Review of EDM-Based Machining of Nickel–Titanium Shape Memory Alloys

Abstract

Shape memory alloy (SMA) materials are valued for their shape memory effect, superelasticity, and biocompatibility, making them an ideal choice for applications in biomedical, aerospace, and actuator fields. Nickel–titanium (NiTi) SMA is a promising biomedical material. It is widely used in the manufacture of biomedical instruments, devices, implants, and surgical tools. However, its complex thermo-mechanical behavior and poor machinability pose challenges for conventional machining. To manufacture high-quality nitinol parts, traditional machining processes are being replaced by advanced machining technologies. Electric discharge machining (EDM) is an advanced machining technique whose mechanism of material removal involves erosion caused by plasma formation and spark generation. It has proven effective for processing difficult-to-machine materials. This review summarizes EDM and its variants, including hybrid EDM, with a focus on machining NiTi-SMA materials for biomedical, aerospace, microelectromechanical systems, and automotive applications, and systematically explores key factors such as process parameters, material removal mechanisms, surface integrity, tool wear, and optimization strategies. This review begins with an introduction to nitinol (i.e., NiTi-SMA) and its variants, followed by an in-depth discussion of plasma formation, spark generation mechanisms, and other key aspects of EDM. It then provides a detailed analysis of notable past research on the machining of NiTi SMA materials using EDM and its variants. This paper concludes with insights into future research directions, aiming to advance EDM-based machining of SMA materials and serve as a valuable resource for researchers and engineers in the field.

1. Introduction

1.1. Introduction to Shape Memory Alloys (SMAs)

Continuous research and innovation on developing new materials and processing technologies are essential to fulfil emerging product requirements, especially in the biomedical, microelectromechanical systems (MEMS), aerospace, and automotive fields [1,2]. SMA materials, which belong to the group of non-ferrous metallic materials, due to their unique and superior properties, are important materials for such application fields. These alloys are intermetallic compounds with superlattice structures, combining the properties of metals and ceramics. Also known as memory metals, smart alloys, or muscle wires, they represent a unique class of advanced smart materials [3,4,5]. These alloys exhibit two distinctive properties: (i) the shape memory effect (SME), where deformation at low temperatures in the martensitic phase is reversed upon heating due to a phase transformation to austenite, and (ii) superelasticity (or pseudoelasticity), where SMA materials subjected to large strains at temperatures slightly above their transformation point can instantly recover their original shape without heating [6,7,8]. This reversible shape change in response to temperature is known as the memory effect. SMA materials exhibiting shape recovery only upon heating are classified as having a one-way shape memory effect, capable of up to ~10% recoverable strain. In contrast, the two-way shape memory effect enables the material to remember and switch between shapes during both heating and cooling cycles, though with a typically lower recoverable strain of around 3% [9,10,11]. Common SMA materials are (i) nickel–titanium (NiTi) SMA, (ii) copper-based SMA, (iii) gold–cadmium (Au-Cd)-based SMA, and (iv) iron-based SMA. NiTi-SMA materials, commonly known as nitinol, are notable for their high elasticity, excellent fatigue resistance, and biocompatibility, making them widely used in medical devices [11,12,13]. Major applications include stents, guidewires, orthodontic arch wires, and actuators [14]. NiTi-SMA is a very expensive and important material for biomedical products, namely, stents, implants, and surgical tools. NiTi-SMA is made of 55% nickel and 45% titanium. Its shape memory behavior, coupled with excellent corrosion resistance, superior biocompatibility, and superelasticity, has been found very effective for its widespread applications. NiTiCu, NiTiPd, NiTiFe, NiTiNb, NiFeGa, and NiTiCo are some of the types of NiTi-SMA materials [15]. Copper-based SMA materials include copper–zinc–aluminum (Cu-Zn-Al) and copper–aluminum–nickel (Cu-Al-Ni). Cu-Zn-Al is less expensive than NiTi but has lower fatigue strength and is commonly used in actuators and switches. Cu-Al-Ni exhibits higher transformation temperatures and increased brittleness compared to NiTi, which makes it suitable for aerospace and industrial actuator applications [16]. Iron-based SMA materials include iron–manganese–silicon (Fe-Mn-Si) and iron–nickel–cobalt–titanium (Fe-Ni-Co-Ti) systems. Fe-Mn-Si alloys are less expensive and exhibit lower performance in cyclic applications, making them suitable for use in civil engineering structures and seismic dampers. In contrast, Fe-Ni-Co-Ti alloys are utilized in high-temperature SMA applications due to their superior performance under thermal and mechanical stress [17,18]. Gold–cadmium (Au-Cd)-based SMA materials exhibit excellent high-temperature performance but are seldom used because of their high cost and toxic nature. Other less common SMA materials include Ni-Mn-Ga, Ni-Fe-Ga, and Ti-Nb. Ni-Mn-Ga and Ni-Fe-Ga are magnetic SMA materials actuated by magnetic fields, while Ti-Nb is used in biomedical applications due to its low elastic modulus and excellent biocompatibility [19].

Shape memory effect, superelasticity, biocompatibility (particularly in nitinol), corrosion resistance, thermal conductivity, electrical resistivity, and fatigue resistance are key characteristics of SMAs [15,20]. These characteristics make SMAs highly suitable for a wide range of applications [11,12,13,20], including

• Medical devices: stents (especially nitinol), orthodontic archwires, surgical tools, and guidewires;
• Aerospace: actuators for morphing structures, temperature-activated couplings, and fasteners;
• Automotive systems: temperature control actuators, crash sensors, and variable-geometry engine components;
• Robotics and automation: micro-actuators, soft robotics, and artificial muscles;
• Consumer electronics: shape-retaining eyeglass frames and mobile device components (e.g., lens actuators);
• Civil engineering: seismic dampers in buildings and self-healing structures.

The key advantages of SMAs lie in their ability to deliver compact, silent actuation without moving parts, offering lightweight alternatives to traditional motors and gears, which is ideal for aerospace and portable systems [21]. Their biocompatibility makes certain SMA materials especially valuable in medical devices, while their excellent fatigue and corrosion resistance ensure long-term durability. Moreover, as multifunctional materials, SMAs uniquely integrate sensing, actuation, and structural functions, enabling innovative solutions across diverse industries [22,23].

1.2. Machining Challenges of SMA (NiTi Alloy)

SMAs can be manufactured using both traditional methods (e.g., micro-milling and micro-grinding) and non-traditional machining techniques, such as abrasive water jet machining (AWJM), EDM, laser beam machining (LBM), electrochemical machining (ECM), and electrochemical polishing (ECP) [24,25]. However, machining SMA materials with traditional techniques is challenging due to their unique properties, such as high strength, low thermal conductivity, and superelasticity. These characteristics lead to rapid tool wear, poor surface finish, and difficulties in chip formation. Moreover, NiTi alloys often undergo phase transformations caused by machining-induced heat and stress, which can alter their microstructure and mechanical properties, complicating precise machining. To address these challenges and achieve the desired dimensional accuracy and surface quality, it is crucial to control cutting temperatures, select appropriate tool materials, utilize coated tools, and optimize machining parameters. The limitations and specific challenges associated with the traditional machining of SMA materials have driven researchers, engineers, and scientists to investigate advanced and hybrid machining processes that can provide enhanced dimensional accuracy and surface integrity. Among these, spark erosion-based machining has emerged as one of the most important and widely adopted advanced techniques, demonstrating significant potential for effectively machining SMA materials.

