Machining of nickel-based super alloy inconel 718 using alumina nanofluid in powder mixed electric discharge machining

Inconel 718 is a nickel-based superalloy with exceptional mechanical qualities, including high tensile and impact strength, as well as good oxidation and corrosion resistance. EDM is widely used to cut hard materials of this type by common parameters affecting spark, current, duty cycle etc, whereas MRR, SR can be greatly improved by using sufficient fluid properties of the dielectric. Incorporation of alumina nanoparticles with deionized water may improve machining performance, with no prior research on machining characteristics of Inconel 718 using alumina nanofluid in EDM. The machinability of Inconel 718 via the electrical discharge machining process, as well as the performance of alumina nanofluid as a dielectric medium, have been evaluated using deionized water with a concentration of 0.5%. In this study, based on the effect of various input parameters such as Ip, Pon and Gv on the MRR, TWR and SR, and the RSM has been used to conduct the experiments. A maximum MRR of 0.048 g min−1, a TWR of 0.00632 g min−1 and an SR of 3.271 μm are achieved. Overall, the use of alumina nanofluid can improve machining performance in EDM, due to the reduction of abnormalities such as crack formation and molten metal debris in the machined surface of Inconel 718.


Introduction
Non-conventional machining overcomes the limitations of traditional machining in terms of material processing quality and prompts industries to shift to non-conventional machining. At present, non-traditional machining is generally used for better machining of particularly hard materials [1,2]. The optimization of nonconventional machining attracts many investigators because of its superior material removal rate (MRR). Electric discharge machine (EDM) is widely used to cut hard materials such as Inconel 718 [3] at which electrical conductivity occurs and is affected by spark, current, duty cycle, pulse on time, gap voltage, electrode material, workpiece, and fluid properties of the dielectric [4][5][6]. In the past few decades, Powder mixed electric discharge machine (PMEDM), dielectric with various nano and microparticles were studied to increase the MRR and improve the quality of the surface roughness SR [7]. It is also observed that employing nanoparticles in the dielectric medium increases the MRR and improves the quality of the SR while decreasing the cracks and porosity of the workpiece [8]. However, EDM has some drawbacks, such as the amount of time spent in EDM machining. In addition, SR is also a major issue when using high discharge [9]. A study investigated the performance of EDM with graphite powder in kerosene where it was reported that MRR increased by 60% and tool wear rate (TWR) by 15% [10]. The various additives to improve machining performance on EN31 steel using EDM, at 7.5 amps, 20 μsec pulses on time, 100 μsec pulse off-time and 700 MPa compaction pressure maximum 1.4 mg min −1 deposit rate achieved [11]. An experiment performed on alumina and TiC with kerosene as the dielectric medium where it was reported that the variation in the current accounted for the observed effect on the surface finish [12]. Apart from this, analyzed the effect of input parameters on nickelbased superalloy Inconel 718 with Additive mixed electric discharge machine (AEDM) where the One variable at a time (OVAT) method has been used to analyze MRR, SR and wear ratio (WR). Furthermore, it is observed that the machining of Inconel 718 was mostly affected by the nanoparticle concentration and size [13]. The machining performance of superalloys with SiC microparticles was studied in EDM, where it was observed that PMEDM offers better performances such as higher MRR, lower TWR and better SR than EDM [14]. Inconel 718 was machined in kerosene with multi wall carbon nano tube (MWCNT) in EDM, where a significant improvement in machining performance was observed [15]. EDM machining performance was studied with graphene nanofluid for Inconel 718 where the machining performance was improved by using nanofluid in EDM [16]. EDM process was studied for optimization using grey relational analysis where based on the ANOVA test, it was concluded that current has a significant effect on MRR after pressure [17]. AISI D2 die steel machining effect was investigated with silicon PMEDM, where optimal conditions for MRR were found in terms of current, pulse on time, pulse off time [18]. Graphite PMEDM saw to optimize the machining of Inconel 718 where it was concluded that graphite PMEDM shows low micro-hardness and residual stress on the machined surface [19]. An experimental analysis was performed for abrasive mixed surface discharge of Inconel 718 using RSM where it was concluded that the wheel speed parameter is the most important influence for MRR [20]. The machining of composite (Al-B 4 C) in EDM was studied with input parameter of current, pulse on current ( ) T , on pulse off time ( ) T off and electrode material where it was found that current is the major input parameter for optimum performance [21]. PMEDM also improves the machining performance of Aluminum metal matrix composite [22]. EDM analysis with MWCNT additive using response surface methodology (RSM) was performed for the machining of Al-10% SiCp Metal matrix composite (MMC), where it is observed that the MRR and SR improved by 38.22% and 46.06%, respectively [23].
The single carbon nanotube (SWCNT) using PMEDM were performed for AISI D2 Tool Steel where it was noted that the use of SWCNT additives in EDM leads to enhanced surface finish and reduction of micro-cracks [24]. EDM analysis using CNT additives was performed on workpieces of copper, copper tungsten and silver tungsten with brass tools where it is observed that both MRR and surface finish are improved by using the additives [25]. The effect of CNT additives on EDM for stainless steel was demonstrated with a tungsten tool where it is observed that using the additives improves MRR and reduces black spots [26]. Inconel 825 was investigated with SWCNT in dye sinking EDM, where it is observed that using SWCNT additives results in better surface finish, better MRR, and fewer cracks relative to SWCNT without [27]. The effect of tool rotation were performed using EDM and observed a uniform and improved MRR with improved surface finish [28]. The effect of SiC additives was studied on EDM machining for MMC Al 6061 where it is observed that timed pulse results in an increase in heat generation which leads to an improvement in MRR and also showed that the additions to MRR and SR provide better results [29]. The additive effect on EDM was studied and found that 'particle size, concentration, density, thermal conductivity and electrical conductivity' were the most influential characteristics of PMEDM [30]. Apart from this, various methods of flushing in EDM for optimization of machining of AISI 5160 where the generated correlation with Non-dominating sorting genetic algorithm (NSGA) found a more accurate prediction for MRR and SR relative to RSM method [31]. Thermophysical properties of nano particles added in the dielectric significantly affect the machining performance as surface characteristics of machined surface [32]. Coated and uncoated electrode material used in the machining process influence the machine performance [33]. The study was also performed for the optimization of EDM using RSM for machining AISI D2 tool where optimal input parameters for improved MRR and SR were explained [34].
Through numerical simulation and experimental testing, many researchers have done significant work in the field of EDM. As seen from these researches, EDM machining nanofluids have several advantages over conventional fluids and are widely used in both conventional and non-conventional machining operations. This is because the nanoparticles increase the specific surface area, and they improve the nanofluid thermal conductivity, resulting in higher surface finishes. In addition, the nanoparticles can improve the electric current properties of the nanofluid, thereby improving the EDM machining performance. Apart from this, Inconel 718 is an improved super alloy known for its particularly high strength to tensile and impact strength. This nanofluid can further improve machining efficiency by reducing crack propagation and molten metal debris formation when mixed with alumina nanoparticles. However, no research has yet been reported on the machining characteristics of Inconel 718 using alumina nanofluid in EDM. Thus this prompted the authors to explore the use of nanofluid, by which the machining performance of EDM could be improved. So, it is important to investigate the use of nanofluid in EDM machining when deionized water will mix with alumina nano particles. Therefore, Inconel 718 via the electrical discharge machining process, as well as the performance of alumina nanofluid as a dielectric medium, is explored as a novelty in this study. The experiment is designed and conducted based on RSM which has been used on inconel 718 using three input parameters (Peak current in (A)( ) I , p Pulse on time in (μs) ( ) P , on and Gap voltage in (V) ( ) G . v In addition, the possible experiment results of MRR, SR and TWR have been investigated which are expected to be useful for the current work industries.

