Investigation of machining rate and surface roughness in wire EDM of Al6063/WC/ZrO2 composite using response surface methodology

The aim of the current work is to explore the machinability behavior of Al6063/WC/ZrO2 metal matrix composites (MMC) produced through a stir casting process through wire electrical discharge machining (WEDM) process. In order to examine the effects of process parameters such as voltage (V), pulse on time (Ton) and pulse off time (Toff) on material removal rate (MRR) and surface roughness (SR), the experiments were conducted by adapting Response Surface Methodology (RSM) in conjunction with the central composite design (CCD). A second-order regression model has been developed to predict the response parameters and an analysis of variance (ANOVA) was performed to validate the significance of the models. Using the desirability function approach, the parameters were set for the highest MRR and the minimum SR value. The prediction was within a tolerable average error range. A total of 19 sets of tests were developed to obtain six possible solutions. The most suitable solution among these six has been found by a confirmatory experiment. The results indicate that voltage and its interactions have significantly affected both the responses than Ton and Toff. Overall, it can be said that the study of the Al6063/WC/ZrO2 MMCs using WEDM process parameters demonstrated improved casting and machining qualities.


Introduction
The need for novel materials with outstanding properties in modern and high-tech industries is unavoidable.A diverse spectrum of materials is being used in numerous industrial applications, including nuclear applications.The materials chosen should have excellent ductility, low density, and high strength, as well as good machinability.Composites are new materials that are proven to be a replacement for typical ferrous and nonferrous materials due to their superior physical qualities [1].Composites are made up of two or more dissimilar materials that are mixed to provide superior physical characteristics including high strength, good machinability, low density, enhanced hardness, lower thermal expansion coefficient, lighter weight and so on [2,3].MMCs are broadly applicable in applications where strong mechanical qualities are a necessity.Numerous papers emphasize the improvement of the mechanical behavior of various materials, particularly aluminum alloys, in light of the dramatically increased industrial sector.Aluminium and its alloys are renowned for having low densities but high strengths that are on par with those of steel [4].The use of aluminum composites is unavoidable in all engineering fields, including aviation and vehicle ventures, because of its higher quality-toweight ratio, resistance against corrosion and machinability [5].However, there are some general issues related to poor high-temperature performance and inadequate wear protection.Recently, there has been a need to increase the mechanical properties of aluminium due to rapid development in industrial activities and subsequent applications, which can be accomplished utilizing MMCs.Therefore, scientists and researchers have been attempting to add suitable components to the aluminium matrix in order to reinforce it without compromising its inherent qualities.Aluminium Metal Matrix Composites (AMMC) are novel design materials that have been developed by strengthening hard support material on delicate aluminum grids to address these difficulties [6][7][8].
Wire EDM is a different process of EDM, which is a crucial non-conventional machining process.It was first used in the mid-1960s.Since the inception of WEDM techniques, the process and materials of the metalworking industries have all undergone significant changes [9].Typically, this approach makes it easy to produce components with intricate profiles, geometries and sharp edges that are challenging under conventional processes [10].WEDM is broadly used to machine MMCs because of its effective operation skills for hard materials [11].This process is utilized in numerous industrial sectors since it can create complex structures.However, the output rate and quality are significantly impacted by the machining parameters.When reinforcement is added to MMC, it improves its mechanical properties.Therefore, it is necessary to optimize the process parameters for each MMC to improve the machining process.Lal et al [12] developed an aluminium hybrid composite combining 7.5% Al 2 O 3 and 7.5% SiC with Al7075.The authors optimized the WEDM process parameters.The results showed that the increase in roughness was caused by pulse time and current from longer, higher-intensity sparks that melted more material per spark on the surface.In another study, Prakash et al [13] produced an Al862 composite by combining different concentrations (1.5%, 3.0% and 4.5%) of B 4 C particles.According to their report, Toff is the factor that affects MRR, followed by Ton.Using an artificial neural network (ANN), Surya et al [14] examined the machinability of an Al7075.In, aluminium matrix composite, TiB 2 was employed as reinforcement.For process optimization, they considered bed speed, current, pulse on and off time.According to their report, the ANN-predicted machining parameter and characteristic were in good agreement with the experimental results.The machining of Al6351/5%SiC with different percentage of B 4 C was studied by Kumar et al [15].Through GRA, the authors optimized the process variables such as current, Ton, wire feed rate, and the weight percentage of B 4 C particle against kerf width and SR.They claimed that peak current (12 A), Ton (100 μs), wire feed rate (6 m min −1 ) and 5 wt% B 4 C were found to be the optimal combinations for good machining processes.Additionally, they imply that Ton had a more profound impact on the output responses.The Ton in this machining process contributed to 96.19% of the total.In order to reduce SR on Al/SiCp MMCs, Guleryuz et al [16] used powdered metallurgy during the machining process.For evaluation, the authors considered various process parameters such as current, voltage, electrode type, Ton, weight ratio of particle reinforcement and voltage.The results of the tests revealed that the most influential characteristics were Ton (34%) and current (31.2%).In addition, the reinforcement contributes to SR up to 6.7%.Using the Taguchi optimization technique, Chen et al [17] improved EDM process variables for milling A6061-T6 aluminum alloy.The study focused current, duty cycle, machining time and Ton to analyze.For the analysis, analysis of means (ANOM) and ANOVA were used.The current and duty cycle in this process were found to be the most influential factors on SR.Under optimized operating conditions, CuZn40 brass alloy specimens had lower SR than A6061-T6.In a WEDM investigation on die steel, [18] Ton and pulse-peak current have an inverse relationship with the SR.On the other side, they noticed that the wire tension and Toff showed lower impact.
According to the literature, only a small number of studies have been done to examine the effect of reinforcement particles on the machinability of the WEDM process.In current research, Al6063/WC/ZrO 2 MMCs have been considered a novel material for the WEDM process since no research has been reported with the above combinations.For the analysis, input voltage, Ton and Toff were considered input parameters and MRR and SR were considered as the output parameters.To achieve a better output response, RSM was offered for determining the most suitable input parameters.To find the impact of each process variables and its contribution to the performance of the output, an ANOVA was also carried out.

