Performance and surface integrity studies in electric discharge machining on Ti-6Al-4 V using neem oil dielectric and 3D Printed SS316L electrode Authors: Satyabrata Barik, Basanta Kumar Nanda, Swayam Bikash Mishra, Mantra Prasad Satpathy, Seyfu Tiruneh Debebe Published: 2026 ABSTRACT Bio-dielectric fluids and advanced 3D printed electrodes are essential for sustainable, precise Electric Discharge Machining (EDM) of critical Ti-6Al-4 V aerospace components. In this study, neem biodiesel dielectric and a precisely manufactured 3D-printed stainless-steel electrode were used to identify optimal EDM parameters that maximize material removal rate (MRR) and minimize tool wear rate (TWR) and surface roughness (SR). The experiments were carried out using response surface methodology (RSM) with Box–Behnken design (BBD). Regression models showed high predictive accuracy (R² > 0.98). An integrated MEREC-CoCoSo multi-response optimization approach was subsequently applied to the experimental results to obtain the best output values. The results revealed the optimal setting as current (Ip) of 30 A, pulse-on time (Ton) of 0.3 s, and voltage gap (Vg) of 60 V. The interaction between Ton and Vg strongly influenced MRR, TWR, and SR. MRR first increased and then slightly decreased as Vg was raised from 60 V to 80 V. It also showed a marginal increase as Ton increased from 0.2 s to 0.4 s. In contrast, both TWR and SR increased sharply with higher Vg and longer Ton. Surface morphology analysis at the optimal setting showed fewer surface flaws, smaller craters, thinner recast layers, and lower porosity. Energy-dispersive X-ray spectroscopy (EDX) analysis confirmed the transfer of material from the 3D-printed electrode to the workpiece. This work provides a practical guide for industrial applications, demonstrating the feasibility and benefits of using novel electrode materials and eco-friendly dielectrics in sustainable EDM. Introduction Electrical discharge machining (EDM) has been recognized as an effective technique for applications requiring the processing of materials with intricate geometries, exceptional hardness, and high strength1. However, elevated process temperatures in EDM can induce various surface defects on the workpiece, including porosity, cracking, residual stresses, and imperfectly recast or re-solidified layers. As manufacturing requirements evolve toward miniaturization and precision, variants such as micro-EDM (µ-EDM) and wire-EDM have become critical for processing advanced alloys. For instance, recent studies have optimized EDM process parameters for superalloy Incoloy 925 using response surface methodology to enhance machining efficiency2. These imperfections typically arise from the severe thermal cycling inherent in the EDM process3. The integration of a bio-dielectric medium with a 3D-printed electrode in EDM could be precisely termed “bio-dielectric aided 3D-printed-electrode EDM”. A significant advantage of employing biologically based EDM machining is its demonstrated ability to improve machinability and surface finish in difficult-to-cut materials4. These additively manufactured electrodes, such as AlSi10Mg, produced by direct metal laser sintering, demonstrated 23% lower tool wear rates and 18% improved surface quality compared to conventional electrodes during the EDM of Ti-6Al-4 V5. This inherent characteristic significantly broadens EDM’s versatility and expands its applicability across diverse industrial sectors, such as automotive, aerospace, and medical engineering6,7. Consequently, bio-dielectric-aided 3D-printed-electrode EDM offers improved surface quality and accuracy, leading to enhanced surface smoothness and reduced recast layer thickness, which are frequently recognized as persistent challenges in traditional EDM operations. A substantial body of research has been dedicated to exploring the EDM of the Ti-6Al-4 V alloy. These investigations have predominantly examined the effects of various dielectric fluids and a range of process parameters on the machining outcomes. Furthermore, the role of tapered tool design in the EDM of Ti-6Al-4 V has been thoroughly assessed through detailed experimental studies8. There were significant improvements in MRR and TWR results. Additionally, a uniform recast layer was also produced. Mustu et al.9 revealed the effect of current and pulse duration in powder-mixed EDM on TiB2 and nanographene-reinforced ZK60 Mg alloys. It was noted that the hardness and machining accuracy increased, while the surface roughness was also improved. In µ-electro discharge milling of Nitinol, sesame oil enables the fabrication of microchannels with a material removal rate of 0.076 g/min and a surface roughness of 0.387 μm10. This sustainable approach enhances biocompatibility, improving cell viability by 85.11%. Bisaria et al.11 provided a comprehensive review of Hydroxyapatite powder-mixed EDM for bio-implant surface modification, aiming to advance HAp-EDM in biomedical applications. Grigoriev et al.12 utilized a water-based dielectric with TiO2 powder suspension as an alternative to traditional hydrocarbon methods in wire EDM of Al2O3. It was observed that there was a 1.5-fold improvement in MRR and a kerf depth of 54.16 μm. Other researchers investigated the influence of various nanoparticles, such as Al2O3 and titanium powder, in the dielectric fluid for machining different types of steel13–16. These studies revealed significant improvements in MRR and reduced TWR, with minimal surface roughness (SR), because of the enhanced heat-absorption capacity of nanoparticles. In the context of EDM of difficult-to-cut materials, Chaurasia et al.17 demonstrated that input parameters must be optimized to achieve a balanced material removal rate and surface roughness for Ni-rich NiTi shape memory alloy. Notably, their investigation using a blend of sesame vegetable oil and black phosphorus at a 2 g/L concentration revealed that post-machining cell viability improved to 79.89%, indicating that properly optimized micro-EDM processes can enhance biocompatibility while maintaining surface quality. Likewise, in the µ-electro discharge milling of Nitinol SMA, canola oil was identified as the optimal dielectric, yielding the highest cell viability of 179.89%18. While sesame oil achieved the minimum TWR, sunflower oil recorded the maximum MRR. Meanwhile, a significant challenge in EDM is maintaining surface quality while achieving high production rates. Thus, optimization of machining parameters is critical. Thus, response surface methodology (RSM) is widely employed in EDM experimentation to obtain the desired responses19. In some EDM literature, optimal parameter combinations are established using TOPSIS-CRITIC and desirability techniques to achieve desired outcomes. The