Materials
Alumina (Al O 2 3 ) nanoparticles are the preferred choice for nanofluid preparation due to mass production maturity, low cost and environmental friendliness. Therefore, many laboratories are looking at alumina-based nanofluids, and they have been identified as one of the best prospects for nanofluid-based coolants [35,36]. Alumina nanofluid is prepared by mixing alumina nanopowder of size 30-50 nm (Parshwamani Metals, India) in deionized water with ultrasonication, where these nanoparticles come in various sizes, with the smallest surface area of sphericity. Furthermore, the average size of the particle according to the SEM image is approximately 42 ± 10 nm spherical in shape as shown in figure 1(a). Therefore, spherical nanoparticles have been chosen in this study. The specific heat, viscosity and thermal conductivity of the nanofluid were 3035 J Kg −1 .K , 6.564 mPa.S and 0.4019 W m −1 .K, respectively. The breakdown voltage of alumina nanofluid has been measured using Megger an automatic break down voltage (BDV) oil tester to be approximately 20 kV mm −1 which is lower than that of deionized water. The reduction in BDV is due to the good bridging effect and the low density of the alumina nanoparticle. This bridging effect is also shown in figure 1(b) in the BDV oil tester for alumina nanoparticle. Therefore, alumina based nanofluids are extensively investigated in many laboratories and are recognized as one of the best candidates for nanofluid based coolants [35,37]. Water-based Al2O3 nanofluids are reported to show superior thermal performance (ranging from 2 to 10%) with a particle size of 40 nm [38].