Materials and tools
2.1.Al6063 alloy Aluminium alloys are metallic materials.They are used widely due to their favourable mechanical and thermal properties [19].Aluminum alloys can also be easily moulded and are the most machinable lightweight metal alloys when compared to magnesium and titanium alloys.Density, tensile strength, yield strength, weldability, ductility, formability, workability and corrosion resistance are the common factors that should be considered while choosing an alloy for a certain application.Other than Al in the alloy, Si, Mn, and Mg are the other common elements of the Al6063 alloy, which is extensively employed in structural applications [20].Al6063 is commonly used in ship construction as a result of its increased strength and greater capacity to withstand sea environment.It is also employed in nuclear power plants because of its higher strength and improved mechanical properties.Table 1 represents chemical compositions of Al6063 alloy.

Tungsten carbide (WC)
WC is also referred to as tungsten (IV) carbide and tungsten tetra carbide.It is made up of tungsten and carbon atoms in equal amounts.It is a fine gray powder in nature and may be crushed and shaped using the sintering process.WC is broadly used in industrial machineries as cutting tools, abrasives, armor-piercing rounds and high-hardness materials.They are used for the production of friction pads and liner tubes for furnaces [21].The MMCs are prepared for this research using WC of 5 microns in size.The weight percentage of WC in MMCs is 2 wt%.The material was purchased from Zee Precision Carbographite Industries, Coimbatore, India.

Zirconium oxide
ZrO 2 is commonly known as zirconia.It is a white, crystal-like solid.Mineral baddeleyite is the natural form, which has a monoclinic crystalline structure.Zirconia is frequently more useful when it is phase-stabilized.It exhibits disruptive phase transitions when heated.This phase shift can then compress the crack caused by the application of load, slowing its propagation and increasing fracture toughness.ZrO 2 has a density, melting point and boiling point of 5.68 g cm− 3 , 2715 °C and 4300 °C respectively.The percentage of ZrO 2 added to the composite is 5 wt%.The properties of WC and ZrO 2 are displayed in table 2. The ZrO 2 used for this study was purchased from Coimbatore metals, Coimbatore, India.