effectiveness of these determined combinations is subsequently validated by confirmatory tests20,21. A novel Combined Compromise Solution (CoCoSo) optimization method is employed across numerous manufacturing sectors, yielding significant results22. CoCoSo and CODAS techniques have been applied in the WEDM process to identify optimal parameter combinations23. Likewise, the EDM process was analyzed using Cu, Cu-W and Cu-Zn electrodes. The result analysis was conducted using the rank and CoCoSo methods. The Cu electrode demonstrated a higher MRR than the other two electrodes. Meanwhile, the Cu-W tool exhibited a lower TWR, and the Cu-Zn tool demonstrated a lower SR. Previous scholarly articles have explored different strategies to improve EDM performance. However, limited research currently exists on bio-dielectric-assisted 3D-printed-electrode EDM. The objectives of the current investigation are to explore the influence of neem biodiesel dielectric and sustainably fabricated 3D-printed SS316L tool electrodes in the EDM process. Moreover, optimal parametric combinations are identified using Ti-6Al-4 V as the work material, with a focus on achieving high MRR, minimal TWR, and SR. This research work also employs a method based on the removal effect of criteria (MEREC) and a CoCoSo approach to provide the optimization results. Furthermore, the characteristics of the EDMed surface, such as craters, a recast layer, and porosity, are explored using scanning electron microscopy (SEM) and EDX analyses. This research provides a sustainable pathway for industrial EDM applications. Materials and methods Electrode, workpiece and bio-dielectrics materials In the current research on the EDM process, SS316L electrodes are fabricated using a selective laser melting system (SLM 280, EOS®, Germany). The outer diameter of the electrode is 10 mm. The SLM process involves a protective argon gas supply at 4,000 hPa and 100 L/min, a Yb-fiber laser power of 200 W, a scanning speed of 7.0 m/s, and an adjustable focus diameter range of 100 to 500 μm. The system’s build volume dimensions are 250 mm × 250 mm × 325 mm. Furthermore, during this operation, compressed air of 7,000 hPa is also supplied. This 3D-printed SS316L electrode offers distinct advantages over conventional copper and graphite electrodes, including higher density and better electrical and thermal properties. Table 1 presents the elemental composition of the fabricated electrode. Table 1Chemical composition of the 3D-printed SS316L electrode.ElementsCSiMnCrCuNiSMo N FeWt. (%)0.0260.370.1616.550.16100.0272.020.036Balance Table 1Chemical composition of the 3D-printed SS316L electrode.ElementsCSiMnCrCuNiSMo N FeWt. (%)0.0260.370.1616.550.16100.0272.020.036Balance Chemical composition of the 3D-printed SS316L electrode. The rectangular plate of Ti-6Al-4 V alloy was selected as the workpiece material for this investigation. Its dimensions were 100 mm in length, 60 mm in width, and 5 mm in thickness. The comprehensive mechanical and chemical characteristics of this specific work material are meticulously presented in Table 2. Table 2Mechanical and chemical characteristics of Ti-6Al-4 V work material.PropertiesUnitWork material specifications(Ti-6Al-4 V )Densitykg/m34429Melting point°C1660Thermal conductivityW/mK6.70Tensile strengthMPa8.9 × 10− 6Vickers micro-hardnessHV268Chemical contentWt. (%)Ti-89.71%, Al-6.18%, V-3.94%, O-0.2%, Fe-0.036%, C-0.024% Table 2Mechanical and chemical characteristics of Ti-6Al-4 V work material.PropertiesUnitWork material specifications(Ti-6Al-4 V )Densitykg/m34429Melting point°C1660Thermal conductivityW/mK6.70Tensile strengthMPa8.9 × 10− 6Vickers micro-hardnessHV268Chemical contentWt. (%)Ti-89.71%, Al-6.18%, V-3.94%, O-0.2%, Fe-0.036%, C-0.024% Mechanical and chemical characteristics of Ti-6Al-4 V work material. Biodiesel is widely regarded as an eco-friendly diesel due to its negligible environmental footprint and reduced harmful effects on personnel involved in its operation. Neem oil, extracted from the Azadirachta indica tree, is a vegetable-based oil distinguished by its notably high content of free fatty acids. This bio-oil was subsequently converted to biodiesel via transesterification, as described by Moazeni et al.24. The governing chemical equation is presented in Eq. (1). A visual representation of the various stages involved in the preparation of this bio-dielectric is provided in Fig. 1.1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{C}_{3}{H}_{5}{\left(OOCR\right)}_{3}+3ROH\underleftrightarrow{catalyst}3{R}^{{\prime\:}}C{O}_{2}R+{C}_{3}{H}_{8}{O}_{3}$$\end{document} Fig. 1Steps in preparing bio-dielectrics. Steps in preparing bio-dielectrics. The transesterification procedure commenced by heating 1 L of neem oil to 75–80 °C for 30 min. Following this, a methoxide solution, prepared by dissolving 4 g of NaOH in methanol (representing 25% of the total volume), was slowly introduced into the preheated oil. The mixture was thoroughly agitated, then allowed to settle for 24 h. Subsequently, the lighter biodiesel phase was carefully decanted into a clean beaker, thereby separating it from the denser, brownish glycerine-based byproduct that had accumulated at the bottom (Fig. 2). The biodiesel underwent further purification through several cycles of mixing and washing with distilled water. Subsequently, the solution was allowed to settle for at least 72 h, yielding transparent, purified biodiesel. The physicochemical properties characterizing the dielectrics derived from neem biodiesel are comprehensively detailed in Table 3. Moreover, bio-dielectrics have been shown to lower specific energy consumption and overall CO2 emissions compared to kerosene-based dielectrics, facilitating a more sustainable life-cycle profile for the EDM of hard-to-cut alloys like Ti-6Al-4 V25. Fig. 2Transesterification process to produce neem bio-diesel. Transesterification process to produce neem bio-diesel. Table 3Physicochemical properties of neem bio-dielectric.PropertiesUnitNeem bio-dielectricDensityg/cm30.868Viscositycm2/s4.71Thermal conductivityW/mK0.15Specific heatkJ/kgK1.962Breakdown voltagekV27 Table 3Physicochemical properties of neem bio-dielectric.PropertiesUnitNeem bio-dielectricDensityg/cm30.868Viscositycm2/s4.71Thermal conductivityW/mK0.15Specific heatkJ/kgK1.962Breakdown voltagekV27 Physicochemical properties of neem bio-dielectric. Experimental setup Experimental trials were conducted utilizing an Electronica Electraplus PS 50ZNC Die Sinking Machine. The Servomax pulse generator was integrated to facilitate this experimental process. A custom-fabricated chamber, measuring 30 cm × 22 cm × 17 cm, along with a dedicated pump system, was employed for the experiments. Peak current (Ip), pulse-on time (Ton), and gap voltage (Vg) are