Optimization of Process Variables in the Drilling of LM6/B4C Composites through Grey Relational Analysis

The objective of this investigational analysis was to study the influence of process variables on the response during the drilling of LM6/B4C composite materials. Stir casting was employed to produce the LM6/B4C composites. A Vertical Machining Center (VMC) with a dynamometer was used to drill the holes and to record the thrust force. An L27 orthogonal array was used to carry out the experimental work. A grey relational analysis (GRA) was employed to perform optimization in order to attain the lowest Thrust Force (TF), Surface Roughness (SR) and Burr Height (BH). For minimal responses, the optimum levels of the process variables viz. the feed rate (F), spindle speed (S), drill material (D) and reinforcing percentage (R) were determined. The process variables in the drilling of the LM6/B4C composites were indeed optimized, according to confirmational investigations. The predicted Grey Relational Grade was 0.846, whereas the experimental GRG was 0.865, with a 2.2% error—indicating that the optimization process was valid.


Introduction
Composites are typically comprised of a polymer matrix, metal matrix or ceramic matrix, based on the physical and chemical properties of the matrix. Metal matrix composites (MMCs) provide numerous benefits over metallic alloys, which include their greater functional strength, high resistance to wear and low coefficient of thermal expansion [1][2][3][4]. The production of composites has therefore been a field of major concern over the last thirty years. The goal in the manufacture of composites is to attain a variety of characteristics that can never be attained in any one of the components alone. The production of composites thus provides an opportunity to modify the characteristics of individual chosen constituents to suit particular requirements. Aluminium Matrix Composites (AMCs) are a type of aluminium-oriented material system that is light and high-performing [5]. Due to their unique properties, composites are commonly used in industry [6]. MMCs have recently received lot more attention than typical metallic materials, and they are now widely used in a variety of commercial applications because of their good strength, light weight, high resistance to wear and exceptional mechanical characteristics [7][8][9][10][11].
The industrial sector demands ultra-lightweight products for load-bearing applications. Therefore, it is essential to substitute with components that would not lead to heavy weights, but which are highly compatible with current economic circumstances, in order to meet current industrial demands. Throughout the field of composites, specifically in AMCs, all these requirements are properly followed. Due to their less-costly existence, AMCs have a wide series of applications in the vehicles, buildings, factories and sports sectors. The most crucial component that affects their characteristics is the physical characteristics of the

Materials and Methods
The materials used in this study were chosen based on their properties, cost and use. Three sets of composites were fabricated in the form of plates with dimensions of 100 mm× 100 mm × 10 mm and a reinforcement weight percentage of 3, 6 and 9, respectively. The aluminium alloy LM6 was also cast for comparison studies with the composites. The composition of the LM6 alloy is depicted in Table 1 [28]. Stir casting is a fluid state process of material manufacturing, where reinforcement is combined with a melted matrix (LM6 alloy) by virtue of mechanical stirring [29]. Figure 1 displays the stir casting setup used for the fabrication process.

Fabrication of the LM6 Alloy/Boron Carbide (B 4 C) Composites
LM6 alloy ingots were heated to 850 • C and melted in an electric furnace. The molten aluminium was mixed well to generate a vortex, and the second phase material (B 4 C) was preheated to (250 • C) and then added. The oxide film that was formed on the surface of the molten metal was removed by adding Titanium (Ti) in the form of Potassium Hexa Fluro Titanate (K 2 TiF 6 ; this improves the wettability between the Aluminium alloy and Boron Carbide). The slurry was mixed well at 600 rpm for 10 min for a homogeneous distribution of the ceramic particles. The composite plates, with dimensions of 100 mm× 100 mm× 10 mm, were obtained by transferring the molten mixture to 650 • C pre-heated cast iron moulds and allowing it to solidify. rials 2022, 15, x FOR PEER REVIEW Figure 1. Fabrication of the LM6 alloy using a stir casting setup.

Fabrication of the LM6 Alloy/Boron Carbide ( ) Composites
LM6 alloy ingots were heated to 850 °C and melted in an ele aluminium was mixed well to generate a vortex, and the second preheated to (250 °C) and then added. The oxide film that was the molten metal was removed by adding Titanium (Ti) in the Fluro Titanate (K2TiF6; this improves the wettability between th Boron Carbide). The slurry was mixed well at 600 rpm for 10 distribution of the ceramic particles. The composite plates, with 100 mm× 10 mm, were obtained by transferring the molten mixt cast iron moulds and allowing it to solidify.