1.3. Introduction of EDM Processes

1.3.1. Fundamentals and Working Principles of EDM-Based Machining Processes

EDM is an advanced, non-contact machining process in which the tool electrode and the workpiece are separated by a dielectric fluid, and material is removed from the electrically conductive workpiece through a series of rapid, recurring electrical discharges (sparks) that occur across the inter-electrode gap (IEG) [26]. The high-energy electrical sparks generate intense localized heat—approximately 12,000 °C—which causes the material to melt and vaporize, resulting in its removal [27]. The IEG refers to the precise gap maintained between the conductive workpiece and the tool electrode, within which the key phases of EDM—such as dielectric breakdown, plasma formation, spark generation, material erosion, and flushing of eroded particles—occur. Maintaining a minimum IEG (i.e., 0.025 mm) is essential to prevent short-circuiting caused by direct contact between the workpiece and tool electrodes and to ensure stable and efficient machining [28]. Excess material from the workpiece is removed using thermoelectric erosion in this process [29]. Therefore, the mechanism of material removal in EDM is the melting and evaporation of material due to the generation of frequent sparks in minute intervals between the workpiece and tool materials in the presence of circulating dielectric fluid, such as hydrocarbon oil or deionized water. The dielectric fluid serves as an insulator until it is ionized by a specific spark gap and voltage, allowing a spark to travel to the workpiece. Generally, the workpiece and the tool act as electrodes made from electrically conductive materials, referred to as the anode (positive pole) and cathode (negative pole), respectively [30]. Copper, graphite, brass, tungsten, and molybdenum are commonly used electrode materials [31]. EDM is ideal for forming complex and intricate shapes and is capable of machining extremely hard materials. However, the major limitation is that both workpiece and tool materials should be electrically conductive. As shown in Figure 1, EDM utilizes a direct current (DC) supply for machining by connecting the tool electrode as the cathode and the workpiece as the anode. The electrodes are partially or completely immersed in the dielectric fluid and separated by a spark gap, also referred to as an interelectrode gap. A servo motor is used to maintain the constant gap between electrodes. Individual stepper motors are used for the movement of the worktable in the X-, Y-, and Z-axes. In this process, the tool electrode is stationary and can move up and down directions along the Z-axis only, whereas the workpiece is mounted on the worktable with the help of clamps and can move along the X- and Y-axes. When extreme electro-thermal heat is generated in the interelectrode gap (IEG), known as the spark gap zone, due to the occurrence of repetitive sparks, the workpiece surface melts and vaporizes, and erosion of the workpiece surface takes place [30,32,33].

Plain/conventional EDM is commonly known as die-sinking EDM. There are several allied or variant processes that use the same fundamental principle, controlled erosion of electrically conductive materials through electrical discharges (sparks), including wire electric discharge machining (WEDM), electric discharge drilling (EDD), electrical discharge turning (EDT), wire electrical discharge turning (WEDT), and their micro-scale versions, such as µ-EDM, µ-WEDM, and µ-EDD. The micro-scale version of EDM refers to miniaturized variants of EDM processes designed for high-precision machining of micro-features and microparts. These techniques are widely applied in MEMS, biomedical devices, and precision tooling. They utilize tiny tool electrodes or fine wires (typically <100 µm in µ-WEDM) and are well-suited for machining hard-to-machine materials such as NiTi, titanium, and tungsten carbide. µ-EDM is used for generating micro-cavities and complex features, µ-WEDM for cutting intricate micro-shapes, and µ-EDD for drilling micro-holes with high aspect ratios. Other variants include hybrid and sustainable spark-erosion-based processes. Hybrid processes integrate spark-erosion-based processes with other machining techniques to enhance overall performance. Prominent examples include ultrasonic-assisted electrical discharge machining (US-EDM), electrochemical discharge machining (ECDM), and laser-assisted electrical discharge machining (L-EDM), which combine multiple physical mechanisms to improve machining efficiency and precision. Sustainable techniques, including dry-EDM, dry-WEDM, powder-mixed electrical discharge machining (PMEDM), and cryogenic-assisted electrical discharge machining, aim to improve process efficiency while minimizing environmental impact [26,34]. In PMEDM, conductive powders such as aluminum, graphite, or silicon carbide are added to the dielectric fluid to improve surface finish, enhance the material removal rate (MRR), and minimize tool wear. In contrast, dry-EDM and dry-WEDM utilize a gaseous dielectric medium—typically, compressed air, nitrogen, or argon—instead of conventional liquid dielectrics, such as kerosene-based oils (used in EDM) or deionized water (used in WEDM). While this substitution offers significant environmental and operational advantages, it also introduces technical challenges, such as reduced dielectric strength, lower cooling efficiency, and difficulties in debris removal, all of which require careful process optimization. Figure 1 illustrates the working principles of die-sinking EDM and WEDM processes. As shown in Figure 1a, the die-sinking or plain EDM process utilizes a fabricated tool electrode to produce a cavity in the workpiece that is the opposite of the shape and size of the fabricated tool electrode. Each electrode can form only one type of cavity in the workpiece. Therefore, different tool electrodes are required for machining different shapes and sizes in the workpiece. The electrodes are typically made of copper or graphite. Hydrocarbon oil is used as a dielectric fluid.

WEDM is similar to EDM except for the use of a fine wire (100–250 µm diameter) as the tool electrode and deionized water as the dielectric fluid (Figure 1b). The same wire is used for machining any kind of complex and intricate parts. Brass wire is commonly used for machining in WEDM. The stratified wire or diffused wire is also used to improve the machining speed. It is a zinc-coated wire that has copper as the base material. WEDM can fabricate any kind of complex parts and components. In another variant, i.e., electric discharge drilling, a hollow cylindrical tool electrode is used for drilling. Different sizes of tool electrodes are required for drilling different sizes of holes in the workpiece. Dielectric fluid also flowing through the tool electrodes during drilling. Brass and copper cylindrical tool electrodes are commonly used in electric discharge drilling.

1.3.2. Material Removal Mechanism of EDM-Based Processes

EDM consists of the following components: a power supply system (including a DC pulse generator, voltmeter, and ammeter), a tool holder, a dielectric supply system (including a pump and filter), a servo mechanism, a control unit (to manage operations by adjusting input parameters), and a worktable. The tool and workpiece electrodes are positioned to avoid contact while maintaining a set distance called the spark gap or interelectrode gap (IEG), through which dielectric fluid flows to initiate sparks. A servo mechanism maintains this gap during machining. The electrodes are fully or partially submerged in a tank with continuously flowing dielectric fluid, which fills the spark gap. The principle of material removal in EDM is thermoelectric erosion, occurring without physical contact between electrodes. A DC generator produces pulses that create sparks between the tool electrode and workpiece, melting material, also known as debris, which is then flushed away from the spark gap by dielectric fluid. During erosion, proper flushing is necessary to prevent debris accumulation in the spark gap, which can cause short-circuiting, and to avoid the redeposition of a thin, hard layer of eroded particles (i.e., also known as a recast layer) on the machined surface that contains voids and microcracks, thereby deteriorating the surface quality and integrity. Figure 2 shows the sequential key phases, such as dielectric breakdown, plasma formation, spark generation, material erosion, and flushing of eroded particles after machining by EDM.