Preparation of MMCs
In this study, the MMCs were prepared using stir casting process.It is a low-cost method of producing aluminium matrix composites [22].Micron-sized WC and ZrO 2 particles were employed as reinforcement to form Al6063/WC/ZrO 2 composites.In stir casting process, the raw material was placed in a graphite crucible and which was then kept in a furnace.The aluminium particles were heated up to 1000 °C, after that, the preheated 2 wt% WC and 5 wt% ZrO 2 were mixed with base metal and the mixer was stirred continuously for up to 2 h to get a uniform distribution.The materials are exposed to higher temperatures throughout the fabrication process for pre-heating in order to achieve good bonding between the materials.The crucible is removed from the furnace once the reinforcements have been combined with aluminium.The molten material is then poured into the die and permitted to harden.The cast specimen is taken out of the mold when it has solidified and machined according to ASTM requirements for testing.Table 3 indicates the parameter settings used for the casting process.Figure 1 shows a photographic view of the fabricated MMCs.The dimensions of the metal are 100 mm × 100 mm with 10 mm thickness.

Material characterization techniques
The mechanical properties of materials are determined through laboratory tests that simulate service circumstances as closely as possible.The source of loads applied to a material in real life is complicated by a variety of factors.These loads can be applied in a variety of ways, including tensile, hardness, compressive and shear loads.These qualities are crucial in design and engineering material selection.The material was examined for its mechanical characteristics using a universal testing machine (SICMUTM-01, Shambhavi Impex, Mumbai, India), brinell hardness tester (ASI-3000O, A.S.I.Sales Ltd, New Delhi, India) and an impact testing machine (SICMPIT-01, Shambhavi Impex, Mumbai, India).The morphologies were found in SEM images (VP 03-04, Zeiss FESEM SIGMA, Germany).

WEDM process
The Concord wire EDM (Bengaluru, India) machine was used for the WEDM machining process.WEDM is working based on thermo-electric spark erosion principle and produces a series of sparks among the electrode and metallic working plate made up of Al6063/WC/ZrO 2 composites with a thickness of 10 mm.Here, the work plate acts as an anode and the zinc-coated brass wire acts as a cathode.The dimension of the brass wire is 25 mm.Although a number of factors influences the WEDM process, the current work concentrates on factors such as voltage, current, Ton and Toff.Ton is considered one of the important parameters for optimization because material removal takes place at this stage due to anodic breakdown, which takes place at the Ton stage.Toff and voltage are also considered important parameters for the WEDM process.Optimum Toff is also necessary to completely remove all dissolved products.It also prevents the workpiece from becoming too hot.
The factors Ton and Toff, have a significant impact on MRR and SR.Voltage (110-130 V), Ton (110-130 μs) and Toff (50-60 μs), are recognized as ranges of input parameters for optimization.For the experimental studies, the current was kept constant at 6 A. Small amounts of material from the component are melted and vaporized by the sparks.The dielectric fluid pushes the eroded materials away.Figure 2 shows the work tool profile and WEDM machine.Figure 3 shows working material during and after WEDM process.In order to design an L19 orthogonal array to machine the Al6063/WC/ZrO 2 composite, table 4 lists the input parameters and their levels.

Development of RSM
Experimental design is a powerful statistical approach that is employed when an experiment intends to characterize the variation of information under variables that directly affect the output response.The persistence  of this method is to compare the differences in output responses for distinct groups of input modifications.A group of statistical and mathematical techniques known as RSM is a helpful method for modeling and analyzing complex engineering problems.For metal cutting processes, the RSM is recommended to optimize the input variables [23].The L19 orthogonal array was preferred using Design Expert based on the number of factors and levels.MRR and SR were chosen as the response variables.All of these data are utilized to analyze and evaluate the best parameter combination.This experimental procedure is done in the following order: (i) Defining the levels of input parameters and response variables (ii) Experiment design strategy; (iii) Use an orthogonal array;   and (iv) Assess the contribution of components.MRR and SR are calculated at the end of each experimental work.SR is measured with a roughness tester and MRR is calculated using the standard method given below: In order to determine the relationship between machining parameters and process parameters, the RSM has been used in the WEDM process.Y = f (V, Ton, Toff) is the quantitative form of the relationship between responses and input parameters.In the above equation, Y represents the intended reaction, and f is the response function.The fitted second-order polynomial regression model, often known as the quadratic model, was used to propose an estimate of Y for analytical purposes.The common quadratic model is represented in the following equation: where x i (1, 2, k) are the independent of k quantitative process variables.The constant β 0 is accompanied by the coefficients of linear, quadratic and cross-product terms , i b ii b and .ij b  is the error function.According to the selected orthogonal array, the experiments were carried out for each combination of factors.There is only one replication for each combination of parameters.The experimental findings are displayed in table 5.