Drilling
Drilling MMCs is a useful machining method. The studies a VMC with predetermined process parameters. Data from the ex and recorded using a computer-based data collecting device [30 with a Kistler dynamometer. HSS, Carbide and TiN-coated carb ent drill materials used. For all three drills, the diameter, point a 6 mm, 118° and 30°, respectively. Images of the VMC and drills Figures 2 and 3. The parameters were chosen based on relevan studies. Table 2 displays the drilling variables and their levels.

Drilling
Drilling MMCs is a useful machining method. The studies here were carried out on a VMC with predetermined process parameters. Data from the experiments was captured and recorded using a computer-based data collecting device [30]. The TF was measured with a Kistler dynamometer. HSS, Carbide and TiN-coated carbide were the three different drill materials used. For all three drills, the diameter, point angle and helix angle was 6 mm, 118 • and 30 • , respectively. Images of the VMC and drills utilized are presented in Figures 2 and 3. The parameters were chosen based on relevant research as well as trial studies.

Grey Relational Analysis
Multi-criteria decision-making (MCDM) tools are extreme compromise solutions [31][32][33][34][35][36][37][38]. GRA is much simpler than most similar task, obtaining balanced solutions when multiple confli sent. GRA was carried out to optimize the process variables in o mum TF, SR and BH. The various steps involved in GRA are d experimental data were initially standardized on a scale of 0 to coefficient (GRC) was found with respect to the normal investiga was found by averaging the GRC with respect to the TF, SR and with the highest GRG (Rank 1) was the optimum.

Grey Relational Analysis
Multi-criteria decision-making (MCDM) tools are extremely popular for obtaining compromise solutions [31][32][33][34][35][36][37][38]. GRA is much simpler than most MCDMs but performs a similar task, obtaining balanced solutions when multiple conflicting responses are present. GRA was carried out to optimize the process variables in order to obtain the minimum TF, SR and BH. The various steps involved in GRA are described below: (1) The experimental data were initially standardized on a scale of 0 to 1 (2) The grey relational coefficient (GRC) was found with respect to the normal investigational data (3) The GRG was found by averaging the GRC with respect to the TF, SR and BH. The process variable with the highest GRG (Rank 1) was the optimum.

Thrust Force (TF) Measurement
The response TF was measured in real time utilising a computer-controlled data collecting system and a KISTLER dynamometer, type 9257 B. The TF created while drilling was correspondingly transformed into voltage signals and the data was collected in the form of a graph using the KISTLER Dynoware software. Figure 5 depicts the usual TF encountered when drilling hybrid MMCs, with the hole-making process divided into six stages. The first stage is the drill's entry into the MMC, where the TF is minimal, and the second stage is the initiation of the drilling process. The third stage is the penetration of the drill into the specimen, where the hole-making process is completed. The fourth stage denotes the hole-making phase, in which the highest TF is displayed. The fifth stage is the tool's exit through the back side of the specimen, indicating a significant drop in force. The sixth stage is the drill's exit from the specimen, where there is no TF. The TF signal acquired in this study was nearly identical to that obtained in previous studies; the variance detected was only in the centre of the hole [39].

Thrust Force (TF) Measurement
The response TF was measured in real time utilising a computer-controlled lecting system and a KISTLER dynamometer, type 9257 B. The TF created while was correspondingly transformed into voltage signals and the data was collecte form of a graph using the KISTLER Dynoware software. Figure 5 depicts the usual TF encountered when drilling hybrid MMCs, with making process divided into six stages. The first stage is the drill's entry into th where the TF is minimal, and the second stage is the initiation of the drilling proc third stage is the penetration of the drill into the specimen, where the hole-making is completed. The fourth stage denotes the hole-making phase, in which the high displayed. The fifth stage is the tool's exit through the back side of the specimen, in a significant drop in force. The sixth stage is the drill's exit from the specimen, wh is no TF. The TF signal acquired in this study was nearly identical to that obtained ous studies; the variance detected was only in the centre of the hole [39].

