Assessment of machinability behaviour of LM24 – nano Al2O3 – gr hybrid composites through stir casting technique

This study is focused on developing a novel combination of a hybrid MMC made of LM24 aluminium alloy, nano Al2O3 and Graphite for enhanced machinability. The study conducts a series of experimental runs and evaluates surface roughness and material removal rate. The process parameters of spindle speed, feed, and depth of cut have been varied, while different weight percentages of nano Al2O3 (1%, 3%, and 5%) were incorporated into the LM24 metal matrix alloy along with 1 wt% of Gr. Response Surface Methodology (RSM) is used to optimize the turning parameters and enhance the machining performance for better quality and productivity. The derived optimal machining parameters have been verified through confirmatory tests. An analysis of variance (ANOVA) was used to determine the individual contributions of each parameter to the machinability characteristics. Surface morphology analysis revealed the uniform distribution of nanoparticle reinforcements in the metal matrix. The surface roughness and material removal rate of the machined nanocomposites were studied. The optimized outputs are a minimum surface roughness Ra 0.522 μm and maximum material removal rate MRR 110.2 mm3s–1 The maximum hardness of the MMC composites has been 60.5 HRB.


Introduction
Metal matrix composites (MMC) are tailor-made materials with a metal matrix and one or mo reinforcements that may contribute to the specific requirements.MMCs are mainly made for lightweight applications with high specific strength, high fatigue strength, wear resistance, creep resistance, electrical conductivity, etc As material processing is also essential, machinability is another demanded characteristic.Hybrid MMCs [1] are developed by adding more than one reinforcement material to achieve the desired properties.Ceramic reinforcements such as Al 2 O 3 , SiC, TiO 2 , etc, will increase the hardness and wear resistance [2].The wear resistance leads to poor machinability characteristics.Hence, to balance the hardness and the machinability, reinforcements such as graphite (Gr) and MoS 2 are added [3,4].Researchers have used MoS 2 [5], eggshell [6], graphite [7], etc, along with reinforcement materials such as SiC [8], B 4 C [5], HSS [9], ZrB 2 [10] as to improve the machining characteristics.In particular, Gr reinforcement plays a vital role due to its high thermal and electrical conductivity, low friction, low density and low wear.Kathirvel and Palanikumar [11] studied the effect of SiC reinforcement in AA6061+SiC+Gr MMCs on surface roughness.The results show that the surface roughness is reduced by increasing the % of SiC reinforcement, and there is an improvement at higher speeds and lower feed rates.Bharat et al [12] investigated the effect of SiC and TiO 2 nanoparticles and found that the feed has the greatest effect on the surface roughness.The machinability studies conducted by Siddesh Kumar et al [5] on AA2219-nano B 4 C-MoS 2 MMCs show that the microhardness increases from 72 to 104.Also, cutting force and surface roughness were increased due to nano B 4 C reinforcement, which cannot improve machinability.Aluminum MMCs reinforced with SiC and B 4 C are more rigid and more wear-resistant when compared with Al 2 O 3 reinforcement [13].In the context of MMCs reinforced with SiC or B 4 C having poor surface roughness, as reported by Kathirvel and Palanikumar [11] and Siddesh Kumar et al [5], Al 2 O 3 can be a better reinforcement to improve machinability characteristics.The reinforcement of 5 vol % nano alumina particle is equivalent to 35 vol % in the micron size range, and hence, nano alumina is the preferred reinforcement to improve the strength [14].Machinability characteristics have been improved by reinforcing with SiC+Gr in MMC made of LM25 aluminium alloy [15].The Grey Fussy Algorithm optimized the turning parameters and improved machining performance in terms of measuring surface roughness, flank wear and material removal rate (MRR).The overall machining performance is improved, showing a higher gray-fuzzy reasoning grade of 0.891.The grey-fuzzy method has been employed by Gnanavelbabu et al [8] in the parameter optimization for machining the hybrid composites made of AA6061+B 4 C+CNT.Aruna [16] reported RSM studies for the optimization of cutting parameters.Masooth et al [17] and Ravikumar and Suresh [18] implemented Taguchi and ANOVA approaches for machinability studies on AA6061/AA7075 hybrid MMCs reinforced with SiC and alumina.Ajin et al have studied the machining characteristics of AA7075-based hybrid MMC and optimized the input parameters using a modified PROMETHEE approach [19].These studies suggest adopting experiment design and optimization design in the machinability studies.Adding Al 2 O 3 in fabricating MMCs with AA6070 has increased hardness, and adding Gr made MMCs soft [20].The mechanical and wear properties can be tailored in AA7075 MMCs by the suitable reinforcement % of Al 2 O 3 and Gr [21].Machinability studies on MMCs made of aluminium alloys with Al 2 O 3 and Gr are rare.Summarising the literature, it can be observed that significant research has been carried out on optimizing parameters and machining characteristics of aluminium metal matrix composites with various reinforcement materials.Regarding LM24 aluminium alloy composites, limited research has been conducted on composite machining studies and design experiments.So far, the machining characteristics of LM24-nano Al 2 O 3 -Gr hybrid metal matrix composites have not been reported.
Hence, the present study attempts to develop a novel combination of MMC with LM24, nano-Al 2 O 3 and Graphite for enhanced machinability.Al 2 O 3 is preferred to SiC and B 4 C to balance wear resistance and machinability as already reviewed above.In particular, nano-Al 2 O 3 is a newer attempt towards this goal.Gr is also a better selection as compared to MoS 2 and egg-shell for the enhancement of machinability, thus making a hybrid nano MMC.Due to the involvement of multiple process parameters, DOE and RSM [22] became essential.