selected as the EDM process parameters based on the preliminary pilot study. Table 4 presents the discrete levels of the selected process parameters. A detailed summary of the works carried out in this study is represented in Fig. 3. Table 5 outlines the Box-Behnken design, which enumerates 15 possible combinations of these input variables. Each experimental run is replicated 3 times to ensure the correctness of the results and to reduce measurement error. In this study, the output indicators, namely MRR, TWR, and SR, are assessed with great precision. Table 4Control parameters and their levels.ParametersUnitNotation-10+ 1CurrentAmpsIp202530Pulse ON timesTon0.20.30.4Voltage gapVoltsVg607080 Table 4Control parameters and their levels.ParametersUnitNotation-10+ 1CurrentAmpsIp202530Pulse ON timesTon0.20.30.4Voltage gapVoltsVg607080 Control parameters and their levels. Fig. 3Overall representation of current research methodology. Overall representation of current research methodology. Table 5MRR, TWR and SR results under different parametric conditions.Exp. runIp (A)Ton (s)Vg (V)MRR (g/s)TWR (g/s)SR (µm)1200.2700.2160.2988.5002200.3800.1280.1957.3203200.3600.2010.24211.5204200.4700.2030.27313.0405250.2600.2000.2408.1706250.2800.1600.27013.8607250.3700.2200.23011.0308250.3700.2300.22011.8909250.3700.2300.22010.93010250.4600.2200.34016.30011250.4800.1000.2205.65012300.2700.3660.38610.78013300.3600.3210.3458.99014300.3800.2780.3378.23015300.4700.3230.4317.990 Table 5MRR, TWR and SR results under different parametric conditions.Exp. runIp (A)Ton (s)Vg (V)MRR (g/s)TWR (g/s)SR (µm)1200.2700.2160.2988.5002200.3800.1280.1957.3203200.3600.2010.24211.5204200.4700.2030.27313.0405250.2600.2000.2408.1706250.2800.1600.27013.8607250.3700.2200.23011.0308250.3700.2300.22011.8909250.3700.2300.22010.93010250.4600.2200.34016.30011250.4800.1000.2205.65012300.2700.3660.38610.78013300.3600.3210.3458.99014300.3800.2780.3378.23015300.4700.3230.4317.990 MRR, TWR and SR results under different parametric conditions. Measurement of responses Measurement of material removal rate (MRR) The MRR is determined by evaluating the weight difference of the workpiece prior to and following the machining process, as illustrated in Eq. (2). This measurement is carefully performed employing a digital weighing balance (Make: Metler Toledo®, US).2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:MRR=\left[\frac{{W}_{i}-{W}_{f}}{t}\right]$$\end{document} Where Wi is the initial weight of the sample, Wf is the final weight of the sample after EDM operation, and t is the machining time in seconds. Measurement of Tool wear rate (TWR) Accurate measurement of TWR is crucial for evaluating electrode performance and the overall machining efficiency in the EDM process. Analogous to the MRR calculation, the TWR is assessed by determining the weight difference of the electrode before and after each experimental trial, as shown in Eq. (3).3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:TWR=\left[\frac{{T}_{i}-{T}_{f}}{t}\right]$$\end{document} Where Ti is the initial weight of the electrode, Tf is the final weight of the electrode after EDM operation, and t is the machining time in seconds. Surface topography The SR (Ra) of the machined face produced by the EDM process was accurately measured using a portable stylus profilometer, the Talysurf (Model: Taylor Hobson®, Surtronic 3+). A cut-off length (Lc) of 1 mm, a sampling length (Ln) of 4 mm, and a CR ISO filter were used during the measurements. Surface images of the specimens were captured using an optical microscope (Model SCD313 BPD, Radical Instrument) at 5× magnification. For a more detailed, three-dimensional evaluation of surface roughness and nanoscale topography, Atomic Force Microscopy (AFM) (Brukers® INNOVA SPM) was employed with a 100 μm scanner. Further in-depth analysis of the EDM-treated surface morphology was performed using Scanning Electron Microscopy (SEM) with integrated Energy Dispersive Spectroscopy (EDS) (Model: EVO MA18 with Oxford EDS detector). It operates at a fixed voltage of 10 kV. MEREC method for criterion weight allocation The field of multiobjective optimization is characterized by considerable mathematical complexity. This complexity stems from the necessity to concurrently optimize multiple performance indicators, which are often in conflict with one another. To effectively address this intricate issue, the MEREC is systematically applied. Step 1 The initial phase of this methodology involves systematically creating a decision matrix that encapsulates various response parameters. 4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D=\left[\begin{array}{c}\begin{array}{ccc}{x}_{11}&\:{x}_{12}&\:\begin{array}{cc}.\:.\:.&\:{x}_{1n}\end{array}\\\:{x}_{21}&\:{x}_{22}&\:\begin{array}{cc}.\:.\:.&\:{x}_{2n}\end{array}\\\:.\:.\:.&\:.\:.\:.&\:\begin{array}{cc}.\:.\:.&\:.\:.\:.\end{array}\end{array}\\\:\begin{array}{ccc}{x}_{m1}&\:{x}_{m2}&\:\begin{array}{cc}.\:.\:.&\:{x}_{mn}\end{array}\end{array}\end{array}\right]$$\end{document} Step 2 This stage involves normalizing the previously constructed decision matrix using Eqs. (5, 6). For a beneficial (higher-the-better) criterion:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{ij}=\frac{Min\left({x}_{ij}\right)}{{x}_{ij}}$$\end{document} For non-beneficial (lower-the-better) criterion:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{ij}=\frac{{x}_{ij}}{Max\left({x}_{ij}\right)}$$\end{document} Step 3 The performance of the alternatives (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\eta\:}_{i}$$\end{document}) can subsequently be calculated using the following mathematical expression. 7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\eta\:}_{i}=\mathrm{ln}\left(1+\left(\frac{1}{n}\sum\:_{j}^{m}\left|\mathrm{ln}\left({N}_{ij}\right)\right|\right)\right)$$\end{document} Where n is the number of responses. Step 4 The overall performance (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\eta\:}_{ij}^{{\prime\:}}$$\end{document}) is then estimated using the subsequent equation. 8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\eta\:}_{ij}^{{\prime\:}}=\mathrm{ln}\left(1+\left(\frac{1}{n}\sum\:_{k,k\ne\:j}^{m}\left|\mathrm{ln}\left({N}_{ij}\right)\right|\right)\right)$$\end{document} Step 5 The removal effect (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{j}$$\end{document}), and its value can be determined using the subsequent equation. 9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{j}=\sum\:_{i}^{m}\left|{\eta\:}_{ij}^{{\prime\:}}-{\eta\:}_{i}\right|$$\end{document} Step 6 Ultimately, the final priority weight (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{W}_{j}$$\end{document}) is calculated as follows. 