Surface Roughness (SR) Measurement
The average SR was measured using an SR tester (Surfcorder SE 3500). The da gathered thrice at various positions on the drilled surface, with the average of th

Surface Roughness (SR) Measurement
The average SR was measured using an SR tester (Surfcorder SE 3500). The data was gathered thrice at various positions on the drilled surface, with the average of the three readings used to calculate the SR. Figure 6 shows the SR tester that was utilised in this study. The direction of a calculation of SR was normal to the drilled hole.

Measurement of Burr Height
A burr is a plastically deformed material created at a drilled hole's in These burrs lead to many problems with product performance and quality can interconnect during component assembly and can end up causing a ja At the end of a cut, the development of burrs is equivalent to that of chips is essential, as it is bigger in size and hard to remove and is likely to ca issues. Figure 7 shows the experimental setup for measuring BH. A typica using a video measurement system (VMS) is shown in Figure 8. The drill specimen was kept under the VMS. The exit burr did not have the sam points, so the BH was measured at five different places and the average v for the analysis.

Measurement of Burr Height
A burr is a plastically deformed material created at a drilled hole's inlet and outlet. These burrs lead to many problems with product performance and quality because they can interconnect during component assembly and can end up causing a jamming effect. At the end of a cut, the development of burrs is equivalent to that of chips. The exit burr is essential, as it is bigger in size and hard to remove and is likely to cause deburring issues. Figure 7 shows the experimental setup for measuring BH. A typical BH observed using a video measurement system (VMS) is shown in Figure 8. The drill exit side of the specimen was kept under the VMS. The exit burr did not have the same height at all points, so the BH was measured at five different places and the average value was used for the analysis.

Microstructure
The main focus of the microstructural analysis was to establish th dispersion of the second phase material in the matrix. The composite mat examined using light microscopy. The material was taken out from eac precise polishing work was carried out to achieve a reflective surface fin rial. The optical images of the given composite demonstrate the unifor matrix second phase material; the microstructures are displayed in Figu  Figure 9a reveals the microstructures of the LM6 aluminium alloy. T particles look like spikes and scripts, owing to the higher percentage of silico in the LM6 alloys. Figure 9b reveals the microstructure of the AMCs, with Boron carbide in the LM6 alloy with 3% B C. The composite particles were due to the higher fluidity of the LM6, with a higher silicon content and a low Figure 9c,d exhibit the microstructures of the MMCs with 6% and 9% B C; content of reinforcement particles and the uniform distribution can be seen.

Microstructure
The main focus of the microstructural analysis was to establish the homogeneous dispersion of the second phase material in the matrix. The composite material Al/B 4 C was examined using light microscopy. The material was taken out from each composite and precise polishing work was carried out to achieve a reflective surface finish on the material. The optical images of the given composite demonstrate the uniform mixture of the matrix second phase material; the microstructures are displayed in Figure 9.

Hardness
The hardness of the Al LM6/B C composites and the LM6 alloy was tested b Rockwell hardness test scale E (1/8-inch steel ball with a 10 kgf minor load and 9  Figure 9b reveals the microstructure of the AMCs, with the dispersion of Boron carbide in the LM6 alloy with 3% B 4 C. The composite particles were in close proximity due to the higher fluidity of the LM6, with a higher silicon content and a lower melting point. Figure 9c,d exhibit the microstructures of the MMCs with 6% and 9% B 4 C; an increase in the content of reinforcement particles and the uniform distribution can be seen.

Hardness
The hardness of the Al LM6/B 4 C composites and the LM6 alloy was tested by the Rockwell hardness test scale E (1/8-inch steel ball with a 10 kgf minor load and 90 kgf major load) [1]. It can be seen from Figure 10 that the composite hardness was considerably more than that of the LM6 alloy. Furthermore, the composite's hardness is known to increase with rises in the wt. percent of boron carbide [40]. The surface areas of the composites were enhanced by implementing the second phase material (B 4 C) into the composite material as well as by reducing the grain sizes of the continuous phase (Al); this prevents plastic deformation. The grain boundaries rose to the highest point, and the misalignment of the atoms was decreased by maximisation of the weight percentage of the second phase material, which increased the matrix strength and thereby increased the composite's hardness. Kalaiselvan

Density Measurement
Displacement techniques are used to measure the density of a material [42]. A 0.001 g-precision electronic weighing balance was used to determine the sample masses. The displaced liquid was used to calculate the volume of the samples. The effect of B 4 C on the density is shown in Figure 11.
The composite density dropped owing to a rise in the mass percentage of the boron carbide [43]. The reason for this is that boron carbide densities (2.52 g/cm 3 ) are smaller than LM6 alloy densities. Monolithic B 4 C ceramic is a very hard, solid and stiff material with a low density. the density is shown in Figure 11.
The composite density dropped owing to a rise in the mass perc carbide [43]. The reason for this is that boron carbide densities (2.5 than LM6 alloy densities. Monolithic B C ceramic is a very hard, sol with a low density.