Materials
The materials used for the present investigation are the base matrix metal of the LM24 alloy and reinforcement materials of nano-Al 2 O 3 and Graphite (Gr).These materials were purchased from Coimbatore Metal Mart, Coimbatore.The average size of the nano aluminium oxide powder was 40 nm.Graphite is a self-lubricating material that is reinforced to enhance its machining properties [23].

Fabrication of LM24 -nano Al 2 O 3 -Gr hybrid composites
In order to achieve the superior properties of MMCs for machinability studies, the stir casting technique is preferred to make LM24-nano Al 2 O 3 -Gr hybrid composites due to their low manufacturing cost with highquality composites [13].The chemical composition of LM24 aluminium alloy [24] is shown in table 1.The ingot of LM24 is cut into the slices of 25 mm × 50 mm and preheated in the graphite crucible up to 200 °C to remove oil and moisture.The sliced samples of LM24 aluminium alloy were placed in a graphite crucible and melted in an electrical resistance furnace at 800 °C.Reinforcement particles of Al 2 O 3 nanoparticles and Gr were preheated to 300 °C for 10 min in a muffle furnace to remove the impurities.The Al 2 O 3 nanoparticles (various wt% viz.1%, 3% and 5%) and Gr (1% in wt) were added through the hopper arrangement to the molten aluminium alloy material during the formation of a vortex in the melt via mechanical stirring.The molten metal is stirred at 400 rpm for 5 min at 800 °C before being put into the warmed metallic die.The molten state mixture is poured into the preheated die by the bottom pouring facility, and the pressure is applied through a squeeze casting attachment.The squeeze casting technique removes the porosity in the cast composites.The cast specimen is solidified at room temperature before removal from the die.

Preparation of specimens
Composite specimens are prepared with three different compositions, i.e., Sample 1: LM24 + 1 wt% nano Al 2 O 3 + 1 wt% Gr, Sample 2: LM24 + 3 wt% nano Al 2 O 3 + 1 wt% Gr, and Sample 3: LM24 + 5 wt% nano Al 2 O 3 + 1 wt% Gr .All the samples are prepared to a size of 30 mm in diameter and 300 mm in length as shown in the mould in figure 1.Then the unwanted portions are removed and surfaces are cleaned before investigation.

Experimental studies
Turning operations were performed on the center lathe (Make: Kirloskar; Model: Turn master 35) with variable spindle speed, depth of cut and feed rate to analyze the machinability studies.Center lathe machined specimen with tool insert of polycrystalline diamond (PCD) SNMG brand with a nose radius of 0.8 mm has been used in the machining operation.Figure 2 shows the methodology of the research work, and figure 3 shows the machining experimental setup and its measurements.

Measurement of surface roughness
The surface roughness Ra of the machined surface was measured using the Mitutoyo surface roughness (Model: SJ 210).The measurements were repeated three times at three different places on the turned surface, and the average of these three values was taken as Ra.

Measurement of hardness
Rockwell hardness tests were conducted according to the ASTM standard, 1/16′ ball indenter with an applied load of 100 kgf and 15 s of dwelling time.Testing was performed on polished specimens, readings were taken in three places, and average values were reported.The average hardness results of reinforced MMC composites are tabulated in table 2. The unreinforced pure matrix has a hardness of 44 HRB.