Methods
The alumina nanofluid and Electronica CNC S50 die sink EDM machine have been used for the experimental analysis. Figures 2 and 3 show the experimental setup of PMEDM. It consists of an EDM control panel, tool holder, work holding fixture, working tank, and dielectric supply. Alumina nanopowder (30-50 nm) mixed with deionized water using ultrasonication for the dielectric supply unit. The experiment was carried out using a workpiece made of Inconel 718 with dimensions of 33 40 150 mm .
3´I nstrument electrodes are manufactured from 99% pure copper and have a length of 55 mm, with diameters of 12 mm. Copper was chosen as a tool material because of its excellent thermal and electrical conductivity, ease of fabrication, and low cost.    a) and (b) shows the tool and workpiece before and after machining. To maintain the concentration level of the powder, manual mechanical stirrer shown in figure 3(b) is used to avoid settlement of powder particles. Furthermore, the microstructure of these samples has also been observed using SEM for differentiation with and without the use of alumina nanofluid. In nano powder mixed electric discharge machine (NPMEDM), the I , p P , on and G v input parameters are widely used which has been chosen for this study. Other characteristics like as polarity, duty cycle (50%) and pulse off time (vary according to the pulse on time because the duty cycle = (Ton/(Ton + Toff)) is constant), are fixed. Based on the experiments, I , p P , on and G v were chosen as input parameters. Tables 1 and 2 show the values of the input parameters for experimental analysis. Further, the response parameters were chosen to be MRR, SR,   and TWR. The initial and end weights of the work material were measured with a model SHIMADZU AUW220D. Before and after machining, the work piece sample shown in figure 4 is weighed, and the MRR (gm min) −1 is calculated using the following equation:

= -
Here, T is the machining time and w 1 and w 2 are initial and final weight of the workpiece in machining.
Similarly, TWR (gm min) −1 can be calculated by using equation (2), in which w t1 and w t2 are the initial and final weight of the tool electrode. Most surface finish standards required for engineering are supported by the Stylus profilometer [40]. SR was  The experiments are conducted according to L21 orthogonal array and the results are shown in table 3.

Response surface modelling (RSM)
RSM is one of the most popular and commonly used methods for designing experiments nowadays. The RSM is a generalised quadratic model that establishes a relationship between input (peek current, gap voltage and pulse on time) and output (MRR, TWR and SR) parameters. Table 3 shows the experimental pattern of response values. Currently, the mathematical models were created using Minitab-20 software and were based on EDM experimental data. Studies were conducted based on the regression equations (3)-(5) to determine the impact of input factors on MRR, TWR and SR, respectively.

Microstructure characterization and Recast layer thickness
The use of alumina nanofluid can also be observed for differential visualization, compare with no alumina nanofluid as shown in figure 9. This figure clearly shows abnormalities such as excessive microcrack formation and molten metal debris compared to alumina nanofluid. This powder particle breakdown causes the voltage of the dielectric to drop, which distributes the energy due to the alumina powder particle in the spark gap and produces multiple sparks in a single pulse over time. Whereas in case of deionized water without adding powder particles it crates less area volatile and more concentrated spark which affects the surface characteristics. Therefore, alumina nanofluid mixed dielectric powder gives better surface characterization than without adding particles. This is also due to the more organized distribution of alumina nanoparticles with a mixture of deionized water in the machined surface of Inconel 718. Hardness and sub-surface damage can be reduced due to high heat output dispersion for the best case with nano fluids, while hardness and sub-surface damage due to low heat production for the worst case without nano fluids may increase [41]. But in both cases, the hardness of the machined surface obtained with (found to be the best) and without (found to be the worst) nano powder is higher than that of the base material. The recast layer is formed due to the deposition of molten metal or debris on the machined surface, where its thickness depends on several factors such as dielectric properties, machining parameters and the powder particle used. In case of powder composite dielectric, the layer thickness is less compare to without using powder particle [42]. The same results were found in this study with nanofluid reducing the recast layer thickness compared to without using nanofluid as shown in figure 10. According to the analysis, the thickness of the recast layer is 1-2  μm for NPMEDM and 4-5 μm for conventional EDM. This is due to multiple sparks in a single spark over a large area so the intensity of energy is distributed evenly [43]. Furthermore, adding nanofluid breakdown voltage reduces the dielectric and increases the gap between the electrode and the workpiece due to this molten metal or debris that is easily removed from the machined surface.