Mechanical properties analysis
The presence of WC and ZrO 2 seems to have raised the hardness of the material from 25 HB to 51.9 BHN when compared with untreated Al6063.The elongation percentage seems to have grown from 18% to 33%.The addition of WC and ZrO 2 increased the hardness and smoothness.The ultimate tensile strength also increased and recorded as 245 MPa for Al6063 and 271 MPa for Al6063/WC/ZrO 2 composites.The presence of WC and ZrO 2 in the matrix improves the material's characteristics due to changes in particle size, orientation of the molecules, strain fields between matrix phases and interactions.The presence of WC and ZrO 2 in the matrix increased the boiling point, melting point and thermal conductivity to 3015 °C, 94 °C and 152 W m −1 K −1 respectively.SEM analysis had been carried out to inspect the morphology of the base material and the image is shown in figure 4. A closer review of the SEM image reveals the presence of reinforcements and their homogeneous distribution throughout the matrix.It was achieved by a uniform stirring process during material preparation.The lack of voids surrounding WC/ZrO 2 in the matrix is evidence of positive interfacial bonding.The white dots in the image specify the existence of WC.During casting, the porosity reached its maximum which was identified from SEM analysis.During the stirring process, the air bubbles were found to enter the mixture independently or as an air envelope surrounding the reinforcement particles.Table 6 shows the properties of Al6063/WC/ZrO 2 composites and untreated Al6063.
According to Ekici et al [24], the thermal conductivity and density determine the MRR in a WEDM process, even though it is dependent on the material's thermal diffusion.The density of the material reduces as the amount of particles increases, but the melting temperature rises as well, making the material more effective on the MRR.Equation (4) indicates that the most important factor influencing the MRR is V, followed by Toff and Ton.In the above equation, three factor interactions have been identified between V, Ton and Toff.Within the bounds of controlling factors, the MRR can be predicted by the quadratic functions of V, Ton and Toff, which have a substantial impact.The most significant factor in MRR is indicated when the F value is less than 0.05.Table 7 displays the projected coefficients of regression and ANOVA for MRR using 95% of the confidence interval.The significant data, R 2 , known as the determination coefficient in the final ANOVA table, is an indicator of the degree of fit and is described as the ratio of the explained variation to the overall variation [25].The response model fits the real data more closely and exhibits less variation between the expected and actual values when R 2 gets closer to unity.Results show that the adjusted R 2 of 0.9615 and the predicted R 2 of 0.8711 are in reasonable agreement.The results also reveal that the quadratic model exhibits the highest predicted R 2 value and adjusted R 2 value.Conversely, other models display unacceptable negative R 2 values.The sequential P-value, which is less than 0.0001 solely for the quadratic model, indicates a confidence level exceeding 95%, signifying the model's statistical significance.Therefore, the software recommends the quadratic model, while the quartic model is observed to be aliased.Figure 8 illustrates the impact of process factors on MRR.According to the equation, the most important factor influencing the MRR is voltage, followed by Toff and Ton.When the Ton was raised from 50 μs to 53.5 μs, the MRR increased considerably from 16.991 mm 3 /min to 18.706 mm 3 /min.When Toff was raised from 110 to 116.3 μs, there was a small increment in MRR (17.013 to 18.706).In the meantime, it was found that the MRR increased from 16.763 mm 3 /min to 18.706 mm 3 /min as the voltage increased from 110 V to 129.9 V.
In this study, voltage is the most significant parameter determining MRR since it shows a higher percentage contribution than the other two components (Ton and Toff).From surface plots (figures 5, 6 and 7) it was also identified that the Toff and Ton also contributed to MRR.The maximum MRR was obtained when Ton and Toff were set to 53.53 μs and 116.33 μs.Further increases in Ton and Toff reduced the MRR of the composite.It is also identified from the perturbation plot (figure 8).Similar behavior was also observed in the previous studies.According to Nilesh et al [26], in Al/SiC composites, the MRR increased as the Ton increased.An increase in Toff above the optimum point was found to reduce the MRR of aluminum composites [27].The work plate and electrode are melted by higher thermal conduction at a high Ton; nevertheless, the higher plasma pressures may limit the amount of vaporization of the melted electrode.The existence of hard WC particles in the work plate causes built-up edges to form at higher expulsion energies, which may also lead to lower MRR [28].A tiny amount of the matrix metal dissolves during the machining process with elevated energy and is expelled and washed away by dielectric fluid.The left-over material forms a recast layer.According to Goswami and Kumar, these higher frequency pulses initially evaporate a significant amount of metal [29].When the work plate gets heated above the melting point, a significant portion redeposits as recast layers on the outer layer.

