Mechanical Properties, Microstructure, and Actuation Behavior of Wire Arc Additive Manufactured Nitinol: Titanium Bimetallic Structures

Abstract

Nitinol (NiTi) is well known for its corrosion resistance, shape memory effect, superelasticity, and biocompatibility, whereas Titanium (Ti) is well known for its high specific strength, corrosion resistance, and biocompatibility. The bimetallic joint of NiTi and Ti is required for applications that require tailored properties at different locations within the same component, as well as to increase design flexibility while reducing material costs. However, because of the formation of brittle intermetallic phases, connecting NiTi and Ti is difficult. In the present study, a systematic experimental investigation is carried out to develop NiTi–Ti bimetallic joint using wire arc additive manufacturing (WAAM) for the first time and to evaluate its microstructure, mechanical properties, martensitic transformation, and actuation behavior in the as-built condition. The defect-free joint is obtained through WAAM and microstructural studies indicate the formation of intermetallics at the NiTi–Ti interface leading to higher microhardness values (600 HV). Shape recovery behavior and phase transformation temperature were also enhanced in comparison to NiTi. An improved actuation and bending angle recovery is observed in comparison with NiTi. The present study lays the way for the use of WAAM in the construction of NiTi and Ti bimetallic structures for engineering and medicinal applications.

Introduction

Bimetallic components involve joining two different materials to obtain tailored properties at different locations of the same component. These components allow for the optimization of overall performance while avoiding the restrictions of each other.¹ Casting, explosive welding, diffusion bonding, powder metallurgy, and other methods are used to fabricate bimetallic structures.²

The bimetallic combination of Nitinol (NiTi) and Titanium (Ti) is one of the potential multimaterial combinations. Because of its functional features, such as superelasticity and shape memory effect (SME), NiTi alloy is the most widely used shape memory alloy (SMA). Although medical devices are the most common use for NiTi, the aerospace and energy sectors are providing new challenges for SMAs in demanding applications requiring high temperatures and hostile environments.^3,4

Several research initiatives aimed at joining NiTi alloys have recently been carried out, with successful findings reported by several research groups.^5,6–8 These properties attract their deployment in several sectors ranging from medical, aerospace, automotive, etc. Joining NiTi and Ti is significant for a broad range of applications due to Ti's high specific strength, corrosion tolerance, and biocompatibility and NiTi's shape memory characteristics, superelasticity, superior corrosion resistance, and biocompatibility.⁹ However, each material has different thermophysical properties, resulting in different heat transfer rates.^6–8,10,11

The primary issue of dissimilar welding of NiTi and Ti-based alloys is Ti's high-temperature reactivity, which produces brittle intermetallics, such as NiTi₂, NiTi, and Ni₃Ti, in the weld metal during dissimilar welding of NiTi and Ti-based alloys.⁶ The formation of brittle intermetallic phases causes the joint to become brittle, and the thermal expansion mismatch associated with dissimilar welding causes transverse cracks in the brittle weld metal, resulting in the loss of mechanical property. Literature indicates that attempts have been made by researchers to join NiTi and Ti using laser welding.¹²

Miranda et al. investigated the use of laser welding to combine NiTi and Ti-6Al-4V plates in a butt joint configuration. The weld joint got cracked as it cooled; therefore, a structural analysis of the fracture surface was performed to examine the cracking phenomena to develop successful welding approaches.⁵ Without the presence of an interlayer material that regulates the chemical composition of the molten metal, preventing Ni migration and the formation of brittle intermetallic, this cracking phenomenon is impossible to prevent.⁶ NiTi and Ti-6Al-4V plates were joined by explosion welding, and a good joint was found without any cracks. However, this technique requires a special firing field and explosive that is not convenient for mass production.² Moreover, only plate samples can be joined by explosive welding, which does not allow the ability to produce objects with complex shapes.