Surface morphology analysis
Field emission-scanning electron microscopy (FE-SEM) instrument (Make: ZEISS; Model: RA-ZEI-001) was used to investigate the microstructure of the composite specimens.The specimens were cut into cubes of size 10 mm, and composite specimens were polished using emery paper grades ranging from 500 to 2000.After polishing, specimens were etched with Keller's reagent (95 ml H 2 O, 2.5 ml HNO 3 , 1.5 ml HCl, 1.0 ml HF).The surface morphology of the MMC composite was analyzed with an operating accelerating voltage of 20 kV, a working distance of 8.6 mm to 13.7 mm, and a magnification value ranging from 250x to 500x with a resolution of 1.5 mm. Figure 3 shows the FE-SEM image of fabricated composites and section 4.1 describes the same.

Design of experiments
Design of experiments (DOE) can potentially be used to effectively minimize the number of runs required for a more significant and realistic design challenge.These techniques have been effectively coupled with response   surface models (RSMs) to optimize actual (rather than simulated) issues.As a result, this study used a response surface approach for DOE since it requires fewer tests.Instead of random trials, a predetermined set of tests with precise cutting settings was carried out so that the process could be further analyzed for different response characteristics separately or in combination [25].An L27 orthogonal array of Box-Behnken RSM design was utilized for the turning operation.Machinability is affected by several parameters, including spindle speed, feed, depth of cut, tool geometry, work/tool materials, and cutting circumstances.Spindle speed, feed, depth of cut, and weight % of nano alumina were selected as regulating parameters.These variables were chosen from research paper on machining optimization issues and machinability [26].Table 3 shows the machining parameters and levels employed as control variables.27 turning operations were completed and subsequently, the specimen's surface roughness (Ra) and material removal rate (MRR) have been recorded.For lathe turning operation, the material removal rate of  Clustering occurs due to the early solidification of aluminium alloys, leading to increased reinforcing content excluded by the solid-liquid interface.Agglomeration is a phenomenon that occurs when there is an increase in the quantity of reinforcing particles inside the matrix [28].The presence of elements is indicated in EDS image of sample 1 as in figure 5.

Rockwell hardness (HRB)
Figure 6 shows the hardness (HRB) values of pure LM24 matrix samples 1, 2, and 3 before and after machining.It is noticed that the significant increase in hardness, i.e., 16.8%, 23.4% and 30.4%, respectively, due to the reinforcements of 1, 3 and 5 wt% of nano Al 2 O 3 reinforcements along with 1 wt% Gr in the LM24 matrix.The hardness values are also increased due to strain hardening on the surface of the samples during the machining operations.The percentage increases in hardness due to the machining operation are 5.2, 7.36 and 5.4 for Samples 1, 2 and 3 respectively.A similar range of hardness values is noticed in experimental studies investigated by Kotteda et al [29].The maximum hardness of Sample C: LM24 + 5 wt% nano Al 2 O 3 + 1 wt% Gr Composite is 37.5% greater than that of the pure matrix.The present result indicates that complex nano Al 2 O 3 was observed in the grain structures of MMC composites and confirmed with FE-SEM analysis.Also, nano Al 2 O 3 reinforcement provided better inter-bonding between them, which led to an arrest of grain boundary dislocation movement and improved hardness.

Mathematical model for surface roughness
The notations of input process variables A, B, C and D are as indicated in table 3.For the output response, namely surface roughness, statistical parameters have been observed as p < 0.0001, adjusted R 2 = 0.9466 and predicted R 2 = 0.8600.The difference between the two values of R 2 is < 0.   ( ) The above quadratic model predicts another 27 solutions corresponding to the supplied values as in table 4. Figure 7 shows the error between the actual roughness and predicted roughness.As the points lie very much along the 45°perfect line, the model is confirmed to be significant.Though the parameters A, C, AC and C 2 have  evolved significantly, A is the most important, with a p-value <0.0001 and more for others.This fact is supported by the one-factor graph shown in figure 8.It shows that the surface roughness is highly sensitive to the wt% of nano alumina.However, it is to be kept in the background that this observation is under the conditions of fixed values of B, C, and D indicated in the graph.To this effect, the software also displays a warning in the graph.Figure 8 also shows that the model is with 95% confidence level.The similar displays of the software show that the surface roughness is least sensitive to the depth of cut (B) and feed (D).The surface roughness is somewhat sensitive to the spindle speed (C) but not as wt% of nano alumina (A).that for the surface roughness to be minimal, wt% of n-alumina must be less than 3 for all speeds and feeds, respectively.Figure 12 shows the 3D surface contour for surface roughness plotted with A versus B. Almost the same information as from figure 9 may be inferred from this contour, for the fixed values of C = 840.5 rpm and D = 0.09 mm rev -1 .

