Machining Performance of AA2024/5Al 2 O 3 /5Gr Hybrid Composites under Al 2 O 3 Mixed Dielectric Medium

In this research work, AA2024/5Al 2 O 3 /5Gr hybrid composites fabricated through stir casting were machined utilising an electric discharge machine (EDM). Experiments were performed by varying current, pulse on time (POT), gap voltage (GV), and Al 2 O 3 powder concentration (PC). The experiments were designed using response surface methodology in which material removal rate (MRR), tool wear rate (TWR), and surface roughness ( R a ) were recorded as responses. The addition of Al 2 O 3 particles has a positive impact on MRR and Ra, whereas it has a negative impact on TWR. The interaction impact of process parameters (p-p) on responses was thoroughly analyzed using contour plots. A mathematical model was developed and validated for all the responses. The experimental results were compared with the predicted values. It was found that all the values have a maximum deviation of 3.5%. The ANOVA table reveals that the PC was the most inﬂuential factor followed by the current.


Introduction
Composites have a high strength-to-weight ratio and are the potential candidates for aerospace, marine, and defence sectors [1]. Powder metallurgy, casting, and in-situ fabrication were the distinct methods used for the fabrication of composites [2][3][4]. Liquid stir casting process was the most capable method for producing composites at a low cost. Attaining homogeneous distribution was the most challenging task in the manufacturing of composites [5]. In stir casting, preheated particles are mixed using a mechanical stirrer for a prescribed period of time at constant speed to achieve uniform distribution [6,7]. Machining of composites using conventional machining is a tedious task as the presence of hard ceramic particles causes excessive tool wear [8]. To overcome this problem, composites were machined using an unconventional machining technique. EDM, abrasive jet machining, ultrasonic machining, and electron beam machining are the distinct unconventional machining processes [9]. Of which EDM was utilized for machining harder materials with complex geometry [10].
Current, POT, pulse off time (POFT), voltage, and gap distance (GD) were the vital p-p of EDM [11]. e influence of electrical p-p on MRR, TWR, and R a on Inconel alloy was analyzed [12]. Mineral oil with a flash point of 82C was utilized as the dielectric liquid and copper as the electrode material. Positive polarity primes to increased MRR, whereas negative polarity precedents to lower SR values. e authors [13] etched at the effects of process factors on MRR and TWR in EDM of SKH 57 high speed steel using copper electrode material. MRR rose with Ip and peaked at approximately 100 s ton. As ton climbed, MRR decreased [14]. Current and POT are the most important factors for influences the MRR and SR, respectively [15]. e incorporation of foreign particles improves the machining performance of EDM [16]. Boron carbide (B 4 C), silicon carbide (SiC), aluminium oxide (Al 2 O 3 ), and graphite (Gr) are the distinct particles incorporated in the dielectric fluid [17]. EDM of tool steel (SKH-51) with graphite nanopowder mixed dielectric was studied by [18] utilising Gr, Al, and Al 2 O 3 nanopowders and developed an analytical model for EDM mechanism. According to [19], boron carbide (B 4 C) powder combined with kerosene and deionized water had a significant impact on microperformance EDM's. Tan et al. [20] investigated the machining performance of particle blended EDM and analyzed recast layer as the response. e EDM of cemented tungsten carbide utilising graphite nanopowder blended dielectric was investigated by [21]. It was reported by [22] that titanium powder combined with deionized water dielectric fluid was used to modify machined surfaces. e authors [23] investigated the impact of SiC mixed dielectric on MRR and TWR. Numerous works were reported by the researchers by adding powder particles in the dielectric fluid; however, literature linked to machining of AA2024/5Al 2 O 3 /5Gr hybrid composites was sparse. In this work, an attempt was made to machine AA2024/5Al 2 O 3 /5Gr hybrid composites by incorporating Al 2 O 3 particles into the dielectric fluid. e electrical process parameters altered were current, POT, and GV, in which the most influential factor was analyzed with the aid of an ANOVA table.

Materials and Methods
Aluminium alloy designated as AA2024, as procured from perfect metal alloys Bangalore, having a chemical composition as shown in Table 1, was kept in the electric furnace. e alloy was heated to 770°C. Preheated Al 2 O 3 and Gr particles purchased from Bhukhanwala Industries, Mumbai, were added to the melt and stirred at a speed of 1000 rpm for 180s. e equal weight proportion of magnesium powder flux was added to the melt and, once again, the mixture was stirred for 120s. e mixture was poured into the preheated mould made of die steel. Composites of dimension (L 105 mm X ? 12 mm) were fabricated and machined to the dimension of (L 100 mm X ? 10 mm) to eliminate the surface defects. Experiments were performed on the E2K Elektra die sink EDM machine by varying current, POT, GV, and PC. e copper of dimension (L 25 mm X ? 10 mm) used as the tool material was purchased from Coimbatore Metal Mart. A new setup was laid to conduct the experiments. It comprises of dielectric tank, stirrer, pump, and motor [24] in which Al 2 O 3 particles of an average size of 5 µm were added to improve the machining performance. Machining characteristics were accessed in terms of MRR, TWR, and R a .
Experimental runs were designed using response surface methodology in which each parameter was varied at five different levels, as shown in Table 2.