Furthermore, only plate samples may be linked by explosive welding, limiting the potential to create things with complicated forms. Due to inherent limitations, the processes normally used for fabricating multimaterial components cannot be used to generate complex-shaped components; thus, additive manufacturing (AM) is being investigated globally for such applications.¹³ AM is one of the advanced manufacturing processes that can build complex-shaped engineering components. Wire arc additive manufacturing (WAAM) is a large-scale metal AM process that uses arc welding to melt wire feedstock and deposit the material layer by layer to build a complex 3D structure. Compared with typical metal AM techniques, WAAM has a lot of benefits, including low cost, high deposition rate, design flexibiltiy, and high catchment efficiency.¹⁴

WAAM is employed in sectors, such as aircraft, marine, automobiles, and architecture, due to the obvious benefits mentioned above. It can handle a broad range of metals since they consume source materials in wire form, allowing them to deposit welding wires. Complex-shaped NiTi and Ti components are difficult to fabricate using conventional manufacturing processes such as turning, milling, forging, and casting, which limits the potential applications of these smart functional materials.^15,16 Fabrication of NiTi components by WAAM has been proven to be successful in manufacturing parts based on these materials with reduced overall costs and high deposition rates when compared with several frequently used fusion-based AM processes, such as Selective Laser Melting and Electron Beam Melting.^17,18

To the best of the author's knowledge, no work on the fabrication of multilayered NiTi and Ti bimetallic structures using WAAM has been reported. As a result, the current work employs WAAM to fabricate a defect-free bimetallic structure of NiTi and Ti, as well as explain its microstructure, martensitic transition, mechanical properties, and actuation behavior.

Materials and Methods

NiTi (Ni50.9Ti49.1) and Ti with a wire diameter of 1.2 mm were used as feedstock materials in this investigation. A Ti block with 100 × 100 × 10 mm served as the substrate. WAAM system based on Gas Metal Arc Welding was used to manufacture bimetallic structures of NiTi and Ti (Fig. 1a). A wire feeder coupled to the X-Y stage was used to feed the feedstock wire to the welding torch. An electrode on the X-Y stage was allowed to move in a different direction. The X-Y controller associated with Repitier host software was programmed using G and M code. The process parameters were optimized for continuous and uniform deposition of NiTi and Ti separately. Table 1 shows the optimized parameter used for depositing bi-metallic wall structures. Before the deposition, the substrate was preheated to a temperature of 400°C to increase bonding between the two dissimilar materials.

FIG. 1.

(a) WAAM setup. (b) Electrical actuation set-up and (c) scheme of sample location. CNC, Computer Numerically Controlled; WAAM, wire arc additive manufacturing.

Table 1.

Wire Arc Additive Manufacturing Parameters Used for Bimetallic Joint

Materials	Wire feed rate (m/min)	Argon gas flow rate (L/min)	Stand-off distance (mm)	Voltage (V)
Ti	5.8	15	15	16.5
NiTi	5

NiTi, Nitinol.

The deposited samples were removed from the substrate by Computer Numerically Controlled wire electrical discharge machining. The standard metallographic technique was used to polish the samples, and Kroll's reagent solution was used to etch them (H₂O:HNO₃:HF: 5:4:1).Optical Microscope (Make & model: LOMO “Metam 31-LV”) was used to capture the panoramic view of the fabricated joint in the macro scale. An X-ray diffractometer (BRUKER-D8 Advance) with a step size of 0.02° and a dwell time of 0.5 s was used for phase analysis. A scanning electron microscope (SEM) with energy-dispersive spectroscopy (EDS) (Make & Model: S-4800 Hitachi; Carl Zeiss “Merlin™”) was used to perform microstructural and compositional studies.

A Vickers microhardness tester (Make and Model: WlterUhl-VMHT002) with a load of 1.96 N and a dwell time of 10 s was used to assess the microhardness of the samples. Small samples (3 × 3 mm) were cut from the NiTi side's edge, center, and joint regions, and the martensitic transformations were investigated using differential scanning (Make and Model: Mettler Toledo 822^e).

Samples were heated and cooled within the temperature range of 100–110°C at a rate of 10°C/min. The compression test was performed on the sample, which is shown in Figure 5b (length of 6 mm and a width of 3 mm) using the Universal Testing Machine (Make and Model: WlterUhl-VMHT002).

FIG. 5.

Mechanical behavior of bimetallic joint. (a) Microhardness. (b) Compression strength. (c) Fractography.

