Effect of Al7075 and activated carbon reinforced composite on optimizing WEDM responses

This paper presents wire cut electrical discharge machining (WEDM) response characteristics of Aluminium 7075 (Al7075) reinforced with powdered activated carbon (PAC) composite. In recent days WEDM has become a significant machining process in targeting its benefits of contributing improved material removal rate (MRR) and low surface roughness (SR). This is due the rising need for intricate, accurate, and superior structural components, the WEDM process emerges as a formidable alternative to traditional machine tools. In this work Pulse-on time (Ton), pulse-off time (Toff), discharge current (IA) and servo speed rate (SS) are the variables to be given as input and machining responses such as MRR and SR are studied. From Analysis of Variance (ANOVA) study it is found that discharge current and servo speed is the significant parameters. The optimal desirability condition is obtained with input parameters Ip: 2000 mA; Ton: 8.9 μs; Toff: 25 μs and SS: 150 rpm for the precision machining. The optimum response parameters are found as MRR 10.46 mm3/min and SR 3.32 μm. Results shows that the model designed for the prediction of MRR produces an above 98.27% and the prediction of SR is above 97.17%. The error percentage among the experimental and predicted MRR and SR were estimated. Additionally confirmatory test is performed with optimal results achieved from response surface methodology (RSM) and desirability technique. Metallurgical tests like electron backscatter diffraction analysis (EBSD) and microstructure are conducted to confirm the surface properties and atomic force morphology (AFM) analysis is applied to clarify the structural features of machined composites. The results revealed that the variation of hard deflection is caused by depression of eroded materials on the top layers of machined surface.