Preparation phase: DC power is applied across the inter-electrode gap (IEG), and an intense electric field develops at the narrowest point between spark gaps. Microscopic contaminants suspended in the deionized fluid are attracted toward the developed intense electric field at the narrowest point and accumulate at the strongest point of the electric field.

Plasma formation: A highly conductive bridge across the spark gap is built by the accumulated contaminants. As the applied voltage increases, the material of the conductive bridge heats up. A plasma channel between the tool electrode and the workpiece is formed due to ionization.

Discharge phase: A spark occurs due to the increased temperature and pressure of the plasma channel in the narrowest gap between the tool electrode and the workpiece. This extends to other portions of the gap, and material from the workpiece starts eroding, melting, and vaporizing. Vaporization creates gaseous by-products, and gas bubbles are formed and attempt to expand outward from the plasma channel.

Interval phase: Once the pulse ends, i.e., during the spark duration, sparks and heating actions stop, causing the collapse of the plasma channel. Then, continuously flowing dielectric fluid flushes the debris eroded from the tool and workpiece from the spark gap. The EDM residue consists of tiny, solidified material balls and gas bubbles. This cycle repeats until the machining is completed [35,36,37].

1.3.3. Key Variable Parameters and Their Functional Roles in EDM-Based Processes

EDM-based processes are complex and influenced by numerous variable parameters. Key electrical parameters include the spark duration, spark-off time, peak current, servo gap voltage, discharge voltage, and open-gap voltage. Common non-electrical parameters include flushing pressure, electrode feed rate, electrode material, and dielectric fluid type. In wire-assisted EDM (WEDM), additional factors such as wire feed rate and wire tension are also important [32,33,38].

Figure 3 shows the variable parameters of EDM-based processes. Spark duration and spark-off duration are the most influential parameters controlling the timing of the spark in EDM. The servo gap voltage (also called the working gap or inter-electrode gap voltage) is critical for maintaining the proper distance between the electrode and workpiece, ensuring stable spark generation. The discharge voltage, which is the actual voltage, is lower than the open-circuit voltage due to dielectric breakdown that allows current flow. The open-gap voltage is the maximum voltage before spark initiation, used to ionize the dielectric and enable discharge. The peak current indicates the maximum current during spark discharge. Higher peak current increases spark energy, speeding up material removal but causing rougher surfaces and more tool wear. Lower peak current produces smoother finishes but slower machining. The flushing pressure forcefully delivers dielectric fluid into the spark gap to remove debris and stabilize sparking, significantly affecting surface quality, tool wear, and machining efficiency. The electrode feed rate is the speed of the electrode advancement toward the workpiece. The electrode material—commonly copper, graphite, or copper–tungsten—affects tool wear, machining speed, and surface finish depending on the application. The dielectric fluid serves as an insulator, coolant, and flushing medium; hydrocarbon oils are typically used for die sinking, while deionized water is common for wire-assisted EDM, selected based on machining and environmental requirements [32].

In wire-assisted EDM, the wire feed rate controls the supply of fresh wire, and wire tension maintains wire straightness and stability, both essential for accuracy, surface finish, and preventing wire breakage [33].

1.3.4. Benefits and Applications of EDM-Based Processes for SMAs

EDM offers several advantages, including [32,33] (i) the capability to machine a nitinol-type hard and tough material [39]; (ii) SMA strip-type fragile and thin workpieces; (iii) precise machining of complex profiles; (iv) absence of mechanical stresses or distortion during machining; (v) minimal thermal damage and reduced heat-affected zones; (vi) improved surface finish and dimensional accuracy with tight tolerances; (vii) reduction in the need for secondary machining processes; (viii) minimized tooling and fixture requirements; (ix) enhanced control over processing parameters; and (x) the possibility of unattended machining for extended periods.

EDM and its variants are widely used for machining SMAs to produce parts and components for a variety of applications, as illustrated in Figure 4. This includes [30] (i) bioengineering, including stents, medical implants, and high-precision surgical tools; (ii) microelectromechanical systems (MEMSs) and microsystem applications, including micro and miniature actuators; (iii) aerospace, including components such as morphing wings and precision connectors; (iv) automotive, including parts for mirror and seat adjustment mechanisms, as well as valve and vent control systems; and (v) robotics and smart systems, where precision and functional integration are essential.

1.3.5. Effect of EDM-Based Processes on Phase Transformation of NiTi-SMA

During EDM-based processing of NiTi alloy, the intense localized heat causes microstructural changes, such as residual stresses, a heat-affected zone, and a recast layer, which can impact the functionality of the NiTi-SMA, leading to shifts and broadening of the martensite–austenite transformation temperature range [40]. These changes can alter the alloy’s shape memory and superelastic properties by modifying the temperatures at which phase transformations occur, often reducing the efficiency of these effects. Additionally, the formation of recast layers and surface oxides can impact mechanical properties like hardness, fatigue resistance, and corrosion behavior, ultimately affecting the overall performance of the NiTi alloy after EDM machining.

1.4. Other Non-Traditional Processes for Machining SMAs

Other popular non-traditional machining processes for SMAs include electrochemical machining (ECM), laser beam machining (LBM), ultrasonic machining (USM), electrochemical polishing, and abrasive waterjet machining (WJM) [25]. ECM removes material through anodic dissolution, avoiding thermal and mechanical stresses, making it ideal for chemically stable materials and delivering an excellent surface finish. LBM uses focused laser energy for precise cutting but may cause thermal damage to the SMA microstructure; it works best on non-reflective materials. USM employs abrasive particles with ultrasonic vibrations, minimizing heat effects but potentially affecting microstructure due to mechanical impact; it is best suited for brittle materials. AWJM uses a high-pressure water jet to cut SMA without heat generation, preserving thermal properties and allowing bulk cutting, though with lower precision for fine details; it is ideal for difficult-to-cut materials [41].

This review uniquely focuses on the application of EDM and its variants, including hybrid EDM, specifically for nickel–titanium (NiTi)-based SMA materials. It offers a detailed examination of the complex machining challenges associated with NiTi-SMA materials and connects fundamental EDM mechanisms—such as plasma and spark generation—with practical outcomes, such as surface integrity and tool wear. By systematically analyzing recent research and optimization strategies, this review provides new insights and future directions to enhance EDM processes for machining SMA materials, particularly in biomedical and advanced engineering applications.

2. Past Research Work on Machining SMAs by EDM and Allied Processes

Over the years, significant research has been conducted on machining SMA materials—particularly NiTi-based alloys—using EDM and its variants, including wire-assisted EDM (WEDM), die-sinking EDM, electric discharge drilling, their micro-scale versions, and hybrid EDM-based processes, such as electrochemical discharge machining (ECDM) and ultrasonic-assisted EDM (US-EDM). Previous studies have primarily concentrated on (i) investigating the effects of key EDM parameters—such as peak current, pulse duration, duty cycle, and electrode material—on the surface integrity, MRR, and phase stability of SMAs, and (ii) optimizing EDM parameters to enhance surface finish, material removal rate (MRR), and tool wear during the machining of SMAs. A systematic approach was adopted to collect, select, and analyze the relevant literature on the machining of NiTi-SMA using EDM and its variants. The methodology for the literature selection was based on the following criteria: (i) the literature was sourced from reputable scientific databases, including Scopus, Web of Science, ScienceDirect, IEEE Xplore, and Google Scholar; (ii) relevant keywords were used during the search, such as machining of NiTi-SMA using EDM, EDM of SMA materials, and hybrid EDM for NiTi alloys; (iii) research focused on the application of EDM and its variants for NiTi-SMA materials based devices; and (iv) EDM-based past studies involving experimental investigations, modeling, optimization strategies, or performance evaluations considering the material removal rate (MRR), surface integrity (including surface roughness), and tool wear rate, for machining NiTi alloys. This section provides a comprehensive summary of past research efforts aimed at understanding and improving the EDM process for various types of SMAs. The objective of this review study is to (i) highlight the potential of EDM-based processes in the machining of SMAs; (ii) identify the significant variable parameters and their influence on performance measures; and (iii) summarize the findings of previous studies to identify research gaps and suggest future directions. A detailed analysis and discussion of past work are reported in the following sub-sections.