Mathematical model for MRR
For the output response, namely material removal rate (MRR), statistical parameters have been observed as p < 0.0001, adjusted R 2 = 0.9982 and predicted R 2 = 0.9867.The software has suggested the 2FI model.Moreover, the ANOVA for the 2FI model has resulted in F = 1418 with p < 0.0001.It means that the chance for the occurrence of this much F-value is only 0.01%.Adequate precision is 136.42 > 4, indicating a proper signal.The 2FI model above, predicts another 27 solutions corresponding to the supplied values as in table 4. Figure 13 shows the error between the actual MRR and the predicted MRR.As the points lie very much along the 45°perfect line, the model is confirmed to be significant.All the parameters B, C, D, BC, BD and CD have evolved significantly.This fact is supported by the one-factor graphs shown in figure 14.In figure 14, the plots    show that the MRR is sensitive to the depth of cut, spindle speed and feed.MRR is insensitive to wt% nano alumina.It is to be noted that this trend is contrary to the one observed for surface roughness, where wt% nano alumina was the most significant.it is to be remembered that these observations are under fixed values of A, B, C, and D, as indicated in the figure.To this effect, the software also displays warnings in the graphs.From figures 15-17, it may be observed that, for the MRR to be maximum, spindle speed, feed and depth of cut need to be maximum.Figure 18 shows the 3D surface contour for MRR plotted with wt% nano alumina versus depth of cut.Almost  the same information as from figure 15 may be inferred from this contour, for the fixed values of C = 840.5 rpm and D = 0.09 mm rev -1 .However, there are grids over the surface with which the visualization of MRR is better.

Optimization by box-behnken method
After the process modeling, the Box-Behnken optimization method is performed using the Design Expert software.The software generates the number of solutions that satisfy the model equations (1) and ( 2).An experienced analyst can select one or a few from the set of optimum solutions, considering the possibility of experimentation.For one of such solutions, a desirability graph is generated and shown in figure 19.At the point of desirability = 1, A = 3.114%, B = 0.732 mm, C = 1037.3rpm and D = 0.101 mm rev -1 .The associated values of surface roughness = 0.522 μm and MRR = 110.2mm 3 s -1 .These are represented in figures 20 and 21, respectively.An experiment has been conducted using these input variables, and the resulting values of surface  roughness and MRR are recorded.Almost an ideal surface roughness value of Ra = 0.1 μm could be observed in the review by Bardwaj et al [28], wherein aluminium MMC with SiC was machined using the diamond tool.A surface roughness value range of Ra = 0.6 to 1.2 could be observed in the research by Kathirvel and Palanikumar [11] while conducting studies with MMC made of A6061 A1+SiC+Graphite.Hence, the surface roughness value achieved in the present work is appreciable.The optimum MRR value reported in this work can be verified with the calculated theoretical value as 119.21 mm 3 s -1 .

Summary of results
By using the Box Behnken optimization method, the machining output process parameters [30] were optimized individually for Ra and MRR.The maximum Material removal rate and minimum surface roughness were found as 110.2 mm 3 s -1 and 0.522 μm respectively to the corresponding machining parameters of the depth of cut 0.732 mm, Spindle speed 1037.3 rpm and feed rate 0.101 mm rev -1 respectively.Table 5 summarizes the results for RSM Predicted and experimental values of individual machining characteristics.The relative error percentage based on predicted values is also shown in table 5.

Conclusion
The following conclusions are drawn based on the experimental and statistical studies on the Hybrid metal matrix composites made of LM24 aluminium matrix with varying weight % of nano alumina and graphite.i. Development of the proposed novel combination MMC has been successful with a uniform distribution of constituents as evidenced from FESEM micrographs.
ii.The material integrity of the MMC is evident with hardness enhancements 16.8%, 23.4% and 30.4% respectively, due to the reinforcements of 1, 3 and 5 wt% nano-alumina and 1 wt% Gr in the LM24 matrix.Moreover, the increase in hardness due to machining is 5.2%, 7.36% and 5.4% in the MMC with 1, 3 and 5 wt% n-alumina, respectively.
iii.Experimental machinability studies have been performed based on the design of experiments and surface roughness and MRR have been evaluated.By response surface methodology, a quadratic equation was evolved with valid statistical indices for surface roughness.The input variable, weight % of nano alumina, is the most significant factor.For MRR, a 2FI equation was evolved.Spindle speed, feed and depth of cut have been found to have high significance.
iv. Optimum process parameters have been identified for the minimum surface roughness and maximum material removal rate of the novel combination of the hybrid MMC with n-alumina and Gr in the LM24 matrix.Box Behnken optimization provided the optimum input parameters as 3.114 wt% nano alumina, depth of cut 0.732 mm, Spindle speed 1037 rpm and feed 0.101 mmrev -1 .The achieved surface roughness is Ra = 0.522 μm and MRR = 110.2mm 3 s -1 .Confirmation tests achieved the values within a relative error of 8.62% and 3.72% respectively.