10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{W}_{j}=\frac{{E}_{j}}{\sum\:_{k}{E}_{k}}$$\end{document} CoCoSo method The CoCoSo technique, a novel methodology, was introduced by Yazdani et al.26 and combines the simple additive weighting method (WSM) with the exponentially weighted product model (WPM). The computational procedure for CoCoSo adhered strictly to the guidelines established as follows: Step 1 The corresponding decision matrix is initially constructed, as indicated in Eq. (4). Step 2 The above-constructed matrix undergoes a normalization procedure. The equations are as follows. For a beneficial (higher-the-better) criterion:11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{ij}=\frac{{x}_{ij}-Min\left({x}_{ij}\right)}{{Max(x}_{ij})-{Min(x}_{ij})}$$\end{document} For non-beneficial (lower-the-better) criterion:12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{ij}=\frac{Max\left({x}_{ij}\right)-{x}_{ij}}{{Max(x}_{ij})-{Min(x}_{ij})}$$\end{document} Step 3 For each alternative, the corresponding scores for the WSM comparability sequence \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}$$\end{document} and the power of weighted comparability \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{i}$$\end{document} sequences are subsequently determined using Eqs. (13 and 14). 13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}=\sum\:_{j=1}^{n}\left({W}_{j}\times\:{N}_{ij}\right)$$\end{document} 14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{i}=\sum\:_{j=1}^{n}{\left({N}_{ij}\right)}^{{W}_{j}}$$\end{document} Step 4 The proportional weights of the alternatives are computed utilizing the Eqs. (15, 16 and 17). 15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ia}=\frac{{P}_{i}+{S}_{i}}{\sum\:_{i=1}^{m}\left({P}_{i}+{S}_{i}\right)}$$\end{document} 16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ib}=\frac{{P}_{i}}{\underset{i}{\mathrm{min}}{P}_{i}}+\frac{{S}_{i}}{\underset{i}{\mathrm{min}}{S}_{i}}$$\end{document} 17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ic}=\frac{{\lambda\:\times\:S}_{i}+\left(1-\lambda\:\right)\times\:{P}_{i}}{\lambda\:\times\:\underset{i}{\mathrm{max}}{S}_{i}+\left(1-\lambda\:\right)\times\:\underset{i}{\mathrm{max}}{P}_{i}}0\le\:\lambda\:\le\:1$$\end{document} Equation (15) demonstrates the arithmetic mean of the cumulative scores derived from both the WSM and the exponential WPM models. Furthermore, Eq. (16) quantifies the composite relative scores from the WSM and WPM. The harmonic combination of the scores obtained from the WSM and WPM is illustrated in Eq. (17), where λ is set to 0.5 (the default value). Step 5 The final appraisal score for each alternative is estimated using Eq. (18). 18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{i}={\left({\gamma\:}_{ia}\times\:{\gamma\:}_{ib}\times\:{\gamma\:}_{ic}\right)}^{\frac{1}{3}}+\frac{1}{3}\left({\gamma\:}_{ia}+{\gamma\:}_{ib}+{\gamma\:}_{ic}\right)$$\end{document} The alternative with the highest appraisal score is deemed the optimal choice for final selection. Figure 4 provides a comprehensive flowchart outlining the application of the integrated MEREC-CoCoSo methodology. This novel methodology is used to effectively optimize the current research work. Initially, the MEREC method is used to establish multiple sets of weights for the considered responses based on the preferences for input parameters (such as Ip, Ton, and Vg), thereby facilitating the evaluation of their influence on the final rankings of alternative experimental trials. The CoCoSo method then uses these weights to calculate the optimal sequence. Fig. 4Application of MEREC-CoCoSo methodology for multi-optimization of EDM process parameters. Application of MEREC-CoCoSo methodology for multi-optimization of EDM process parameters. Results and discussion Statistical analysis of MRR, TWR and SR The present study employs analysis of variance (ANOVA) to assess the impact of various machining parameters on the resultant outcomes. The regression equation for each response is developed to model the relationship between the response and input variables. The significance of the variables is evaluated at an accuracy level of 95%, indicating the model terms can produce the desired results27. The machining variable has a substantial influence on MRR, as evidenced by a P-value < 0.05 (Table 6). The F-value of the model is 115.57, which underscores its overall significance. In this model, terms with P-values below 0.05 were deemed insignificant. The significant terms, characterized by values exceeding 0.1, include BC, A2, and C2. The maximum F-value of 22.99 was observed for BC, indicating that the combination of Ton and Vg is the most dominant factor, followed by the Ip-Vg combination. Table 6ANOVA analysis for MRR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel0.07298.042E-003115.57< 0.0001significantA-Ip0.03610.036523.83< 0.0001significantB-Ton1.128E-00311.128E-00316.210.0101significantC-Vg9.453E-00319.453E-003135.85< 0.0001significantAB2.250E-00412.250E-0042.730.1321AC2.250E-00412.250E-0043.230.1321BC1.600E-00311.600E-00322.990.0049significantA20.01210.012170.09< 0.0001significantB21.256E-00411.256E-0041.810.2368C29.541E-00319.541E-003137.12< 0.0001significantResidual3.479E-00456.958E-005Cor Total0.07314\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 99.52, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 98.66, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 93.61. Table 6ANOVA analysis for MRR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel0.07298.042E-003115.57< 0.0001significantA-Ip0.03610.036523.83< 0.0001significantB-Ton1.128E-00311.128E-00316.210.0101significantC-Vg9.453E-00319.453E-003135.85< 0.0001significantAB2.250E-00412.250E-0042.730.1321AC2.250E-00412.250E-0043.230.1321BC1.600E-00311.600E-00322.990.0049significantA20.01210.012170.09< 0.0001significantB21.256E-00411.256E-0041.810.2368C29.541E-00319.541E-003137.12< 0.0001significantResidual3.479E-00456.958E-005Cor Total0.07314\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 99.52, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 98.66, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 93.61. ANOVA analysis for MRR. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 99.52, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 98.66, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 93.61. Likewise, the ANOVA results for the TWR response are presented in Table 7. The Model F-value of 115.02 indicates the model is substantial. In this model, the terms A, B, C, AB, BC, A2, and C2 are identified as significant, while those with P-values > 0.1 are not significant. This analysis corroborates the findings of previous studies28. The term BC exhibits the highest F-value (84.64), suggesting that the interaction between Ton and Vg is the most impactful factor. Following this, the Ip-Ton interaction shows a substantial influence with an F-value of 451.60, particularly when TWR is regarded as a desirable attribute. Table 7ANOVA analysis for TWR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel0.06997.644E-003115.02< 0.0001significantA-Ip0.03010.030451.60< 0.0001significantB-Ton6.125E-00416.125E-0049.220.0289significantC-Vg2.628E-00312.628E-00339.550.0015significantAB1.225E-00311.225E-00318.430.0078significantAC4.000E-00414.000E-0046.020.0577BC5.625E-00315.625E-00384.640.0003significantA20.01710.017259.24< 0.0001significantB20.01110.011173.04< 0.0001C25.004E-00415.004E-0047.530.0406significantResidual3.323E-00456.646E-005Cor Total0.06914\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.42, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.65, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 92.63. Table 7ANOVA analysis for TWR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel0.06997.644E-003115.02< 0.0001significantA-Ip0.03010.030451.60< 0.0001significantB-Ton6.125E-00416.125E-0049.220.0289significantC-Vg2.628E-00312.628E-00339.550.0015significantAB1.225E-00311.225E-00318.430.0078significantAC4.000E-00414.000E-0046.020.0577BC5.625E-00315.625E-00384.640.0003significantA20.01710.017259.24< 0.0001significantB20.01110.011173.04< 0.0001C25.004E-00415.004E-0047.530.0406significantResidual3.323E-00456.646E-005Cor Total0.06914\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.42, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.65, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 92.63. ANOVA analysis for TWR. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.42, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.65, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 92.63. Subsequently, an ANOVA is applied to SR to further elucidate the factors influencing it, as detailed in Table 8. The Model F-value of 53.48 indicates that the model is statistically significant. Within this model, the terms A, C, AB, AC, BC, A2, and C2 are identified as significant. The analysis reveals that BC, representing the interaction between Ton and Vg, is the most influential factor, with an F-value of 292.50. In regression analysis, a high R2 value generally indicates that the model is both sufficient and effective at capturing the underlying patterns. Equation (19) represents the modelled equation for MRR results, with R2 = 0.9952 and a predicted R2 of 0.9361. The analysis reveals strong alignment, with an adjusted R² of 0.9866. Table 8ANOVA analysis for SR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel109.83912.2053.480.0002significantA-Ip2.4112.4110.560.0227significantB-Ton0.3510.351.530.2714C-Vg12.30112.3053.900.0007significantAB13.43113.4358.860.0006significantAC2.9612.9612.960.0155significantBC66.75166.75292.50< 0.0001significantA29.3719.3741.060.0014significantB20.5510.552.420.1802C21.6811.687.380.0419significantResidual1.1450.23Cor Total110.9714\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.97, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.12, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 90.45. Table 8ANOVA analysis for SR.SourceSum of SquaresdfMean SquareF-valuep-valueSignificant or notModel109.83912.2053.480.0002significantA-Ip2.4112.4110.560.0227significantB-Ton0.3510.351.530.2714C-Vg12.30112.3053.900.0007significantAB13.43113.4358.860.0006significantAC2.9612.9612.960.0155significantBC66.75166.75292.50< 0.0001significantA29.3719.3741.060.0014significantB20.5510.552.420.1802C21.6811.687.380.0419significantResidual1.1450.23Cor Total110.9714\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.97, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.12, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 90.45. ANOVA analysis for SR. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}^{2}$$\end{document} = 98.97, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{j}}^{2}$$\end{document} = 97.12, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}.}^{2}$$\end{document} = 90.45. In a similar vein, the TWR analysis (Eq. (20)) shows that the observed R² = 0.9842 aligns well with the predicted R² (0.9263) and the adjusted R² (0.9765). Additionally, the SR analysis (Eq. (21)) indicates that the observed R² = 0.9897 is in good agreement with the predicted R² of 0.9045 and the adjusted R² of 0.971. It is observed that the difference between the predicted R2 and adjusted R2 values is small in both MRR, TWR, and SR analyses, and they are within the acceptable range.19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{gathered} MRR = - 1.233\: - \:0.1057\:I_{p} \: + \:2.006\:T_{{on}} \hfill \\ \: + \:0.0699\:V_{g} \: + \:0.0022\:I_{p} ^{2} - 0.583\:T_{{on}} ^{2} \: - \:0.0005\:V_{g} ^{2} - \: \hfill \\ - \:0.0150\:I_{p} \: \times \:\:T_{{on}} \: + \:0.00015\:I_{p} \: \times \:\:V_{g} \: - \:0.0200\:\:T_{{on}} \:\: \times \:\:V_{g} \hfill \\ \end{gathered}$$\end{document}20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{gathered} TWR = 1.482\: - \:0.1489\:I_{p} \: - \:1.511\:T_{{on}} \: + \:0.02074V_{g} \hfill \\ + \:0.0027\:I_{p} ^{2} + \:5.581T_{{on}} ^{2} \: - \:0.0001\:V_{g} ^{2} \: \hfill \\ + \:0.0350\:I_{p} \: \times \:\:T_{{on}} \: + \:0.0002\:I_{p} \: \times \:\:V_{g} - \:0.0375\:\:T_{{on}} \:\: \times \:\:V_{g} \hfill \\ \end{gathered}$$\end{document}21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{gathered} SR = - 130.5\: + \:2.972\:I_{p} \: + \:356.4T_{{on}} \: + \:1.617\:V_{g} \:\: \hfill \\ - \:0.06372I_{p} ^{2} \:\: + \:38.7\:T_{{on}} ^{2} \: - 0.00675\:V_{g} ^{2} \hfill \\ - 3.665\:I_{p} \: \times \:\:T_{{on}} + \:0.01720\:I_{p} \: \times \:\:V_{g} \: - \:4.085\:T_{{on}} \:\: \times \:\:V_{g} \: \hfill \\ \end{gathered}$$\end{document} Fig. 5Residual analysis plots for MRR. Residual analysis plots for MRR. Fig. 6Residual analysis plots for TWR. Residual analysis plots for TWR. Residual plots play a crucial role in validating the satisfactory outcomes derived from ANOVA. The reliability of the ANOVA model is confirmed by examining the residuals, which reveal discrepancies between expected and observed responses. Figure 5(a), Fig. 6(a), and Fig. 7(a) demonstrate a linear correlation between normal probability and externally studentized residuals. It suggests that the data points should follow a normal distribution for optimal model performance. However, in the current study, the data curve is linear. Furthermore, Fig. 5(b), Fig. 6(b), and Fig. 7(b) illustrate the trends in the experimental data, with the residuals appearing random relative to the expected response. This scatter plot indicates an absence of a consistent pattern, highlighting the atypical nature of the residuals. The distribution plots in Fig. 5(c), Fig. 6(c), and Fig. 7(c) further validate the models’ suitability