2.1. Past Work on Machining SMAs Using Conventional EDM and Its Variants

This subsection highlights the previous studies on machining SMA materials using traditional EDM-based processes, such as wire electrical discharge machining (WEDM), die-sinking EDM, and their micro-scale variants.

Table A1. Summary of recent research work conducted on the machining of shape memory alloy (NiTi) using EDM and its variants.

Authors (Year) [Ref. No.]	Methodology and Optimization	Machining Details (i.e., Workpiece and Tool Materials)	Selected Process Parameters	Selected Responses	Key Findings
Machining of NiTi-SMA using WEDM
Chaudhari et al. (2021) [42]	Straight cutting, L16 (44) OA, TLBO, and MOTLBO	Φ 6 mm NiTi rod Molybdenum wire (φ: 0.18 mm) Deionized water mixed with MWCNTs	Current, spark duration (Ton), spark-off duration (Toff), and powder concentration	MRR, SR (Ra), and RLT	MRR and SR improved by 75.42% and 19.15%, respectively, using 1 g/L MWCNTs in WEDM. TLBO and MOTLBO were employed for single-objective and multi-objective optimization, respectively. Maximum MRR of 0.5262 g/min and minimum SR of 1.27 µm were achieved using TLBO. FESEM analysis revealed that MWCNTs significantly reduced RLT and surface defects.
Manjaiah and Laubscher (2016) [43]	Curved cutting	Ti50Ni40Cu10 SMA Zinc-coated brass wire (φ: 0.25 mm) Dielectric: deionized water	Spark duration, spark-off duration, and servo voltage	RLT and residual stress	Identified servo voltage and spark duration as the most significant parameters affecting RLT. Recommended low voltage and short spark duration to reduce RLT.
Vakharia et al. [39] (2022) [44]	2 mm thick straight cutting, and L9 (33) OA replicate twice	Ni55.8Ti SMA (Φ 6 mm); Molybdenum wire Dielectric: deionized water	Current, spark duration, and spark-off duration	SR and surface morphology	Observed current as the most influential parameter for SR. Used a novel SinGAN technique to generate multiple images from the original ones and then fed them into four ML models.
Goyal and Rahman (2021) [45]	Straight cutting, Taguchi L27 (54), ANN, and DFA	NiTi-SMA (100 mm × 100 mm × 6 mm) Brass wire (φ: 0.25 mm) Dielectric: deionized water	Spark duration, spark-off duration, peak current, wire speed, and wire rigidity	SR and kerf width	Observed spark duration as the most influential parameter for SR and kerf width. Observed an increase in RLT with higher discharge energy. ANN models predict responses with higher accuracy than DFA.
Bisaria and Shandilya (2023) [46]	Straight cutting	Ni55.95Ti44.05 SMA Brass wire (φ: 0.25 mm) Dielectric: deionized water	Spark frequency, spark energy density, and spark gap voltage	Surface roughness (SR)	Observed that higher spark energy, resulting from a longer spark duration, leads to increased recast layer thickness, more cracks, and greater work surface deterioration.
Bisaria and Shandilya (2018) [47]	Straight cutting	Ni55.7Ti) SMA, Brass wire (φ: 0.25 mm) Dielectric: deionized water	Peak current, spark duration, and spark-off duration	Wire wear ratio (WWR) and dimensional deviation (DD)	Observed high discharge energy results in high WWR and DD. WWR and DD increased with longer spark duration and higher peak current but decreased with increase in spark-off duration.
Chaudhari et al. (2022) [48]	1.5 mm thick straight cutting, Pareto analysis, heat transfer search (HTS) algorithm	Ni55.8Ti SMA (φ 6 mm) Brass wire (φ: 0.25 mm) Dielectric: deionized water	Current, spark duration, and spark-off duration	MRR, SR, and microhardness (MH)	Recommended machining of NiTi-SMA at Ton of 40 μs, Toff of 12 μs, and current of 1 A, using a molybdenum wire by WEDM. Wire rupture occurs with an increase in crater size.
Kowalczyk and Tomczyk (2022) [49]	Straight cutting	NiTi-SMA Brass wire (φ: 0.25 mm) Dielectric: deionized water	Amplitude of the current, voltage, and energy	SR (i.e., Ra and Rz)	Achieved improved surface quality with fewer defects and no wire material contamination at optimum parameters. Identified a smoother finish at the bottom of the machined surface.
Roy et al. (2020) [50]	Straight cutting and factorial design	TiNiCu SMA Brass wire (φ: 0.25 mm) Dielectric: deionized water	Peak current and pulse peak voltage	SR (i.e., Ra and Rz)	Higher values of Ra (2.91 μm), Rz (18.24 μm), and increased surface defects were observed at elevated peak current (1.28 A) and peak voltage (53 V). A slight increase in Ra at lower peak current and voltage was due to higher Rz caused by less intense discharge conditions.
Roy et al. (2021) [51]	Taper cutting and RSM	NiTi-SMA (φ: 8 mm) Zn-coated brass wire (φ: 0.25 mm) Dielectric: deionized water	Spark duration, spindle rotational speed, and inclination angle	Volumetric material removal rate (VMRR) and Ra	VMRR and Ra improved with an increase in spindle rotational speed. The inclination angle of the wire also enhanced VMRR and Ra.
Kesavan et al. (2021) [52]	Straight cutting, Taguchi L27, and TOPSIS	NiTi-SMA Brass wire (ϕ 0.25 mm) Deionized water	Power, wire speed, spark duration, and spark-off duration	MRR and Ra	MRR increased with an increase in input power and spark duration. Higher discharge energy leads to an increase in Ra.
Kulkarni (2022) [53]	Straight cutting and RSM	NiTi-SMA plate (800 mm × 160 mm × 2 mm) Zn-coated brass wires (diffused wires) (ϕ = 0.25 mm) Deionized water	Spark duration, spark-off duration, wire feed, servo voltage, and different diffused wires	TWR and SR (Ra)	A high TWR of 0.07 g/min was observed for NiTi-SMA using an X-type electrode at higher spark duration, higher wire feed, and lower SV. A-type electrode demonstrated superior TWR, SR, and surface integrity performance compared with other electrodes in machining medical-grade NiTi-SMA.
Gupta and Dubey (2022) [54]	Straight cutting, Taguchi L27 OA, GA, and ANN	Ni54.1Ti45.9 SMA Zn-coated brass wire (ϕ = 0.25 mm) Dielectric: deionized water	Wire feed rate, wire rigidity, spark duration, spark-off duration, and peak current	MRR and SR	Proposed a hybrid genetic algorithm (GA) and an ANN model to enhance quality attributes. ANN models accurately predicted results that closely matched the experimental data.
Hou et al. (2022) [55]	Multistage straight cutting (trim cut)	NiTi-SMA Brass wire (ϕ = 0.25 mm) Dielectric: deionized water	Spark duration, spark-off duration, and peak current	SR (Ra)	Proposed multistage machining of NiTi-SMA by WEDM. Increasing cutting passes reduced crater size and convex edges, lowering Ra from 2.79 µm to 0.12 µm. In the initial machining stage, the recast layer reached 18.16 µm with many microcracks. By the fifth trimming stage, it was reduced to under 0.3 µm.
Xu et al. (2022) [56]	Straight cutting, Taguchi L27 OA, multiple regression (MLR) model, BPNN, and bat algorithm (BA)	NiTi-SMA, Brass wire (ϕ = 0.25 mm) Dielectric: deionized water	Peak current, discharge frequency, wire tension, flushing pressure, and wire speed	Cutting speed (CS) and kerf width (KW)	Found peak current as a key parameter influencing both machining efficiency and surface quality. Identified the optimal combination of process parameters using bat algorithm (BA). Experimental results and the proposed optimization approach show that the prediction errors for both MLR-BA and BPNN-BA methods are within ±2%.
George et al. (2023) [57]	Straight cutting, Taguchi, 3 level each parameter	NiTi-SMA (φ 15 mm and width of cut of 5 mm) Half-hard brass wire (ϕ = 0.25 mm) Dielectric: deionized water	Spark duration, spark-off duration, and voltage	MRR and SR (Ra)	Spark duration increased both SR and MRR. An increase in voltage reduced SR but increased MRR. Achieved minimum SR (2.168 μm) and maximum MRR (1.616 mm3/min) at optimum machining combinations.