Effect of input parameters on MRR
The I , p P , on and G v are shown to be process parameters for MRR after findings were analysed. In figure 11, the influence of input parameters on MRR was displayed using the regression equation in equation (3). MRR enhances as the peak current ( ) I p increases, as shown in figure 8(a). From figure 11(a) (at P s 30 ; ), when peak current ( ) I p is low as 2 A, the MRR is low due to little energy input. With an increase in I p from 2 A to 6 A, the discharge energy increased, generating a lot of heat and forming a lot of craters. As a result, a further increase up to 10 A occurs after the removal of the higher material to obtain the maximum MRR [44]. Further, from figure 11 ), the MRR initially increases with P on from 10 s m to 30 s m than start decreasing from 30 s m to 50 s. m The increased P on causes a higher heating effect at the machined surface. As a result, additional debris was formed, reducing machining performance due to the arc and short circuit -identical results were achieved by Tzeng and Lee [45]. It can be seen in figure 11(c) that when G v grows, MRR increases as well, peaking at V 30 and then decreasing. The increased gap voltage resulted in an increase in energy per spark, resulting in higher MRR when all other parameters were held constant, i.e., I A 10 ; p = P s 30 on m = [13]. Figures 12(a)-(c) shows response surface plots of MRR against input parameters. Figure 12

Effect of input parameters on SR
A regression equation, given by equation (4), was created to investigate the impacts of the input parameters i.e., I , p P , on and G v on the Sr figures 13(a)-(c) depicts the effect of important input parameters on Sr As presented in figure 13(a), as the I p (at P s 30 ; ) increases, the SR decreases. The discharge energy increases with I , p resulting in huge craters. As a result, roughness is formed over the machined surface. The roughness is small at a low value (2 A) of I , p presented in figure 13(a). The roughness further increases with I p form 2 A to 10 A. This is owing to the enormous magnitude of crater formation induced by higher heat generation, which resulted in an increase in SR [46]. Figure 13(b) illustrates that the variation of SR with P on (at I A 10 ; p = It is noticed that the SR is minimal at a low value of P on and increases with increasing P on values from 10 s m to 50 s, m as discharge energy increases with each P . on Therefore, lowest SR detected at a lower value of P on and greater SR at a higher value of P on [23]. The G v also is a significant influence over Sr Further, from ), the SR initially increases with G v from 10 V to 40 V than start decreasing from 40 V to 50 V . However, with greater G , v the insufficient cooling of the work material creates unfavourable concentrated discharges, resulting in a higher SR [13].
SR response surface plots against input parameters are shown in figures 14(a)-(c). Figure 14(a) illustrates the mutual effect of I p and P on on SR where the higher I , p and higher P on observed in higher Sr Further, from

Effect of input parameters on TWR
To analyse the influence of the input process parameters i.e., I , p P , on and G v on TWR, by equation (5). Figures 15(a)-(c) depicts the effect of important input parameters on TWR. It is shown from figure 15(a) that as ) increased, the TWR firstly reduced and then improved. Further, it is noted from figure 15(b) that with an increase in P on (at I 10 A; p = G 30V v = ). The TWR decreases with an increase in P on (at I 10 A; p = G 30V v = ). As a result, for smooth machining, a greater P on is preferable [13]. Furthermore, figure 15(c) shows that the variation of TWR with gap voltage ( ) G v (at I 10 A; p = P 30 s on m = ) where the TWR is observed as minimal at low G v values and grows as G v rises from 10 V to 50 V and minimum TWR is obtained at 10 V [13]. Figures 16(a)-(c) shows response surface of TWR against input parameters. Figure 16(a) illustrates the mutual effect of I , p and P on on TWR where it is noted from this figure that lower I p and higher P on found in lower TWR while low TWR were found at lower I p and higher P . on Further, figure 16(b) illustrates the mutual effect of G , v and I p on TWR where it is noticed that the low TWR were obtained at lowest I p and highest and lowest value of G . v However, it is also seen from figure 16(c) that combined effects of G , v and P on on TWR where lowest TWR are obtained at highest G v and highest value of P . on

Conclusion
The experimental study has been carried out on Inconel 718 machined by NPMEDM. The impact of I , p P , on and G v on MRR, SR, and TWR were investigated where all process parameters were relevant. Further, MRR and SR, the peak current ( ) I p was obtained as the most significant parameter whereas I , p P , on and G v were significant parameters for the TWR. The maximum MRR of 0.048 g min ( on v m = = ) are obtained in NPMEDM, which was about 10% and 6% respectively, lower than that of EDM. Overall, the reduced crack formation and the molten metal debris-like abnormalities in the machined surface of Inconel 718 with the use of alumina nanofluid have resulted in better machining performance in EDM.