by examining the residuals. Figure 5(d), Fig. 6(d), and Fig. 7(d) reveal a random distribution pattern when residuals are plotted against the run number. These results are pivotal in determining the suitability of the model applied. Fig. 7Residual analysis plots for SR. Residual analysis plots for SR. Effect of Vg and Ton on MRR, TWR and SR Design of Experiments (DOE) is a systematic method for conducting experiments to improve production or operational processes. Here, the response surface methodology (RSM) is selected as the DOE. It is the combination of statistical and mathematical techniques used to analyze, model, and optimize process parameters. In this study, experiments were conducted following the Box-Behnken design (BBD) of RSM, incorporating 3 center points. This design was selected due to its ability to perform non-sequential experiments with a limited number of design points. Furthermore, this technique provides 3D and contour plots that graphically depict the relationships between influential input factors (i.e., Ip, Ton, and Vg) and the output performance (i.e., MRR, TWR, and SR). This approach was used to assess the intricate interdependencies among these three parameters comprehensively. It specifically examines the individual and combined effects of the process parameters on either MRR, TWR, or SR. The intricate interplay between critical EDM parameters, namely Vg and Ton, and the resulting machining characteristics (MRR, TWR, and SR) is graphically represented in Fig. 8. From Fig. 8(a), it is clear that the MRR initially increases and subsequently declines as the Vg transitions from 60 V to 80 V. This initial rise in MRR can be due to the potential difference across the inter-electrode gap, which intensifies spark energy and promotes more vigorous material melting, vaporization, and erosion from the workpiece surface29. However, beyond a certain threshold value of Vg, there is a slight reduction in MRR. It is due to increased electrode wear, which leads to unstable discharge conditions. Moreover, a less-concentrated discharge channel forms, which reduces the effective energy density at the workpiece30. Concurrently, MRR also demonstrates a marginal increment with increasing Ton from 0.2 to 0.4 s. This behavior is primarily due to the prolonged duration of each electrical discharge, allowing for a greater cumulative energy input per pulse. This extended thermal exposure facilitates more substantial material removal through melting and vaporization. Figure 8(b), the contour plot for MRR, demonstrates the regions of optimal MRR. It also illustrates the combined and interactive effects of Vg and Ton on this response. Furthermore, Fig. 8(c) illustrates a noticeable rise in TWR as the Vg increases from 60 to 80 V. This trend is typically associated with the sharp increase in spark energy at higher Vg values. It is due to the tool electrode, which has undergone more aggressive thermal erosion and wear mechanisms. Furthermore, TWR shows a sharp increase at Ton of 0.2 to 0.3 s. This can be attributed to the extended exposure of the tool electrode to the high-temperature plasma channel. Thus, it leads to increased thermal degradation and subsequent material loss from the tool surface. The contour plot in Fig. 8(d) provides a comprehensive visualization of the TWR response on varying Vg and Ton settings. Moreover, the SR, as shown in Fig. 8(e), sharply increases with Vg from 60 V to 80 V. Higher Vg values result in higher discharge energies and potentially larger, deeper craters on the machined surface. It is due to the more intense and voluminous material removal per spark, thereby deteriorating the surface finish. There is significant variation and a sharp increase in SR with Ton increasing from 0.2 s to 0.4 s. This is due to longer pulse durations, which lead to larger, deeper craters31. The contour plot in Fig. 8(f) is instrumental in visualizing the complex relationship between Vg, Ton, and SR. Multi-response optimization using integrated MEREC-CoCoSo methodology Determining the weight of each response by MEREC Using the MEREC methodology, the weight assigned to each response is systematically determined, as detailed in Sect. 2.4. Table 9 presents the weight values assigned to every response, as determined by the MEREC methodology. Table 9Removal effect (Ej) and weight (Wj) values of each response using MEREC.ObjectivesMRRTWRSREj2.7011.5191.660Wj0.4590.2580.282 Table 9Removal effect (Ej) and weight (Wj) values of each response using MEREC.ObjectivesMRRTWRSREj2.7011.5191.660Wj0.4590.2580.282 Removal effect (Ej) and weight (Wj) values of each response using MEREC. Fig. 8Surface and contour plots for MRR, TWR and SR. Surface and contour plots for MRR, TWR and SR. Application of CoCoSo multiobjective optimization method This section outlines the methodology for segmenting ranking alternatives using the CoCoSo approach. After generating the decision matrix, the normalized values for each response are computed and presented in Table 10 using Eqs. (11 and 12) and for both beneficial and non-beneficial criteria, respectively. Notably, the last two criteria (responses) enumerated in Table 10 represent non-beneficial attributes, for which lower numerical values are desirable. Conversely, the first criterion is identified as a beneficial attribute, with higher numerical values preferred. Following this, the values for the weighted sum (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}$$\end{document}) and the exponentially weighted product (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{i}$$\end{document}) are derived using Eqs. (13 and 14). Based on these values, the coefficients \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ia}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ib}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{ic}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{i}$$\end{document} are computed via Eqs. (15–17), and with their respective values tabulated in Table 11. Table 10Normalized values of responses according to CoCoSo methodology.Exp. No.Normalized MRRNormalized TWRNormalized SR10.4360.5610.73220.1071.0000.84330.3800.7980.44840.3890.6660.30650.3750.8090.76360.2250.6820.22970.4500.8510.49480.4880.8940.41490.4880.8940.504100.4500.3860.000110.0000.8941.000121.0000.1900.518130.8310.3640.686140.6710.3960.757150.8400.0000.780 Table 10Normalized values of responses according to CoCoSo methodology.Exp. No.Normalized MRRNormalized TWRNormalized SR10.4360.5610.73220.1071.0000.84330.3800.7980.44840.3890.6660.30650.3750.8090.76360.2250.6820.22970.4500.8510.49480.4880.8940.41490.4880.8940.504100.4500.3860.000110.0000.8941.000121.0000.1900.518130.8310.3640.686140.6710.3960.757150.8400.0000.780 Normalized values of responses according to CoCoSo methodology. Table 11Weighted, aggregated scores and ranking of alternatives using CoCoSo technique.Exp. No. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{S}}_{\boldsymbol{i}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{P}}_{\boldsymbol{i}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{a}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{b}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{c}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}}$$\end{document} Rank10.5512.4590.0703.4610.9242.095820.5452.3120.0673.3410.8772.011930.5082.3830.0683.2680.8871.9901040.4382.2660.0632.9610.8301.8241250.5972.5110.0733.6430.9542.190560.3452.0710.0562.5270.7411.5821470.5672.4730.0713.5210.9332.126780.5732.4710.0713.5390.9342.134690.5982.5160.0733.6510.9562.1954100.3071.4740.0412.0000.5471.22115110.5121.9710.0583.0020.7621.78613120.6542.4810.0733.8100.9622.2632130.6692.5870.0763.9301.0002.3391140.6242.5440.0743.7540.9722.2483150.6051.8560.0573.2250.7551.86611 Table 11Weighted, aggregated scores and ranking of alternatives using CoCoSo technique.Exp. No. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{S}}_{\boldsymbol{i}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{P}}_{\boldsymbol{i}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{a}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{b}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}\boldsymbol{c}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{\gamma\:}}_{\boldsymbol{i}}$$\end{document} Rank10.5512.4590.0703.4610.9242.095820.5452.3120.0673.3410.8772.011930.5082.3830.0683.2680.8871.9901040.4382.2660.0632.9610.8301.8241250.5972.5110.0733.6430.9542.190560.3452.0710.0562.5270.7411.5821470.5672.4730.0713.5210.9332.126780.5732.4710.0713.5390.9342.134690.5982.5160.0733.6510.9562.1954100.3071.4740.0412.0000.5471.22115110.5121.9710.0583.0020.7621.78613120.6542.4810.0733.8100.9622.2632130.6692.5870.0763.9301.0002.3391140.6242.5440.0743.7540.9722.2483150.6051.8560.0573.2250.7551.86611 Weighted, aggregated scores and ranking of alternatives using CoCoSo technique. Table 11 also presents the rankings of alternative experimental trials, differentiated by various weight allocations derived from the MEREC method. Based on these final rankings, the optimal parameters for achieving superior MRR with lesser TWR and SR results are identified. This table clearly indicates that experiment number 13 yields the most favourable combination of Ip, Ton, and Vg to attain the desired response values. Specifically, these settings involve an initial Ip of 30 A, a Ton of 0.3 s, and a Vg of 60 V. It results in achieving maximum MRR (0.32 g/s), minimal TWR (0.34 g/s), and SR (8.99 μm). Conversely, experiment number 10 is identified as yielding the least desirable process parameter combination when evaluated using the integrated MEREC-CoCoSo methodology. The MEREC-CoCoSo integrated approach demonstrates superior performance over traditional multi-criteria decision-making methods due to its dual-layered robustness. This mechanism ensures that the optimal parameter selection remains stable and valid even under varying preference scenarios. It provides a more reliable compromise between the conflicting objectives of high MRR and minimal TWR/SR23,31. Morphological characteristics of machined surface The validity of these experimental findings is substantiated through a comparative analysis. This analysis utilizes the optical microscopic images (Fig. 9 (a, c, e)), acquired at various Vg settings, and their corresponding 3D surface visualizations generated from the EDMed surfaces (Fig. 9 (b, d, f)). Experimental run 13, which achieved the highest ranking according to the novel MEREC-CoCoSo integrated methodology, is thus selected for detailed examination. The resulting EDMed surface is depicted in Fig. 9(a). As this configuration yielded the best performance, the EDMed surface shows a balanced distribution of peaks and valleys. It is confirmed from the 3D surface plot (Fig. 9(b)). This favorable outcome is attributed to the optimized Vg (60 V), which promotes enhanced stability in the EDM process. Conversely, the second-ranked experimental run showed that increasing Vg to 70 V yielded MRR and SR of 0.36 g/s and 10.78 μm, respectively. Consequently, the EDMed surface obtained under this condition appears rougher (Fig. 9(c)). Its corresponding 3D surface plot (Fig. 9(d)) vividly illustrates numerous peaks and irregularities on the machined surface. This is due to the generation of a more powerful spark, resulting from the intensified concentration of electrons and their increased collision frequency during the EDM process. Such conditions elevate the inter-electrode temperature, prompting greater material melting and vaporization. Thus, the larger craters form on the workpiece surface. Similarly, the rank 3 outcome, in which the Vg was further increased to 80 V (with a slightly higher Ton than in the preceding case), resulted in a comparatively smoother surface (Fig. 9(e)). Under these conditions, the MRR and SR values recorded were 0.27 g/s and 8.23 μm, respectively. The corresponding 3D surface plot (Fig. 9(f)) reveals an almost uniform surface with minimal irregularities. This unexpected improvement in surface quality despite an increased voltage gap might be attributable to various factors such as intermittent power supply disruptions, machine vibrations, or potential contamination of the dielectric medium. Fig. 9Optical microscopic images of EDMed surfaces and their corresponding 3D surface plots at different Vg (a) 6o V, (c) 70 V, and (e) 80 V. Optical microscopic images of EDMed surfaces and their corresponding 3D surface plots at different Vg (a) 6o V, (c) 70 V, and (e) 80 V. Furthermore, the surface morphology resulting from bio-dielectric EDM at varying Vg settings was meticulously examined through SEM (Fig. 10(a-c)). At a constant Ton of 0.3 s, the machined surface obtained with a Vg of 60 V exhibited markedly fewer surface flaws when compared to surfaces generated at Vg values of 70 V and 80 V. This optimally processed surface was characterized by an absence of craters, recast layers, and extensive porosity, as clearly illustrated in Fig. 10(b). Conversely, an uneven distribution of energy within the machining zone can compromise surface quality. It leads to the formation of deeper and broader craters due to concentrated thermal effects. An example of this detrimental effect is observed in Fig. 10(c). Based on these findings, a Vg of 60 V is concluded to be the optimal setting in the current research, as it strikes a favorable balance between minimizing surface irregularities and achieving an acceptable MRR. Fig. 10Surface morphology of EDMed surfaces at different Vg (a) 6o V, (c) 70 V, and (e) 80 V. Surface morphology of EDMed