summarizes relevant previous studies on the machining of NiTi-SMA using WEDM, and

Table A2. Summary of recent research work conducted on the machining of shape memory alloy (NiTi) using die-sinking EDM and µ-EDM.

Authors (Year) [Ref. No.]	Methodology and Optimization	Machining Details (i.e., Workpiece and Tool Materials)	Selected Process Parameters	Selected Responses	Key Findings
Machining of NiTi SMA using EDM, µ-EDM, and PMEDM
Faheem et al. (2023) [58]	Full-factorial design of experiment (33): 27 experiments, 0.5 mm depth of cut, and NSGA-II with TOPSIS	Ni55.65Ti-SMA plate (150 mm × 130 mm × 5 mm) Copper tool electrode (face size: 12 mm × 25 mm)	Spark duration, duty factor, and peak current	MRR and SR	A larger peak current increases MRR, resulting in higher productivity. Found spark duration and peak current as key factors influencing surface roughness, contributing 49.78% and 43.81%, respectively. TOPSIS integrated with NSGA-II simplifies selecting the optimal solution from the Pareto set, addressing the limitation of NSGA-II in single-solution selection.
Abidi et al. (2017) [59]	Micro-hole drilling, grey–Taguchi method, and grey-PCA	µ-EDM NiTi-SMA (3 mm × 1.5 mm × 0.5 mm) Tungsten and brass electrodes (ϕ = 100 µm) Dielectric: kerosene oil	Capacitance, discharge voltage, and electrode materials	Overcut, taper angle, and SR (Ra)	Grey-PCA indicates that the best multi-response was achieved with capacitance at level 2 (475 pF), discharge voltage at level 1 (80 V), and Cu as the electrode material. Electrode material had the strongest influence on the multi-response parameter, followed by discharge voltage and capacitance.
Abidi et al. (2017) [60]	Micro-hole drilling and MOGA-II	µ-EDM NiTi-SMA (3 mm × 1.5 mm × 0.5 mm), Tungsten and brass electrodes (ϕ = 100 µm) Dielectric: kerosene oil	Capacitance, discharge voltage, and electrode materials	MRR, TWR, and SR (Ra)	Identified capacitance as the most influential factor affecting Ra, TWR, and MRR, followed by electrode material and discharge voltage. Found electrode material as the most influential parameter affecting overcut. Taper angle is primarily influenced by capacitance and electrode material, followed by their interaction.
Gaikwad et al. (2015) [61]	Drilling a 3 mm square hole	EDM Cryogenic-treated NiTi-SMA Dielectric: kerosene oil	Gap current, spark duration, and spark-off duration	MRR and TWR	MRR and TWR increase with an increase in current. MRR increases with an increase in both spark duration and spark-off duration for treated and untreated NiTi-SMA.
Gaikwad et al. (2021) [62]	Blind cavity, L9 OA, and Buckingham’s pie theorem	EDM NiTi-SMA Dielectric: kerosene oil	Current, voltage, spark duration, and spark-off duration	RLT	Employed the Jaya algorithm to optimize the EDM of NiTi-SMA. Experimental and predicted RLT values show close agreement. Identified current as the most dominant EDM parameter affecting RLT compared with other factors.
Vora et al. (2022) [63]	Taguchi’s L9(34) replicate thrice, and HST algorithm	PMEDM NiTi-SMA Dielectric: kerosene oil	Current, spark duration, spark-off duration, and nanographene powder concentration (PC)	MRR, SR, and dimensional deviation (DD)	Ton, Toff, and PC significantly affect MRR, while all design variables influence SR and DD. Found PC, Ton, and Toff as the most influential factors, contributing 75.18%, 29.37%, and 45.72% to MRR, SR, and DD, respectively.
Singh et al. (2022) [64]	Central composite design (CCD), DFA, TLBO, and PSO techniques	EDM Fe-based SMA Copper electrode Dielectric: kerosene oil	Spark duration, spark-off duration, peak current, and gap voltage	MRR and SR	Experimental results show MRR ranging from 12.49 to 73.90 mm3/min, and SR ranging from 5.03 to 6.65 µm. Optimized MRR and SR values obtained using the DFA, TLBO, and PSO are 62.69 mm3/min and 5.56 µm; 78.85 mm3/min and 4.34 µm; and 78.91 mm3/min and 4.35 µm, respectively.
Chaudhary et al. (2017) [65]	Taper drilling and Taguchi L18 OA	Die-sinking EDM NiTi-SMA Copper electrode Dielectric: EDM oil	Polarity, peak current, spark duration, and spark-off duration	MRR and taper angle	Identified polarity and spark duration as the most influential process parameters affecting MRR and taper angle.

presents notable research on EDM and µ-EDM of NiTi-SMA. Key past research is discussed and summarized below.

Important research reports the use of carbon nanotubes in a dielectric fluid to facilitate the WEDM of NiTi-SMA by improving the MRR and reducing SR simultaneously [42]. Due to the dielectric effectiveness, the MRR was improved by 75.42% and SR by 19.15% when compared with the traditional WEDM process. The TALBO algorithm was also used to enhance the performance of WEDM by optimizing conflicting responses, MRR, and SR simultaneously. A detailed investigation of the recast layer of the Ti50Ni40Cu10 SMA workpiece cut by WEDM revealed the prominent role of voltage and pulse-on time [43]. To reduce the recast layer thickness, low-voltage and short pulse-on time settings were recommended. A tensile nature residual stress was measured for the machined workpiece having a recast layer. Vakharia et al. [44] used intelligent models to establish the relationship between WEDM variables and surface quality. They conducted nine experiments according to Taguchi L₉ (3³) OA by cutting a 2 mm thick strip from a NiTi SMA cylindrical bar of 6 mm in diameter by varying selected WEDM parameters and recording SR for each experimental run. The research outcomes showed that surface images were accurately predicted by the DenseNet model, with an average accuracy of 100%, and observed current was the most influential parameter for SR.