Figure 2 .
Figure 2. Shows the methodology of the research work.

Figure 3 .
Figure 3. Shows the experimental setup and measurements.

4 .
MMC has been calculated by the equation [27], MRR = ( ) p ´´´´.diameter feed depth of cut Spindle speed 60 Results and discussions 4.1.Surface morphology characterization FE-SEM analysis was conducted to ensure the uniform distribution of reinforcement particles in the MMC.

Figure 4
illustrates micrographs (FE-SEM) of (a) Sample 1: LM24 + 1 wt% nano Al 2 O 3 + 1 wt% Gr (b) Sample 2: LM24 + 3 wt% nano Al 2 O 3 + 1 wt% Gr (c) Sample 3: LM24 + 5 wt% nano Al 2 O 3 + 1 wt% Gr.It is observed from figure 3 that the distribution of nano-Al 2 O 3 and Gr particles are observed with the white and black colours, respectively.Smooth and uniform dispersion of reinforcement particles of nano Al 2 O 3 and graphite (Gr) in MMC composites was achieved by the stir casting technique.The main structure of α-Al dendrite structures was observed in a microscopy investigation.The grain and grain boundary were observed visibly in micrographs.Reinforcement of nano Al 2 O 3 in the cast specimen arrests the movement of dislocation and enhances the properties of the composite, The reinforcement particles of nano Al 2 O 3 and graphite (Gr) in MMC composite improve the strength and self-lubrication properties of the composite.Superior interfacial bonding between matrix and reinforcement was found in figure 4(c) Sample 3: LM24 + 5 wt% nano Al 2 O 3 + 1 wt% Gr.

2 .
The software has suggested the quadratic model[22].Moreover, the ANOVA for the quadratic model has resulted in F = 33.89with p < 0.0001.It means that the chance for the occurrence of this much F-value is only 0.01%.Adequate precision is 19.571 > 4, indicating a proper signal.The model is good enough to navigate the design space.Model terms A, C, AC and A 2 are significant with p values <0.0001, 0.0091, 0.0006 and <0.0001, respectively.The quadratic model equation can be presented in coded form or actual factors.When the coded form is used for making the predictions, the highest levels are coded as +1 and the lowest levels as −1.It helps to identify the relative impacts of the factors by noting the coefficients of the factors.Here, the model equation is presented in terms of actual factors.The values are to be specified in the original units.Relative impacts cannot be observed here as the intercept will not fall at the center of the design space.The quadratic model for the surface roughness is given by

Figure 6 .
Figure 6.Hardness values of pure LM24 matrix, MMC before machining and MMC after machining.

Figures 9 -
11 show 2D contours for surface roughness with input variables A versus B, A versus C and A versus D respectively.These contours have been plotted with constant values of C&D, B&D and B&C respectively, shown on them.From figure8(a), it may be observed that, for the surface roughness to be minimal, wt% of nano alumina needs to be less than around 3 for all depths of cut.Similarly, figures 10 and 11 indicate

Figure 7 .
Figure 7. Actual versus predicted values of surface roughness.

Figure 8 .
Figure 8.Effect of wt% of nano alumina on surface roughness.

Figure 9 .
Figure 9. (A) versus B contour plot for surface roughness.

Figure 10 .
Figure 10.(A) versus C contour plot for surface roughness.

Figure 11 .
Figure 11.(A) versus D contour plot for surface roughness.

Figure 14 .
Figure 14.One factor graphs showing the effect of input variables on MRR.

Figures 14 -
16 show 2D contours for MRR with input variables wt% nano alumina versus depth of cut, wt% nano alumina versus spindle speed and wt% nano alumina versus feed, respectively.These contours have been plotted with constant values of C & D, B & D, and B & C, respectively shown on them.

Figure 15 .
Figure 15.A versus B contour plot for MRR.

Figure 16 .
Figure 16.A versus C contour plot for MRR.

Figure 17 .
Figure 17.A versus C contour plot for MRR.

Figure 18 .
Figure 18.A versus B 3D surface contour plot for MRR.

Table 1 .
Chemical composition of LM24 aluminium alloy.

Table 3 .
Input parameters for machinability studies.

Table 4 .
Design of the 27 experiments with the results.

Table 5 .
Summary of predicted and experimental values of individual machining characteristics.