Grey Relational Analysis ( 6/ )
The GRA is used to optimize the process parameters for LM6/B ing into account the different performance factors of the drilling pro impact of drilling variables when drilling using VMC, performance f D and R were chosen. Experiments were carried out using the suitab For the twenty-seven experiments, the GRG was calculated and p Practically, the greater the grade of the gray relationship, the closer th to the optimum value. Therefore, for optimum performance, a greater process variables' optimum levels are those with the maximum GRG

Grey Relational Analysis (LM6/B 4 C)
The GRA is used to optimize the process parameters for LM6/B 4 C composites, taking into account the different performance factors of the drilling process. To analyse the impact of drilling variables when drilling using VMC, performance features such as F, S, D and R were chosen. Experiments were carried out using the suitable L 27 OA.
For the twenty-seven experiments, the GRG was calculated and presented in Table 3. Practically, the greater the grade of the gray relationship, the closer the product would be to the optimum value. Therefore, for optimum performance, a greater GRG is needed. The process variables' optimum levels are those with the maximum GRG.
The experiments were ranked according to their GRG values. Figure 12 clearly shows that the GRG value was high for the 1st level of 'F', the 3rd level of 'S', the 3rd level of 'D' and the 1st level of 'R'. The experiments were ranked according to their GRG values. Figure 12 clearly shows that the GRG value was high for the 1st level of 'F', the 3rd level of 'S', the 3rd level of 'D' and the 1st level of 'R'. An ANOVA was used to establish the suitable outcome of the process variables in this study by examining their relative contributions to the response [44,45]. The experimental plan carried out was analysed at a 95% level of confidence. Table 4 summarizes the ANOVA of the GRG results. The obtained value for the GRG was 93.1%. The pvalue was smaller than 0.05 for 'F' and 'R', which indicates that they had a significant  An ANOVA was used to establish the suitable outcome of the process variables in this study by examining their relative contributions to the response [44,45]. The experimental plan carried out was analysed at a 95% level of confidence. Table 4 summarizes the ANOVA of the GRG results. The obtained R 2 value for the GRG was 93.1%. The p-value was smaller than 0.05 for 'F' and 'R', which indicates that they had a significant impact on the GRG. Besides the p value, it is also possible to use an F-test to decide which process variables have a major impact on the response. As per the Fischer's F-test, if the obtained F-ratio value is more than the tabulated F-value, it is regarded as significant [46][47][48][49]. It is evident from Table 4 that the F-test values for 'F', 'S', 'D' and 'R' were larger than the F-Table values, and hence they had a significant impact on the GRG. Among the chosen process variables, the drill material (51.44%) had the maximum contribution on the GRG, followed by the spindle speed (23.29%), feed rate (12.92%) and reinforcement percentage (5.45%). The contributions of the interactions were much less, and so they were pooled with the error term.
The ANOVA table shows the relative contribution of each process variable on the GRG values. The degrees of freedom (DoF) are obtained by subtracting one from the number of levels. The total DoF is the number of experiments minus one. SS means the sum of squares. The Mean Square (MS) is given by its sum of squares divided by its degrees of freedom; this is also known as the variance, which is why the test is known as the analysis of variance (ANOVA).

Confirmation Experiments
To determine the optimum parameters, the investigational outcomes were analysed. From Figure 12, the optimum process variables for attaining the maximum GRG (minimum TF, SR and BH) were the variables at levels F 1 , S 3 , D 3 , R 1 , which were an 'F' of 50 mm/min, an 'S' of 3000 rpm, a TiN-coated drill bit and a 3% B 4 C particulate. The predicted GRG was 0.846, whereas the experimental GRG was 0.865. A good agreement attained with respect to the predicted and experimental values can be seen, and the error was 2.2%, so the methodology of optimization held well.