Goyal and Rehman [45] developed models by using an artificial neural network (ANN) to predict the SR and kerf width (KW) of a slot fabricated in a Ni49Ti51 SMA rectangular plate (100 mm × 100 mm × 60 mm) by WEDM (Figure 5). They fabricated twenty-seven slots under different machining conditions by varying the peak current, spark duration, spark-off duration, wire speed, and wire rigidity. Pulse duration was identified as the most critical process variable influencing both responses. Higher discharge energy led to an increase in the RLT, with the maximum RLT of 15.88 mm observed at an Ip of 8 A, a T_on of 120 µs, a T_off of 30 µs, a WT of 11 g/cm³, and a W_f of 4 m/mn. The study results showed that ANN models demonstrated a strong correlation between the selected WEDM parameters. Furthermore, a comparative evaluation between desirability function analysis (DFA) and ANN revealed that ANN models predict responses with greater accuracy than DFA.

In a very recent study, Bisaria and Shandilya [46] investigated the influence of spark parameters on the surface integrity of NiTi. They reported that an increase in spark energy density, due to a longer spark duration, increases the thickness of the recast layer, cracks, and work surface deterioration. In another study, they investigated the influence of variable WEDM parameters, namely, peak current, spark duration, and spark-off duration, on the wire wear ratio (WWR) and dimensional deviation (DD) for machining of Ni-rich NiTi (Ni55.7Ti) SMA. They found that the WWR and DD increased with an increase in the spark duration and peak current, while decreasing with an increase in the spark-off duration [47]. During WEDM of nitinol, Chaudhary et al. [48] obtained an improved surface quality with fewer defects and no wire material contamination at optimum parameters. Three-dimensional surface profiling and EDX analysis revealed a smoother finish at the bottom of the machined surface, attributed to increased exposure to the dielectric fluid. This indicates that balanced flushing from both directions enhances surface uniformity. Additionally, the optimized parameters helped preserve the shape memory effect while minimizing surface imperfections. Kowalczyk and Tomczyk [49] highlighted the evaluation of measurement inaccuracies associated with the electrical parameters of WEDM for machining NiTi-SMA. The surface morphology of WEDMed TiNiCu alloy was investigated by Roy et al. [50]. It was reported that different settings of the machining parameters are needed for rough and precise machining. To obtain an improved surface morphology, the sparking parameters, mainly spark duration, peak current, and voltage, should be within limits. They found that an elevated peak current (1.28 A) and voltage (53 V) led to a higher R_a (2.91 μm) and R_z (18.24 μm) and more surface defects. Roy and Mondal [51] turned on a nitinol-60 cylindrical bar having an 8 mm diameter and a 100 mm length by WEDM. Fifteen Box–Behnken design (BBD)-based experiments were performed to identify the effect of the pulse duration, spindle rotational speed, and inclination angle on the volumetric material removal rate (VMRR) and R_a. Moreover, they suggested using WEDM in multiple stages to turn a cylinder with a high erosion rate and a superior surface. They also observed that (i) the VMRR and R_a improved with increasing spindle rotational speed, and (ii) inclining the wire electrode further enhanced both the VMRR and R_a. Kesavan et al. [52] investigated the influence of WEDM parameters and optimized them to secure the best values of the MRR and SR. They observed that (i) the MRR increased with a higher input power and longer spark duration; (ii) a higher discharge energy resulted in an increased R_a. Kulkarni et al. [53] observed that a coated and diffused wire positively affects the machinability of NiTi-SMA. They concluded that (i) a high TWR of 0.07 g/min was observed for NiTi-SMA using an X-type electrode at higher spark duration, higher wire feed, and lower servo voltage, and (ii) an A-type electrode demonstrated superior performance in TWR, SR, and surface integrity compared with other electrodes when machining medical-grade NiTi-SMA. In another important study, WEDM of Ni54.1Ti45.9 was performed using Taguchi L₂₇ [54]. A hybrid genetic algorithm (GA) and artificial neural network (ANN) (integrated GA-ANN) was developed that successfully predicted the chosen responses, which were close to the experimental data. Ho et al. [55] performed multistage cutting of NiTi-SMA to minimize the thermal effects of WEDM. Reduced cutting passes were found to be effective, and low-size craters were formed, which led to a smoother surface. They observed that (i) increasing cutting passes reduced the crater size and convex edges, lowering the R_a from 2.79 µm to 0.12 µm, and (ii) in the initial machining stage, the recast layer reached 18.16 µm with numerous microcracks, which reduced to under 0.3 µm by the fifth trimming stage. Xu et al. [56] investigated the effects of the peak current, discharge frequency, wire tension, flushing pressure, and wire speed on the cutting speed (CS) and kerf width (KW) in Ni-Ti SMA machining. They developed multiple regression (MLR) and back-propagation neural network (BPNN) models to predict the CS and KW under varying parameters. The bat algorithm (BA) was applied to optimize process parameters, achieving prediction errors within ±2% for both the MLR-BA and BPNN-BA methods. Peak current was identified as the key factor affecting machining efficiency and surface quality, with the BPNN model proving effective for accurate WEDM modeling of NiTi-SMA. George et al. [57] studied the machining of low-carbon NiTi-SMA using WEDM. It was found that surface quality and productivity were both affected by the spark duration, spark-off duration, and voltage. The optimal parameters—110 µs as the spark duration, 55 µs as the spark-off duration, and 30 V as the voltage—minimized surface damage and improved surface finish. They achieved the minimum SR (2.168 μm) and the maximum MRR (1.616 mm³/min) under optimum machining combinations.

Faheem et al. [58] studied the machinability of Ni55.65Ti SMA using die-sinking EDM, focusing on the effects of the spark duration, duty factor, and peak current on the surface roughness and MRR in a full-factorial design study. NSGA-II-based multi-objective optimization produced a 6.828 µm surface roughness and a 4.552 mm³/sec MRR. In their observations, (i) peak current was identified as the most influential factor for the MRR, contributing 77.61%; (ii) spark duration and peak current were the key factors influencing surface roughness, contributing 49.78% and 43.81%, respectively; and (iii) TOPSIS combined with NSGA-II helped to select the most preferred solution from the Pareto optimal set, overcoming the challenge of choosing a single solution when using NSGA-II alone. There are some investigations on the micro-version of EDM of nitinol. µ-EDM of nitinol was conducted by Abidi et al. [59] to make a hole. They reported obtaining the machined nitinol workpiece equipped with good dimensional accuracy and surface finish at a 475 pF capacitance and an 80 V voltage. In another study by Abidi et al. [60], µ-EDM was analyzed and optimized for creating holes in NiTi-SMA using a multi-objective genetic algorithm (MOGA-II). The objective was to identify the optimal machining parameters to enhance both productivity and surface finish. The researchers found MOGA-II to be an effective tool for optimizing process variables. The best results were achieved using a tungsten electrode at a low-to-moderate capacitance and a low discharge voltage. In contrast, while a brass electrode provided a higher MRR, it resulted in increased tool wear and inferior micro-hole quality.