Calculation of Response.
e MRR and TWR were calculated according to equations (1) and (2), respectively, which is the ratio of weight difference before and after machining to the machined time. e surface roughness was calculated at 5 different places using a Mitutoyo tester, and the average value was recorded. e most influential factor was analyzed using ANOVA table; a developed mathematical model was used for the computation of predicted values.

MRR
H b and H a -sample weight before and after machining. I b and I a -electrode weight before and after machining. z-machined time.

Impact of p-p on MRR.
e presence of reinforcing particles and its hard nature make it difficult for machining. e experimental results are shown in Table 3. Materials are removed by means of melting and vaporization. e impact of various p-p on MRR is given in Figure 1. e mean MRR under pure dielectric conditions was 55.75 mg/min, but when 2.5 g of Al 2 O 3 particles were added to the fluid, it rose to 83.75 mg/min, and then to 87.11 mg/min when 5 g/l was added. e improvement in MRR was attributed to two factors: (1) the suspended particles get energized and move in a zig-zag fashion when the voltage is applied which facilitates the bridging effect [25].
It generates an enormous amount of heat, resulting in an improvement in MRR. (2) e addition of powder reduces the insulating strength and results in early dielectric break down [26], hence sparking occurs most frequently. With further increase in PC to 7.5 g/l, the MRR decreases. e suspended particles are packed densely inside the spark gap which hinders the machining process. e particles were not completely flushed away, resulting in the remelting of particles over the surface [27]. e MRR reduces from 101.91 mg/min to 81.16 mg/min when there is a shift in current from 7A to 16A. At 7A current, the plasma channel distributes the heat intensity uniformly over the surface [28]. With further increase in current, this plasma channel gets widened, resulting in a reduction of heat intensity, which leads to a decrease in MRR [29]. e MRR decreases until the saddle point of 45 V, thereafter it starts to reduce. With an increase in voltage beyond that, it accelerates the spark energy, which leads to excessive heat generation, hence more materials removed from the surface. At mean parametric value of voltage, more materials redeposit over the surface, hence MRR decreases. e POT has the very least impact on MRR; the reduction in MRR from 87.08 mg/min to 81.41 mg/min was observed when there was an increase in pulse on from 3µs to 7µs owing to the plasma expansion. e interaction effect of various p-p on MRR is shown in Figure 2.
For the PC of 5 g/l, at 7 A current, a MRR of 101.32 mg/ min was obtained. For the same parametric condition, conventional dielectric fluid offers a MRR of 78.33 mg/min. It further reduces to 63.86 mg/min when the value of current was increased to 19 A. When 5 g/l powder for lower parametric pulse on value offers 44% lower MRR in comparison with high parametric values. e combined increase of GV and PC drastically reduces the MRR by 40%. e optimal combined parametric combination was 15 V and 5 g/l. It was clearly noted that an increase in GV, current, and POT reduces the MRR. e ANOVA Table 4 denoted that PC was the most influential factor followed by current and GV. e maximum difference of 3.5% was observed between experimental and predicted results, and its mathematical model is shown in equation (3).  International Journal of Chemical Engineering