The Effects of Drilling Process Variables on the GRG
The effects of the process parameters on the GRG ( Figure 13) are based on response Table 5. The GRG values of all 27 experiments were calculated and the level 1 value in Table 5  'S'. This is because friction between the drill bit and the specimen is decreased when the 'F' is decreased, resulting in a lower TF. A lower 'F' reduces the temperature generated during drilling, which improves the surface quality. It was observed that a lower 'F' gave a lower TF, which gave a good surface finish at a smaller feed. Due to the rapid heat rise generated by friction at greater spindle speeds, the work piece softens and penetrates smoothly, leading to a smaller TF. As spindle speed increases, the cutting time is reduced, which results in a reduced thrust force and reduced work piece distortion and, hence, the surface finish is improved; this provides better GRG values. The SR value for 3% B C was less than that of the other two wt %. In the LM6/B C composite material, the TF value was larger with the addition of B C. The physics behind this phenomenon is that B C is the hardest material-so by increasing the B C%, the composite's hardness rises as well, which results in a larger TF. The SR of the LM6/B C de-  For example, the feed rate value of 50 m/min was the average of the first nine values (Ex No. 1-9); in the same way, the feed rate value of 100 m/min was the average of the second nine values (Ex No. [10][11][12][13][14][15][16][17][18] and the feed rate value of 150 m/min was the average of the third nine values (Ex No. [19][20][21][22][23][24][25][26][27]. For example, the spindle speed value of 1000 rpm was the average of Ex Nos. 1, 2, 3, 10, 11, 12, 19, 20 and 21.
The GRG was employed as a quality representation of all the responses, including the TF, SR and BH. The process effect of the variables on the GRG is shown in Figure 13. The highest value of the GRG response graph suggests that drilling variables had a stronger impact on machinability features [50,51]. The peak value of the GRG was obtained at an 'F' of 50 mm/min, an 'S' of 3000 rpm, a 'D' of TiN-coated and an 'R' % of 3 wt. %; this are the optimal process parameters for drilling.
It was observed that the highest GRG was obtained at the lowest 'F' and highest 'S', which implies that the response TF, SR and BH were minimal at the lowest 'F' and highest 'S'. This is because friction between the drill bit and the specimen is decreased when the 'F' is decreased, resulting in a lower TF. A lower 'F' reduces the temperature generated during drilling, which improves the surface quality. It was observed that a lower 'F' gave a lower TF, which gave a good surface finish at a smaller feed.
Due to the rapid heat rise generated by friction at greater spindle speeds, the work piece softens and penetrates smoothly, leading to a smaller TF. As spindle speed increases, the cutting time is reduced, which results in a reduced thrust force and reduced work piece distortion and, hence, the surface finish is improved; this provides better GRG values.
The SR value for 3% B 4 C was less than that of the other two wt %. In the LM6/B 4 C composite material, the TF value was larger with the addition of B 4 C. The physics behind this phenomenon is that B 4 C is the hardest material-so by increasing the B 4 C%, the composite's hardness rises as well, which results in a larger TF. The SR of the LM6/B 4 C decreases initially when the weight percentage of the reinforcement increases.
As the 'F' rises, so does the cutting force and BH. For high feed rates, the BH is the largest. The BH decreases with rises in the reinforcement % for LM6/B 4 C composites. The GRG value is based on the average of the GRC of the TF, the GRC of the SR and the GRC of the BH of each experiment; so, the GRG value depends on the response TF, SR and BH.

Conclusions
Drilling experiments on LM6/B 4 C composites were conducted to analyse the impact of drilling process variables such as the feed rate, spindle speed and drilling material, as well as the reinforcement percentage. The experimental and statistical analyses led to the following results: • LM6/B 4 C composites were prepared by the low cost Stir casting method.

•
The uniform distribution of the second phase material in the matrix was confirmed by Optical micrographs.

•
The densities of the LM6/B 4 C composites decreased with rises in the wt. % of the B 4 C, whereas the hardness increased with increases in the reinforcement.

•
Drilling experiments were conducted on LM6/B 4 C composites using Taguchi's DoE and analysed using Grey relational analyses.

•
The TF, SR and BH values decreased with decreases in the feed rate for all the specimens.