Another important study presented the EDM of specialized nitinol developed by cryogenic treatment [61]. It was observed that (i) both the material removal rate (MRR) and tool wear rate (TWR) increased with increasing current, and (ii) the MRR increased with a longer spark duration and spark-off duration for both treated and untreated NiTi-SMA. There is also literature evidence on the effectiveness of optimization techniques like the Jaya algorithm to facilitate EDM of NiTi alloys [62]. The study compared experimental and empirical analyses of the recast layer thickness (RLT) using Taguchi’s L₉ design and Buckingham’s π-theorem, showing close agreement between the results. Current was identified as the dominant parameter affecting the workpiece lateral taper (WLT). The Jaya algorithm optimized the WLT to 4.846 μm, with a voltage of 55 V, a current of 4 A, a pulse-on time of 20 μs, and a pulse-off time of 8.2 μs. In a recent investigation, Vora et al. [63] performed power-mixed electric discharge machining (PMEDM) of NiTi-SMA. They mixed nanographene in the dielectric and found powder concentration as one of the significant parameters that affected the machinability of NiTi. They reported that the addition of graphene powder improved the material removal rate and minimized roughness and dimensional inaccuracy. Singh et al. [64] optimized key process parameters of EDM, including the spark duration, spark-off duration, peak current, and gap voltage, for machining Fe-based SMA using a copper electrode. The experimental results showed an MRR ranging from 12.49 to 73.90 mm³/min, and an SR ranging from 5.03 to 6.65 µm. The optimized MRR and SR values obtained using DFA, teacher learning-based optimization (TLBO), and particle swarm optimization (PSO) techniques were 62.69 mm³/min and 5.56 µm; 78.85 mm³/min and 4.34 µm; and 78.91 mm³/min and 4.35 µm, respectively. Chaudhury et al. [65] studied the influence of EDM variable process parameters—spark duration, spark-off duration, polarity, and peak current—on the MRR and taper angle (TA) for drilling holes in nitinol through Taguchi L₁₈. The results showed that polarity and spark duration significantly influenced EDM performance. The optimal machining conditions for the maximum MRR and minimum TA were (i) a 12 A peak current, a 20 µs spark duration, and a 6 µs spark-off duration, and (ii) a 12 A peak current, a 30 µs spark duration, and a 4 µs spark-off duration, respectively.

2.2. Past Work on Advanced and Hybrid-EDM Processes for SMA

This subsection highlights the previous studies on machining SMA materials using advanced and hybrid-EDM-based machining processes, such as electrical discharge drilling (EDD), ultrasonic-assisted EDM (US-EDM), electrochemical spark-erosion machining (EC-EDM), cryogenic-assisted EDM, dry-EDM, and dry-WEDM.

Table A3. Summary of recent research work conducted on the machining of shape memory alloy (NiTi) using hybrid EDM.

Authors (Year) [Ref. No.]	Methodology and Optimization	Machining Details (i.e., Workpiece and Tool Materials)	Selected Process Parameters	Selected Responses	Key Findings
Om and Singh et al. (2017) [66]	Drilling and one-factor-at-a-time (OFAT)	EDD NiTiCu10 SMA Dielectric: EDM oil	Pulse current, gap voltage, spark duration, spark-off duration, and rotational speed of the tool electrode	MRR, TWR, and SR	Achieved maximum MRR (4.077 mm3/min) at the maximum peak current of 12 A. Obtained the lowest TWR (0.031 mm3/min) and SR (2.8 µm), both at the lowest spark duration of 15 µs.
Chaudhary and Haribhakta et al. (2017) [67]	Micro-hole (through or blind)	EDD SMA Dielectric: EDM oil	-	-	This study highlights prior research on EDD applied to SMA materials.
Mane and Jadhav (2024) [68]	Drilling, Taguchi L18 OA, and RSM	USV-EDM NiTi-SMA Copper electrode Dielectric: EDM oil	Low-frequency ultrasonic vibration	SR	Low-frequency ultrasonic vibration in EDM can reduce surface roughness.
Wang et al. (2018) [69]		USV-MF complex-assisted WEDM-LS TiNi01 SMA Copper electrode Dielectric: EDM oil	Spark duration, spark-off duration, and current	MRR and SR	Proposed ultrasonic vibration (USV) and magnetic field (MF)-assisted WEDM-LS (USV-MF complex-assisted WEDM-LS) to enhance the machining of SMA.
Huang et al. (2003) [70]	Drilling micro-holes	USV µ-EDM NiTi-SMA Dielectric: EDM oil	-	Machining efficiency	Introducing ultrasonic vibration to the µ-EDM process enhanced machining efficiency by over 60 times, with minimal increase in TWR.
Kumar et al. (2023) [71]	Drilling and DFA	Electrochemical arc machining (ECAM) Ni55.7Ti SMA Molybdenum electrode	Supply voltage	Overcut and TWR	Supply voltage significantly influences the machining rate. SEM, EDS, and XRD analysis revealed thermal damage and TiC formation, suggesting potential impacts on the alloy’s shape memory properties.
Chaudhari et al. (2022) [72]	Straight cutting, BBD, and TLBO	Near-dry WEDM NiTi-SMA Molybdenum wire (φ 0.18 mm) Dielectric: compressed gas	Current, spark duration, and spark-off duration	MRR and SR	Near-dry WEDM process is well-suited for achieving better surface quality with fewer surface defects. Near-dry WEDM reduced MRR by 8.94% while improving surface roughness by 41.56%.
Muniraju and Talla (2024) [73]	-	Dry-EDM NiTi-SMA	-	-	This study reviewed sustainable EDM techniques for nitinol SMA, focusing on dry-EDM, bio-dielectrics, and powder additives.
Jatti and Singh (2014) [74]		EDM Cryogenic-treated NiTi-SMA Dielectric: EDM oil	-	MRR and TWR	Cryogenic treatment significantly improved electrical conductivity and increased MRR by approximately 19%, with only a slight change in TWR.

MRR—material removal rate, SR—surface roughness, WWR—wire wear rate, KW—kerf width, CS—cutting speed, RLT—recast layer thickness, TWR—tool wear rate, DA—dimensional accuracy, ANOVA—analysis of variance, OC—overcut, Ra—average surface roughness, Rz—mean roughness depth, T_on—spark duration, Toff—spark-off duration, TLBO—teaching–learning-based optimization, DFA—desirability function analysis, ANN—artificial neural network, GA—genetic algorithm, OA—orthogonal array, and FESEM—field-emission scanning electron microscope.

summarizes relevant previous studies on machining NiTi-SMA using hybrid EDM. The highlights of the relevant literature are discussed below.

There are some previous studies on drilling to make a hole in an SMA using EDM and EDD processes. As discussed in the above subsection, a few researchers used µ-EDM to make holes in NiTi-SMA [53,54]. Om and Singh [66] investigated the EDD of NiTiCu10 SMA using a copper tool electrode. Using a one-factor-at-a-time (OFAT) approach, they evaluated the MRR, TWR, and SR by varying the pulse current, gap voltage, spark duration, spark-off duration, and tool rotation speed. The highest MRR of 4.077 mm³/min was recorded at a 12 A pulse current, while the lowest TWR (0.031 mm³/min) and SR (2.8 μm) were achieved at a spark duration of 15 μs. Scanning electron microscope analysis revealed typical EDM-induced features, such as craters, debris, microcracks, and resolidified layers. The study effectively identified key process parameters and their optimal ranges for improved machining performance. In another study, Chaudhary and Haribhakta [67] reviewed EDM-based processes for micro-hole drilling, emphasizing their role in producing precise, small holes for miniaturized nitinol components and highlighting mechanical micro-drilling as a cost-effective alternative, with modeling aiding in optimal process parameter selection.