A set-up based on Joule heating (electrical actuation) consisted of a laser displacement sensor, a monitor, a data collection device (Agilent DAQ 34790A), a thermal actuator, and a programmable power supply (RIGOL DP1308) was used to examine the strain fluctuation in NiTi–Ti samples, as illustrated in Figure 1b. Figure 1c depicts the region of the samples cut for actuation experiments that were extracted from the NiTi bottom layer near the interface, and it reveals Ti-rich phases. One sample with a cross-section of 1 × 1 mm² was cut in the vicinity of the NiTi and Ti joint (is labeled as NiTi–Ti sample), where the NiTi and Ti layers had a thickness of 0.5 mm.

The other sample was taken far away from the junction on the NiTi side and had the same cross-section (labeled as NiTi sample). In the ice bath, both materials were bent around a mandrel with a radius of 2 mm (in the martensite state).

This results in a 25% bending deformation of the samples. After unloading, both samples were characterized by the same residual strain. Then, at room temperature, one side of the sample was fixed, and the other was connected to a load of 35 g (Fig. 1b). The Arduino relay circuit was used to control the heating and cooling of the samples for a 15-s duty cycle. The actuation response and dynamic displacement characteristics were investigated during the real-time experiment by applying a series of potentials with varying voltage and current with frequency to get their actuation response. A voltage of 5 V and a current of 4 A was used in the entire actuation study. Laboratory view was connected to the power supply. The samples were also heated on a hot plate to investigate the variation in bending angle due to heating.

Results and Discussion

Figure 2a and b depicts a cross-section of the WAAM-built bimetallic structure, demonstrating that the structure is defect free on macroscale, with layer-by-layer morphology visible throughout the structure. Some of the previously deposited layers was found to be remelted during multilayer deposition, resulting in visibly apparent remelted zones between the layers. Remelting the prior layers improved the bonding between the layers. Figure 2d shows the compositional analysis carried out on the WAAM-deposited bimetallic joint interface. Ti-rich zones can be seen from the elemental maps obtained along the bimetallic joint. The deposition moves from 100% Ti at the Ti-rich bottom layers to a zone with a gradual reduction in Ti composition from Ti side to NiTi side.

FIG. 2.

Bimetallic joint. (a, b) Macrostructure, (c) elemental mapping, and (d) compositional analysis. NiTi, Nitinol; Ti, Titanium.

Furthermore, the presence of oxygen is detected during the analysis, with a higher percentage at the Ti-rich zone, which indicates the formation of Ti oxides during the deposition in the Ti side.¹³

Figure 3 presents the microstructure and X-ray diffraction pattern obtained at different locations of NiTi–Ti bimetallic structure. Dendrite structure was formed at the interface, which showed Ti-rich intermetallic compounds as confirmed by EDS results. Two intermediate zones near the joining direction were found on the NiTi side. The first zone was with the Ti layer, and its width was about 100–150 μm. This zone consisted of pure Ti (marked by “A”), Ti solid solution with Ti concentration of 90 at. % (marked by “B”) and Ti₂Ni precipitates (marked by “C”). The second zone (Ref. Fig. 3a) consisted of the NiTi₂ phase and Ti-rich NiTi individual grains (marked by “D”). The grains of Ni_49.5Ti_50.5 (marked by “D”) were surrounded by the NiTi₂ phase, which was far away from the joint. The X-ray diffraction patterns (Fig. 3b, c, e) confirmed the structure of the NiTi–Ti composites in various locations.

FIG. 3.

SEM images (a, c) and X-ray patterns (b, d, e) found in the vicinity of NiTi, joint, and Ti side. SEM, scanning electron microscope.

Furthermore, X-ray diffraction studies reveal that at room temperature, the NiTi phase is in the austenite state, which is consistent with the chemical composition of this phase.

The Ti-rich NiTi phase demonstrates the martensitic transformation at high temperatures and should be in the martensite state at room temperature.¹⁹ WAAM-deposited Ti layers exhibited mainly serrated α phase with limited amounts of intergranular β phase in some areas (Fig. 3d). This microstructure is generally uniform, with a Widmanstätten α phase morphology and small layers of preserved β phase between the lath boundaries. This can be validated by the X-Ray Diffraction (XRD) pattern of Ti, which is usually a mixture of α Ti and β Ti with a hexagonal lattice structure (Fig. 3e).