Surface roughness
SR is a crucial parameter that reveals the quality of the machining component.In order to get a good surface finish, the work piece should be machined at a slower cutting speed.The ANOVA table (table 7) indicates that SR is significantly influenced by voltage and Toff.In some cases, the Ton also impacts the SR of the machined surface since the deeper and longer discharge crater on the workpiece can be attributed to the larger discharge energy produced by the increased Ton, which spoils the surface integrity [24].The table also interacts with the selected three parameters.Backward elimination is used to remove non-significant variables from the model while it is developing.The model is considered significant at a 95% confidence level.The model is more significant, as evidenced by its F value of 55.61 and related P-value of less than 0.05.The F value and p values for lack of fit are 98.41 and 0.003 respectively indicating the model fit with the data well enough.The regression equation provided below describes the relationship between SR and input parameters.From the ANOVA, it is evident that the individual parameters V and Toff and the interaction of V * Ton and V * Toff are significant.The  factor A, which is voltage, is the most contributing factor influencing SR.The results also elucidate that the quadratic model (equations ( 5) & ( 6)) demonstrates the most substantial predicted R 2 value and adjusted R 2 value.Conversely, the remaining models present inadmissible negative R 2 values.The sequential P-value, registered at less than 0.0001 exclusively for the quadratic model, signifies a confidence level surpassing 95%, affirming the statistical significance of the model.ANOVA shows that the voltage and Toff are determined to be significant at a 95% confidence level among the three selected parameters.As can be seen from table 8, voltage has a significant impact, contributing 55.0%.According to 3D surface plots (figures 11-13), increasing voltage decreases the roughness.The SR was reduced as  the voltage increased from 110 V to 129.9 V.In the meantime, when Ton and Toff were raised from 50 μs to 53.5 μs and 110 μs to 116.3 μs, the SR decreased considerably.It is also known from the perturbation plot (figure 14).Based on the aforementioned data, it can be noticed that an increase in Ton above a certain level leads to higher SR.This is due to the release of more discharge energy due to higher-intensity sparks [30].Due to this higher energy, a small part of the sample melts and evaporates.Further, a part of the molten substance is removed by the dielectric, and the heated metal vaporizes, creating a crater on the surface [31].Furthermore, the crater becomes wider in shape due to high discharge energy.It is also obvious that the higher SR of the machined plate is due to the presence of deep craters [32].The relationship indicates that the lower SR on the machined plate can be achieved with the following parameter settings: voltage = 129.9V, Ton = 53.5 μs and Toff = 116.3μs. Figure 15 displays the predicted versus actual value of SR.It demonstrates the typical distribution of the data and the majority of the residuals are grouped near the straight line, suggesting that the errors follow a normal distribution.Additionally, it shows that the model fits the experiential data.It was discovered that errors fell into a straight line and were regularly distributed.SR and input parameters are related in a contour plot (figure 16).In figure, the green and blue areas represent the average and lower SR.

SR
For every response, specific objectives are set to effectively assess the influence of control elements on individual desirability, such as restrictions for controlling variables and goals to maximize, reduce, or maintain  the target value.Weights are applied to goal values or upper and lower boundaries to emphasize them more.An objective is given a weight of 'unit' by default in order to modify the form of its unique desirability function.
Since the vital purpose of the optimization study is to identify a set of environments that will satisfy all the goals, equal weights of various objective functions assign equal attention to all the objective functions.The optimized bar histograms for the total desirability are shown in figures 17 and 18 shows the ramp graph.

Confirmation experiment
To assess the trustworthiness of the optimization outcomes, a confirmation test was conducted.The confirmation experiment was designed upon solution no. 1, reported in table 9.The confirmation experiments were conducted using the control parameters to get optimal solutions.The table lists MRR and SR as the output responses.Three sets of experiments (as reported in table 10) were conducted with optimal parameters and the average of the output responses was provided at the end of the table.

Validation
To evaluate the prediction abilities of each of the responses, the models were evaluated using the best possible set of test data.In order to ascertain whether the optimized parameters were within the permissible range, the experimental and predicted values were compared and checked.The comparison showed the error limit was within 5%.This demonstrates that the provided models for MRR and SR are valid.