surfaces at different Vg (a) 6o V, (c) 70 V, and (e) 80 V. EDX analysis SEM micrographs of the top surface of the sample, processed at a voltage gap of 60 V, showing a very thin deposition layer, are presented in Fig. 11(a). Figure 11(b) delineates the specific area analyzed by EDX. The elemental distributions and spatial maps of Ti, Cr, Ni, and Fe on the workpiece surface are shown in Fig. 11(c-f) using color mapping. Concurrently, EDX analysis was conducted on the untouched top surface of this workpiece immediately after the EDM process. The resultant EDX spectrum is depicted in Fig. 11(g). It exhibits distinct peaks corresponding to various elements, thereby confirming the significant presence of C (20.15 wt%) and Ti (39.85 wt%) on the surface. The presence of Ni, Cr, and Fe elements further corroborates the transfer and deposition of the 3D-printed tool material (SS316L) onto the workpiece surface. The deposited layer comprises a higher concentration of Ni (19.89 wt%) than of Cr (9.80 wt%) and Fe (7.90 wt%). It is hypothesized that some carbonaceous particles, originating from the hydrocarbon dielectric oil, may have migrated to the machined surface. Subsequently, they underwent an allotropic transformation because of the spark discharge during the EDM process. Conclusions This study investigated the influence of neem biodiesel as a dielectric in the EDM of Ti-6Al-4 V. The primary objective is to identify optimal parameters to achieve a high MRR, minimal TWR, and SR. A meticulously crafted 3D printed SS316L electrode played a crucial role in evaluating its combined effect with the bio-dielectric. The following key findings were deduced from this investigation: The ANOVA revealed that the combination of Ton and Vg is the most critical factor affecting the MRR, TWR, and SR. The regression models show exceptional predictive accuracy, with R2 values of 0.9952, 0.9842, and 0.9897, all of which are statistically significant (p ≤ 0.0002).Multi-response optimization using the integrated MEREC-CoCoSo methodology identified experiment 13 as the most favourable machining parameter combination, i.e., Ip of 30 A, a Ton of 0.3 s, and a Vg of 60 V. These settings resulted in a superior MRR, coupled with minimal TWR and SR. The ANOVA revealed that the combination of Ton and Vg is the most critical factor affecting the MRR, TWR, and SR. The regression models show exceptional predictive accuracy, with R2 values of 0.9952, 0.9842, and 0.9897, all of which are statistically significant (p ≤ 0.0002). The ANOVA revealed that the combination of Ton and Vg is the most critical factor affecting the MRR, TWR, and SR. The regression models show exceptional predictive accuracy, with R2 values of 0.9952, 0.9842, and 0.9897, all of which are statistically significant (p ≤ 0.0002). Multi-response optimization using the integrated MEREC-CoCoSo methodology identified experiment 13 as the most favourable machining parameter combination, i.e., Ip of 30 A, a Ton of 0.3 s, and a Vg of 60 V. These settings resulted in a superior MRR, coupled with minimal TWR and SR. Multi-response optimization using the integrated MEREC-CoCoSo methodology identified experiment 13 as the most favourable machining parameter combination, i.e., Ip of 30 A, a Ton of 0.3 s, and a Vg of 60 V. These settings resulted in a superior MRR, coupled with minimal TWR and SR. Fig. 11SEM colour mapping with EDX plot of machined surface processed using neem bio-dielectric. SEM colour mapping with EDX plot of machined surface processed using neem bio-dielectric. Morphological characterization confirmed that a Vg of 60 V produced a balanced distribution of peaks and valleys, with significantly fewer surface flaws. It also demonstrates an absence of craters, recast layers, and extensive porosity, indicating optimized process stability.The integral use of a meticulously crafted 3D-printed SS316L electrode is pivotal to this investigation. It enables the comprehensive evaluation of its distinct material behavior and combined effects with the neem oil bio-dielectric on the EDM process. It is further evidenced by the confirmed transfer of its elemental constituents to the workpiece surface.EDX analysis confirms the transfer of material from the electrode to the workpiece. A significant amount of carbon (20.15 wt%) and titanium (39.85 wt%) is present on the machined surface, suggesting an allotropic transformation of carbonaceous particles due to the neem oil dielectric. Morphological characterization confirmed that a Vg of 60 V produced a balanced distribution of peaks and valleys, with significantly fewer surface flaws. It also demonstrates an absence of craters, recast layers, and extensive porosity, indicating optimized process stability. Morphological characterization confirmed that a Vg of 60 V produced a balanced distribution of peaks and valleys, with significantly fewer surface flaws. It also demonstrates an absence of craters, recast layers, and extensive porosity, indicating optimized process stability. The integral use of a meticulously crafted 3D-printed SS316L electrode is pivotal to this investigation. It enables the comprehensive evaluation of its distinct material behavior and combined effects with the neem oil bio-dielectric on the EDM process. It is further evidenced by the confirmed transfer of its elemental constituents to the workpiece surface. The integral use of a meticulously crafted 3D-printed SS316L electrode is pivotal to this investigation. It enables the comprehensive evaluation of its distinct material behavior and combined effects with the neem oil bio-dielectric on the EDM process. It is further evidenced by the confirmed transfer of its elemental constituents to the workpiece surface. EDX analysis confirms the transfer of material from the electrode to the workpiece. A significant amount of carbon (20.15 wt%) and titanium (39.85 wt%) is present on the machined surface, suggesting an allotropic transformation of carbonaceous particles due to the neem oil dielectric. EDX analysis confirms the transfer of material from the electrode to the workpiece. A significant amount of carbon (20.15 wt%) and titanium (39.85 wt%) is present on the machined surface, suggesting an allotropic transformation of carbonaceous particles due to the neem oil dielectric. These findings underscore the efficacy of neem oil as a sustainable dielectric fluid in the EDM process, offering a pathway to optimize machining performance for Ti-6Al-4 V alloys. The identified optimal parameters provide a practical guide for industrial applications, contributing significantly to the adoption of environmentally conscious manufacturing processes. This research provides an understanding of bio-dielectric-assisted EDM with a 3D-printed electrode for sustainable machining practices.