove, California (1994), p. 221.
`[28] W.A. Johnson, J.A. Domingue and S.H. Reichman, J. de Physique, Coll. C4, Tome 42,
`Supp. 12 (1982), p. C4-285.
`[29] H. Wipf, M. Dietz, D. Aslanisdis, A. Serneels and W. van Mooreleghem, Proceedings,
`International Conference on Shape Memory and Superelasticity Technologies, Antwerp,
`Belgium (1999), p. 375.
`[30] B.S. Murty and S. Ranganathan, International Materials Review, Vol. 43 (1998), p.101.
`[31] D.E. Aslanisdis, Proceedings, International Conference on Shape Memory and
`Superelasticity Technologies, Antwerp, Belgium (1999), p. 384.
`[32] Y. Suzuki, K. Yamauchi, K. Suzuki, K. Tabei, T. Katoh, K. Kusaka and J. Kobayashi,
`Proceedings, International Symposium on Shape Memory Alloys, Guilin, China (1986),
`p.405.
`[33] D. Goldstein, Production of Shaped Parts of Nitinol Alloys by Solid State Sintering, Naval
`Surface Weapons Center Report NSWC TR 84-326 (1986).
`[34] H. Kyogoku and S. Komatsu, J. Japan Society of Powder and Powder Metallurgy, Vol. 46,
`No. 10 (1999), p.1103.
`J.J. Moore and H.C. Yi, Materials Science Forum, Vol. 56-58 (1990), p.637.
`[35]
`[36] H.C. Yi and J.J. Moore, J. Materials Science, Vol.24 (1989), p.3449.
`[37] H.C. Yi and J.J. Moore, J. Materials Science, Vol.24 (1989), p.5067.
`[38]
`S.A. Shabalovskaya, V.I. Itin and V.E. Gyunter, Proceedings, International Conference on
`Shape Memory and Superelasticity Technologies, Pacific Grove, California (1994), p. 7.
`T.W. Duerig, Proceedings, International Conference on Shape Memory and Superelasticity
`Technologies, Pacific Grove, California (1994), p. 31.
`[40] A. Smolders, D. Aslanisdis, A. Serneels, S. Van den Bossche and W. Vandermeulen,
`Proceedings, International Conference on Shape Memory and Superelasticity Technologies,
`Antwerp, Belgium (1999), p.412.
`[41] G. Julien, Post Processing For Nitinol Coated Articles, United State Patent 6,254,458
`(2001).
`[42] A. Sickinger and J.P. Teter, Proceedings, MRS Fall Meeting (1999).
`[43] A.D. Johnson and V.V. Martynov, Proceedings, International Conference on Shape Memory
`and Superelasticity Technologies, Pacific Grove, California (1997), p. 149.
`S. Kajiwara, T. Kikuchi, K. Ogawa, T. Matsunaga and S. Miyszaki, Proceedings,
`International Conference on Shape Memory and Superelasticity Technologies, Pacific
`Grove, California (1997), p. 155.
`S. Miyazaki, K. Nomura and H. Zhirong, Proceedings, International Conference on Shape
`Memory and Superelasticity Technologies, Pacific Grove, California (1994), p. 19.
`L. You, C.Y. Chung, X.D. Han and H.D. Gu, Proceedings, International Conference on
`Shape Memory and Superelasticity Technologies, Pacific Grove, California (1997), p. 189.
`P. Krulevitch, P.B. Ramsey, D.M. Makowiecki, A.P. Lee, G.C. Johnson and M.A. Northrup,
`Proceedings, International Conference on Shape Memory and Superelasticity Technologies,
`Pacific Grove, California (1994), p. 25.
`[48] A. Ishida, A. Takei, M. Sato and S. Miyazaki, Proceedings, MRS Symposium, Vol. 360
`(1995), p.381.
`
`[39]
`
`[44]
`
`[45]
`
`[46]
`
`[47]
`
`8
`
`