Impact of Process Parameter on TWR.
During machining, generated heat was transferred to both the tool and the work piece. In spite of that materials were removed from the both tool and the work piece. e objective of the EDM industry was minimizing the TWR and maximizing MRR. e effect of various p-p is shown in Figure 3. e TWR was minimum when 2.5 g/l of Al 2 O 3 were added to the dielectric fluid, and it increased with further increase in the concentration of foreign particles. When powder is added, the energized particles owing to the bridging effect reduce the spark gap. To maintain the machining condition, a slight increase in GD happened, which led to the complete dissipation of heat which reduced the TWR. At higher PC, the TWR increases owing to bridging and arcing effects [30].
With an increase in current from 7 A to 19 A, a 5% reduction in MRR was observed. It was anticipated that more heat was transferred to the work piece and also confirmed the increasing of spark gap. TWR results were diametrically opposed to MRR results in terms of GV. e minimum TWR of 18.04 mg/min was recorded at the GV of 15 V, and it was drastically increased to 22.13 mg/min at 45 V. e results confirmed that most of the heat generated was transferred to the tool material, resulting in an increase in TWR. e TWR was minimum when current was applied for the longer duration of time. At higher POT, the widening of plasma channel occurs, resulting in a reduction in TWR. e maximum difference of 4.3% was observed between experimental and predicted results, and its mathematical model was shown in equation (4). e interaction effect of p-p on TWR is shown in Figure 4.
Under pure dielectric medium for the 19 A current, a minimum TWR of 19.56 mg/min was obtained, and it was increased to 27.25 mg/min when the PC was increased to 10 g/l. In the case of the combinatorial effect of PC and POT, 16.92 mg/min and 27.25 mg/min TWR were recorded for the concentrations of 0 g/l and 10 g/l, respectively. With regards to POT, it offers higher TWR on low parametric value and it decreases with an increase in duration of time, irrespective of PC. Same case was reported on GV, at higher voltage, pure oil offers 17.89 mg/min and it was drastically increased to 27.55 g/min when 10 g of Al 2 O 3 was mixed in the dielectric fluid. e cumulative increase of voltage and pulse on leads to reduction in TWR; the same results were observed at lower parametric value and 20% increases in TWR was observed at intermediate value. At lower parametric values of current and POT, 16.73 mg/min were recorded and a 20% increase in TWR was observed when there was a swift change in parametric value from lower to higher. e interaction effect of pulse and current had the least impact on TWR; a minor substantial change was observed. e ANOVA Table 5 shows that PC was the most impactful factor followed by GV.
A mathematical model was developed as shown in equation (4); it showed the experimental and predicted values were well allied.
3.3. Impact of p-p on R a . Products with a better surface finish offers the best performance along with the highest life time. e results showed that the R a values decrease until the saddle point of 5 g/l thereafter it starts to declined, as shown in Figure 5.   e improvement in R a was attributed to the fact that suspension of particles increases the GD. is facilitates the complete flushing of machined debris from the spark gap [31]. It eliminates the formation of remelted layer and globules on the surface [32]. e powder particles assist in the uniform distribution of heat over the machined surface which resists the formation of craters, cracks, and pits on the surface [33]. When the powder particles weighed more than 5 g/l were added, machined debris and powder particles were densely packed inside the spark gap, resulting in incomplete flushing and leads to increase in R a [33]. e R a value of 1.85 µm was observed when the parametric value of current was tuned to 7 A. Owing to its low parametric value, it generates heat of low intensity, hence R a increases. A 15% improvement in R a was attained when there was a reduction in GV from 15 V to 45 V. At this parametric value, most of the generated heat is transferred to the tool material which reduces the heat intensity and hence improves surface quality. e POT, either at higher or lower parametric, yields the least R a value. e maximum difference of 4.8% was observed between experimental and predicted results, and its mathematical model was shown in equation (5).
e combined parametric effect of p-p on R a is shown in Figure 6. e minimum surface roughness value of 1.79 µm was observed when the current and PC were set at 7 A and 5 g/l respectively. Under unmixed dielectric conditions, the surface exhibited an Ra value of 6.62 µm when the current value was tuned at 19 A. For the PC of 5 g/l, the minimum surface roughness of 2.74 µm was attained when the POT was tweaked at 3 µs and it was increased to 7.99 µm, when the PC was increased to 10 g/l. e optimal R a value of 3.22 µm was recorded for the PC and voltage was set at 45 V and 5 g/l, respectively. e surface quality  International Journal of Chemical Engineering 7 worsens drastically to 10.16 µm when these values are increased to 75 V and 10 g/l. e best surface quality was expressed when the parametric values of pulse on and voltage was set at 45 V and 7 µs, respectively. Interestingly, a best surface quality of 0.21 µm was obtained for the input variables of 7 A current and 7 µs POT. e most influential factor was identified with the aid of ANOVA, as shown in Table 6. e PC was the most impactful factor followed by the POT. e mathematical model was developed as shown in equation (5), and it was found that predicted values were well correlated with the experimental results.

Conclusion
e AA2024/5Al 2 O 3 /5Gr hybrid composites were successfully fabricated using the stir casting technique. EDM experiments were performed on the composites by adding Al 2 O 3 in the dielectric medium, and the following findings were drawn from the studies.
(1) Incorporation of Al 2 O 3 particles in the dielectric medium improves the MRR owing to the bridging effect and uniform heat distribution. Adding particles beyond the saddle point leads to insufficient flushing which hinders the machining process, hence MRR reduction. (2) Adding Al 2 O 3 particles has a negative impact on TWR. Owing to arcing and short circuits, most of the generated heat was transferred to the tool, resulting in an increase in TWR. e interaction plot reveals that current and POT have very minimal impact on TWR.
(3) Inclusion of Al 2 O 3 particles improves the R a of the machined surface due to the enlargement of the spark gap and complete flushing of materials. e incorporation of particles facilitates the uniform heat distribution and heat dissipation which results in the improvement of R a . (4) A mathematical model was developed for all the responses, and influential factor was identified using the ANOVA table. e experimental results were compared with the predicted results, and it was found that the deviation was below 5%.
Data Availability e datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
ere are no conflicts of interest.