Due to their strong affinity and the occurrence of chemical reactions inside the melt pool, Ti and Ni may easily produce NiTi₂ and Ni₃Ti intermetallic phases. The mechanical properties of the joint can be adversely affected by these brittle intermetallic phases. NiTi has a lower melting point than Ti, it participates more in joint formation, resulting in more intermetallic compounds being formed on the NiTi side.

Figure 4 shows the calorimetric curves found on heating and cooling of NiTi–Ti composite. A heat release peak has been found on cooling and attributed to B2→B19′ transformation. The reverse B19′→B2 transformation causes the heat absorption peak when heated. On calorimetric cures for cooling and heating, certain low-intensity heat flow peaks may be seen (shown by circle in Fig. 4), which indicate that the B2 ↔ B19′ martensitic transformation occurs at various temperatures due to some deviation of the chemical composition of the NiTi phase. Transformation temperatures were determined according to ASTM F2004-05R10 as intersections of tangent lines, and transformation enthalpy was found as square under a peak (accuracy is ±1°C for transformation temperatures and ±1 J/g for enthalpy).

FIG. 4.

Calorimetry curves obtained on cooling (a) and heating (b) of samples cut in various locations of the NiTi layers in NiTi–Ti (a, b) bimetallic composites.

The transformation temperatures, as well as enthalpy did not depend on the location of the samples in the NiTi–Ti composite, except the sample cut from the joint area (Table 2). The transformation enthalpy at the joint area was found to be lower than in samples taken farther away from the joint because this sample had a pure Ti layer as well as the various phases mentioned above in the intermediate zone. Only the NiTi phase underwent the martensitic transformation and contributed to the transformation enthalpy at the same time.

Table 2.

Temperatures and Enthalpy of the Martensitic Transformation in the NiTi Layers of NiTi–Ti Bimetallic Composites

Material	Sample location	Transformation	Ms, °C	Mf, °C	As, °C	Af, °C	Efw, J/g	Erev, J/g
NiTi–Ti	Top	B2 ↔ B19′	80	37	70	109	29	28
	Middle	B2 ↔ B19′	77	37	70	108	29	30
	Joint	B2 ↔ B19′	71	37	70	102	15	14

Ti, Titanium.

In Table 2, M_s and M_f refer to start and finish temperatures of the forward transformation on cooling, A_s and A_f refer to start and finish temperatures of the reverse transformation on heating, and E_fw and E_rev refer to enthalpies of the forward or reverse transformations.

Figure 5a presents the variation in the microhardness in the NiTi–Ti sample, and it was observed that the microhardness of the NiTi zone varies from 250 to 400 HV. The microhardness values at the interface joint are more than that of base metals, indicating the formation of hard and brittle phases. The presence of NiTi₂ confirmed by XRD studies results in higher hardness of weld metal within the interface. The presence of impurities, particularly oxygen, has an effect on the hardness of Ti. Furthermore, because to the presence of distinct components along grain boundaries in microstructures, the hardness of WAAM-generated Ti is primarily regulated by the solid solution and grain boundaries.²⁰ Grain boundaries and dislocations are more abundant in alloys with a faster cooling cycle in the bottom region, leading to enhanced microhardness. In the composition range studied, the strength and hardness of binary Ti-O, Ti-N, and Ti-C alloys are linear functions of alloy concentration.

The microstructure toward the top of the WAAM thin-wall components, on the other hand, includes a lot of martensite structure, which is often hard and more intense.

The hardness of WAAM-built NiTi and Ti layers is greater than that of conventionally manufactured NiTi and Ti layers (296 HV for NiTi and 170 HV for Ti). In addition, the microhardness at the NiTi–Ti joint interface was also higher than the hardness obtained at the interface of NiTi and Ti (550 HV) obtained by laser welding.¹³

Figure 5b shows the stress–strain curve of NiTi–Ti bimetallic under compressive loading. The compressive strength of WAAM-deposited joint was found to be 750 MPa. Dissimilar joints reveal lower horizontal stress plateaus than NiTi's stress plateau. Because the two layers (TiNi and Ti) are linked in bimetallic composites, it is impossible to deform the TiNi layer independently.