Mane and Jadhav [68] applied low-frequency ultrasonic vibration to NiTi-SMA during EDM, finding that it reduces surface roughness. Using an L₃₂ orthogonal array and response surface methodology, optimal process parameters were identified and experimentally validated. In another innovative study, Wang et al. [69] proposed a novel ultrasonic vibration (USV) and magnetic field (MF) complex-assisted WEDM-LS technique that significantly improves machining performance for thick TiNi-01 SMA. Both simulation and experimental results demonstrated that the combined USV-MF approach enhances wire electrode stability by promoting a more uniform discharge distribution, thereby reducing wire breakage. Validated process models for MRR and surface roughness (R_a) confirmed that this method simultaneously increases machining efficiency and surface quality compared with conventional and MF-assisted WEDM-LS. Overall, this innovative machining approach offers promising advantages and practical potential for advanced manufacturing of difficult-to-machine materials. Ultrasonic vibration-assisted EDM was found to be influential in a study conducted by Huang et al. [70]. A super high increment in efficiency was recorded without compromising the electrode wear. Enhanced flushing due to the strong stirring effect of vibration was the key. Electrochemical spark-erosion machining (EC-EDM) is a hybrid process that combines electric discharge and electrochemical dissolution, making it effective for machining hard-to-machine materials, such as Ni55.7Ti alloy, where conventional methods are inadequate. Some prior studies have explored the use of EC-EDM on SMAs. Kumar et al. [71] investigated the electrochemical arc machining (ECAM) of Ni55.7Ti SMA using a molybdenum electrode. Using the response surface methodology and the desirability function approach, they identified supply voltage as the most influential factor affecting the MRR, TWR, and OC. Surface analyses using EDM, EDS, and XRD revealed thermal damage and TiC formation, indicating possible effects on the alloy’s shape memory properties.

Chaudhary et al. [72] studied near-dry WEDM for machining nitinol SMA, showing it as a greener alternative to wet-WEDM, with better surface quality. Discharge energy-based parameters were found to significantly influence the MRR and surface roughness. Near-dry WEDM reduced the MRR by 8.94% but improved surface roughness by 41.56%. EDM analysis confirmed fewer surface defects, making it ideal for high-finish, eco-friendly applications. Muniraju and Talla [73] reviewed sustainable EDM methods for machining Nitinol SMA, focusing on dry/near-dry EDM, bio-dielectrics, and conductive powders. Their study highlights these approaches as effective in reducing environmental impact, improving efficiency, and enhancing surface quality, providing key insights for eco-friendly EDM practices.

Jatti and Singh [74] studied the impact of cryogenic treatment at around −185 °C on the machinability of NiTi-SMA during EDM. The treatment notably improved electrical conductivity and increased the MRR by about 19%, with only slight changes in the TWR.

2.3. Optimization of EDM and Variants for Machining SMAs

Optimization of EDM-based processes is crucial due to the complexity of process parameters and the presence of conflicting performance objectives. This subsection reviews studies using techniques like TLBO, PSO, the desirability approach, RSM, GA, ANN, GA-ANN, BPNN, MOGA-II, NSGA-II, TOPSIS, and grey relational analysis for machining SMAs. Chaudhari et al. [42] improved WEDM performance by simultaneously optimizing conflicting responses, MER, and SR, using the TALBO algorithm. ANN models were developed to predict the SR and kerf width (KW) of slots in Ni₄₉Ti₅₁ SMA fabricated by WEDM [45], while GA-ANN predicted the SR and MRR for Ni_54.1Ti_45.9 SMA [54]. An ANN accurately predicted results that closely matched the experimental data, given sufficient quality input. While the GA can find near-optimal solutions, it may not always reach the exact global optimum [54]. A BPNN was used to predict the CS and KW of WEDM-machined NiTi-SMA [56]. NSGA-II optimized the MRR and mean roughness depth (R_z) during WEDM of a Ni_55.8Ti cylindrical plate. Integrating TOPSIS with NSGA-II simplifies the selection of the optimal solution from the Pareto set, effectively addressing NSGA-II’s limitation in choosing a single best solution. [58]. Additionally, the desirability approach, TLBO, and PSO were applied for the multi-objective optimization of conflicting responses [64]. Each algorithm offers unique strengths: NSGA-II for multi-objective problems, GA-ANN for combined modeling and optimization, and PSO for fast, continuous-space optimization. The choice depends on problem complexity, accuracy needs, and computational cost.

Table 1. Comparative analysis of NSGA-II, GA-ANN, and PSO in terms of efficiency, accuracy, and application suitability.

Algorithm	Efficiency	Accuracy	Application
NSGA-II	High	High	Multi-objective optimization
GA-ANN	Moderate	Very high (prediction and optimization)	Predictive modeling and optimization
PSO	Very high	High	Fast, single-objective, or hybrid optimization

provides a comparison of the NSGA-II, GA-ANN, and PSO algorithms for EDM of SMAs.

3. Conclusions and Future Research Directions

This review highlights the extensive use of EDM-based machining processes, particularly die-sinking EDM and WEDM, for machining NiTi alloys, primarily in biomedical applications. The key findings include the following:

• EDM-based processes, especially WEDM, have been widely used for machining NiTi alloys, mainly in biomedical applications.
• Spark duration, current, and voltage have been identified as the machining variable parameters significantly affecting the MRR, surface roughness, tool wear, dimensional deviation, and overcut in machining SMAs by EDM-based processes.
• Powder-mixed EDM has shown improved efficiency and productivity.
• WEDT has enabled the fabrication of cylindrical NiTi components.
• Various optimization techniques, such as GA, ANN, NSGA-II, and TOPSIS, have been successfully used for multi-objective optimization of EDM, particularly in addressing conflicting responses related to productivity and quality.
• Hybrid and AI-based methods have effectively improved surface quality and reduced thermal damage.
• There is still a need for in-situ monitoring and adaptive control to enhance EDM precision, repeatability, and overall efficiency.

The key research gaps identified from the past work are as follows:

• Surface integrity studies have largely focused on roughness and recast layer thickness; limited work exists on geometrical profile, microhardness, defects, and microstructural changes, including the heat-affected zone.
• Few efforts have been made to address the multi-objective optimization of conflicting machining goals (e.g., surface quality vs. productivity).
• The impact of the electrode material on thermal damage, phase transformation, and shape memory retention remains underexplored.
• Limited research exists on fabricating complete engineering components using EDM for NiTi alloys.

The following points outline potential avenues for future research in the areas of process optimization and control, materials science, sustainability, and environmental concerns:

• Integrating EDM with other advanced machining processes, such as additive processes, for improved capabilities.
• Employing artificial intelligence (AI) and machine learning (ML) for process optimization, especially to balance conflicting responses.
• Developing in situ monitoring and adaptive control systems for real-time parameter adjustment.
• Studying NiTi phase transformation under EDM thermal cycles.
• Exploring the effects of Ni/Ti alloy composition on surface integrity and functional behavior.
• Investigating eco-friendly EDM approaches (e.g., dry-EDM and green dielectrics).
• Conducting life cycle and sustainability analyses of EDM processes.
• Evaluating energy usage, emissions, and resource efficiency in EDM of NiTi.