Machined surface analysis
Figures 19(a)-(f) shows the surface morphology of the machined surface obtained at different optimal conditions reported in table 8.All SEM images were obtained at the 2 μm level.Figure 20 shows the SEM image  The shrinkage porosity can be eliminated through the normal heat treatment procedure [33].On the other side, it has an uneven structure.It induces abrupt variations in surface morphology, which will negatively affect SR.The micro cracks that appeared on the surface are the result of partial material evaporation and thermal mismatch during the WEDM process [34].The micro-voids that appeared on the machined surface are due to the removal of an uneven surface, which in turn builds up the defect.Compared to raw casting materials, the voids and cracks on the machined surfaces are very minimal, as clearly indicated by SEM images.The higher porosity identified in the raw material was considerably reduced by the WEDM process.The machined surface has very few cracks compared to the crater and debris.These cracks have a significant impact on SR because of their irregular surfaces.These cracks may propagate during the working of the material since it provides many nucleation sites.Therefore, in order to lengthen the lifespan of the components, the elimination of these surface fractures is necessary.A few recast layers found on the surface are mainly the result of melting and solidifying.This irregularity occurred due to varied pulse on and pulse off  cycles, which caused the material to melt differently.Craters on the surface are caused by the higher erosive force due to the higher discharge energy [35].It can be avoided by having a uniform particle distribution.Compared to the surface morphology of the pure Al6063 reported by Satyanarayana et al [36], the machined surface was covered in small surface microcracks with a maximum length of 15 μm and minimum craters with varying frequencies.For this particular type of material, individual craters have a unique shape that is very different from craters investigated in other materials, including Inconel [37], steel Cr12 [38] and Al/B 4 C [24].

Conclusion
The current study used the CCD of RSM to experimentally examine the effects of three WEDM process variables, namely voltage, Ton and Toff, on two responses, namely MRR and SR, during the machining of Al6063/WC/ZrO 2 MMCs.The following conclusions were made based on the present investigation: • Due to the presence of reinforcement in the Al6063 alloy material, the physical and mechanical properties of the MMCs were improved.
• Effective use of the CCD of RSM enabled the development of a model with a high degree of prediction at 99% confidence.
• In order to confirm the generated model's competency of the generated model, ANOVA was used, and the results of the experimental and predicted vales are in good agreement.• Voltage was found to have a greater impact on MRR and SR than Ton and Toff.
• It was also found that voltage at 129.9 V, Ton at 53.5 μs and Toff at 116.3 μs were assessed as the optimum parameters for maximum MRR and minimum SR for Al6063/WC/ZrO 2 MMC.
• After comparing the experimental and predicted results, it was determined that the error limit was within 5%.
• The micro cracks on the machined surface are attributed to partial evaporation of the material and thermal mismatch during the WEDM process.
• The crater and debris identified on the machined surface creates impact on SR of the machined component, which can be avoided by uniform particle distribution.

Figure 3 .
Figure 3. Working materials during and after WEDM process.

Figure 4 .
Figure 4. SEM images of the workpiece.
Figure 6 illustrates the effect of Ton and Toff while holding constant voltage.A longer Toff lowers the temperature between subsequent sparks, which lowers the MRR of the material.The relationship indicates that the maximum MRR can be attained with the following parameter settings: Voltage = 129.9V, Ton = 53.5 μs and Toff = 116.3μs.The constructed model is also satisfied by the residual plots in figure 9. Errors were found to be regularly distributed and to lie on a straight line.A contour plot (figure 10) relates MRR and input parameters.The maximum MRR is indicated by the red region, while the average and lower MRR are indicated by the green and blue sections.

Figure 17 .
Figure 17.Prediction of machining characteristics by desirability.

Figure 19 .
Figure 19.SEM image of the work piece obtained at different optimal conditions.

Table 2 .
Properties of WC and ZrO 2 .

Table 3 .
Parameter setting for casting process.

Table 4 .
Input machining parameters and their levels.

Table 5 .
Experimental run order.
3.2.Influence of process parameters3.2.1.Material removal rateDesign expert 13 software and pertinent data from table 5 were utilized to calculate the effects of the process parameters on MRR.The relationship was found by second-order polynomial model (equation (4)).The equation correlated the input parameters and MRR in the following way:
Table 11 provides the error values that are computed as

Table 11 .
Validation of the results.