As a result, both layers deform concurrently during the preliminary compression. The formation of brittle phases, according to XRD data, may be responsible for the reduced compressive strength and elongation compared with base metals. On both sides, failure occurs through the fusion zone dendrites. In the Ti₂Ni granular regions, brittle transgranular cracking with a smooth, almost bright surface due to interdendritic precipitation of Ti₂Ni in the fusion zone, an intergranular failure occurs through the dendrites of the primary solidification phases beta-Ti and NiTi, resulting in a dimple fracture surface morphology Figure 5c.

The displacement versus time curves obtained in real-time experiments is shown in Figure 6a, where zero displacements at t = 0 correspond to the location of the nonfixed predeformed sample at room temperature. From the figure, it can be seen that a displacement of 2.2 mm was found for the NiTi–Ti sample, whereas a displacement of 1.5 mm was obtained for the NiTi sample. A Martensite detwinning and plastic strain occurred in the samples after preliminary deformation to 25% in the martensite stage. The two-way shape memory effect (TWSME) is usually generated as plastic strain causes the formation of oriented internal stress. During Joule heating, the strain recovers because of the reverse transformation from detwinned martensite to austenite (Fig. 6b, c).

FIG. 6.

Actuation studies of NiTi and NiTi–Ti joint (a) shape variation graph for electrical actuation (b) schematic for hot plate actuation and bending angle versus temperature graph, and (c) sample bending image during actuation studies.

Owing to the generation of the internal oriented stress during the previous deformation, strain rises during the forward change from austenite to detwinned martensite and during subsequent cooling.

The load of 35 g that was applied to the sample was too small to affect the TWSME. Thus, the NiTi sample demonstrated the TWSME that is typical for the SMA. The strain variation in NiTi–Ti samples is typical for the NiTi-based bimetallic composited produced by explosion welding.²¹ The martensite reorientation and plastic deformation in the NiTi layer occurs during preliminary deformation of the NiTi–Ti sample (bending around mandrel), whereas elastic and plastic deformation occurs in the Ti layer.

On subsequent heating, the strain recovery occurred in the NiTi layer, leading to the deformation of the Ti layer (as in counter body in actuators), and resulted in internal stress appearing in the NiTi–Ti sample. Following cooling under this stress, the detwinned martensite form increased strain and relaxation of tension. The stress generated during Ti layer deformation added up to the internal stress in the NiTi layer and turned more than the internal stress caused in the NiTi sample by plastic deformation.

As a result, the strain variation in the NiTi–Ti sample during cooling and heating was found to be greater than in the NiTi sample.

This is illustrated in Figure 6b, which shows the bending angle's relation to sample temperature. It can be seen in Figure 6b that at all temperatures, the bending angle in the NiTi–Ti sample is 5°C larger than in the NiTi sample, excluding a temperature of 140°C at which this difference is 15°C. The samples were subjected to a series of cycles in which they were heated by the current for a few seconds before being cooled to test the stability of the recoverable strain variation (Fig. 6a). Both samples demonstrated the stable strain variation and the difference in strain between NiTi–Ti and NiTi samples kept during thermal cycling. Thus, the article results show that the NiTi–Ti bimetallic sample produced by WAAM demonstrates a good functional behavior as observed in NiTi-based bimetallic composites fabricated by other techniques.^22–24

Conclusions

WAAM was used to produce a defect-free NiTi–Ti joint, and elemental mapping revealed the compositional transition from NiTi to Ti. According to X-ray diffraction studies, the NiTi phase was austenite at room temperature, and the martensitic transition was observable at high temperatures. Brittle intermetallic compounds formed at the interface resulted in increased hardness and brittle failure. NiTi–Ti sample demonstrated better shape recovery behavior in comparison to NiTi sample due to the Ti layer inducing an additional stress. This current research work opens the way for complex NiTi–Ti bimetallic structures to be built for several engineering and medical applications.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research is funded by DST-RSF collaboration (RSF No. 19-49-02014, DST No. DST/INT/RUS/RSF/P-36). The X-ray, SEM, and EDS tests were conducted utilizing equipment from the Saint Petersburg State University. This work is supported by joint DST-RSF project (RSF#19-49-02014, DST #DST/INT/RUS/RSF/P-36).