Introduction
Aluminium Metal Matrix Composites (AMMC) are the high-potential materials for numerous applications in manufacturing fields, such as aerospace, automobiles, electrical, sports, military and engineering products [1][2][3].In AMMCs, the reinforcement get mixes with aluminium matrix suggestively raises the wear resistance, elastic modulus, fatigue resistance, and strength.Furthermore, while adding reinforcement the thermal expansion coefficient of the matrix material reduces [4][5][6].Kumar et al studied five different process parameters and three response parameters such as material removal rate, surface roughness, and spark Gap are considered for process optimization.Energy-dispersive spectroscopy and scanning electron microscopy analysis reported the manifestation of the recast layer.The complete AMMCs potential is caught up by the trouble of reduced machinability and wide-ranging tool wear from traditional machining methods such as milling, turning, drilling, and to target at anisotropy, non-homogeneity, hardness, toughness, low ductility, and intrinsic brittleness and also due to the existence of hard abrasive reinforcements [7].Studies reported that low material removal, poor surface finish, excessive tool wear, and high costs of manufacture are usually connected with machining of these recent materials [8].Thus, non-traditional techniques of machining, such as wire electric discharge machining, a thermal process, is being engaged to machine easily involving AMMCs [9][10][11].Recently aircraft, nuclear, aerospace and bio-medical industries are utilizes WEDM in various applications.The most commonly used non-conventional process of machining is the WEDM.Vora et al investigated current, pulseoff-duration (Toff), and pulse-on-duration (Toff) are crucial input parameters for the WEDM process of Ti6Al4V.Additionally, the impact of expanded graphite (EG) nano-powder on the output parameters of material removal rate (MRR) and surface roughness (SR) was explored.Incorporating EG nano-powder was observed to enhance WEDM operations, resulting in a notable increase in MRR by 45.35%, while simultaneously decreasing SR by 36.16%[12].Soundararajan et al presented The material removal mechanism in WEDM process comprises of the complex effect of erosion with repetitive, rapid, and discrete spark discharges over the wire tool and job immersed in the medium of liquid dielectrics.These electrical discharges not only melt but also vaporize minute quantity of material that are ejected and flushed out with dielectric, leaving few craters on the work piece [13].Ramraji et al highlighted the importance of particle reinforcement and explored the influence of wire materials on the machining of AA6061-TiB2 composite using Wire Electrical Discharge Machining (WEDM).The study investigated the use of stir cast AA6061-TiB2 with varying weight percentages of 5, 10, and 15% as the work material due to its exceptional mechanical properties and extensive applications [14].The material gap constantly among the work piece and the wire that fluctuates between 0.025 mm and 0.05 mm.Authors published that enhancement of process parameters was performed using various researchers concerning to increase Material Removal Rate (MRR) and lower Surface Roughness (SR).To obtain effective parametric responses a multi-objective optimization is adopted [15,16].RSM technique is utilized for the strategy of experimentation and utilized for optimization of parameters in process of WEDM of HSLA, CFRP, Al-SiC and TiB 2 (P) metal matrix composite (MMC) materials correspondingly.Molybdenum wire was utilized as the wire electrode in with the idea of machining shape-memory alloys and mathematical models have been brought up by RSM, and these models are additionally utilized in predicting as well as in the idea of optimizing the WEDM process parameters [17][18][19][20].
Dhoria et al conducted an experimental investigation was conducted to analyze the impact of input process parameters on the machinability of metal matrix composites (MMCs) using wire electrical discharge machining (WEDM).The study specifically examines hybrid-reinforced silicon carbide and graphite in combination with the base alloy aluminum (Al-6351) [21].Ubale et al found that machinability of W-Cu composite by WEDM and parametric optimization using desirability function approach involved mathematical models in the design of developing the outcomes for performance characteristics [22].Desirability function approach is also utilized in finding the optimal parametric combination for single-objective optimization together with multi-objective optimization.Optimization of the parameters of machining in WEDM based on desirability and statistical analysis concluded that the feed rate highly change on the MRR [23,24].Verma et al employed desirability optimization to investigate various parameters.Their findings revealed that rotational speed and welding speed emerged as the most significant factors.They also observed significant interaction effects between process parameters.Additionally, fractography analysis of tensile specimens was conducted to comprehend the tensile behavior of friction stir welded (FSW) joints.From literature study it is found that none of research conducted in the realm of activated carbon-reinforced aluminum alloy metal matrix composites specifically applied within the WEDM process with multi objective optimization.Research literature and recent investigations indicate that experimental efforts have been made to enhance the surface finish, material removal rate, and mechanical properties of stir-casted Al-MMCs in WEDM.However, these studies often lack comprehensive optimization tools capable of addressing complex issues with multiple and conflicting responses.The absence of adequate analytical models to predict machine responses further complicates matters, especially given the stochastic nature of WEDM and the added complexity from reinforcements in the host matrix material.Therefore, there is a pressing need for experimental investigations into WEDM using aluminum-based MMC samples, along with the development of predictive regression models.Given the numerous machining parameters at play and the inherent uncertainty of these processes, achieving optimal performance poses a significant challenge, necessitating further research in this area.In industrial production, optimization tools for multi-objective optimization are crucial for maximizing output.To address these challenges, this study focuses on multiobjective optimization, wherein weights are assigned to various performance measures to accommodate the requirements of different sectors.The present paper offers an experimental analysis using the WEDM process with response surface methodology for AA7075%-9% PAC MMC.The response parameters are optimized using desirability techniques, and additional analyses such as EBSD and AFM topology studies are conducted to validate the surface properties of the WEDM processed material.

Material
Matrix material AA7075 is reinforced with 18 μm powdered activated carbon was fabricated by down pouring stir casting process.The chemical composition of the as-received AA7075 matrix material is given in table 1. Activated carbon is used as nano powdered particle to improve the mechanical strength of the material.The morphology of reinforcement powder in composite is exhibited by EDS pattern in figure 1 and mapping by the corresponding major elements is shown in figure 2. The samples are prepared with Al7075 is reinforced with weight of 3%, 6%, 9%, 12% and 15% of powdered activated carbon (PAC) using stir casting process as per literatures study related to the application and properties required to enhance with train and error run experiments [3].Based on mechanical and metallurgical evaluation Al7075 with 9% of powdered activated carbon composite of size 100 × 50 × 6 mm was considered for machining process.
AA7075%-9% PAC metal matrix composite with dimension of 100 × 50 × 6 mm was utilized as workpiece in conducting the WEDM experiments.From this a length of 10 mm, width of 10 mm and height of 6 mm were cut by WEDM machine.All the experiments were made to run for a period of 15 min and the machining duration was estimated using a digital stopwatch with 0.1 s accuracy [16].The composite material arrangement in WEDM setup is shown in Figure 3.

Wire electrical discharge machining (WEDM)
The WEDM machine consists of a work table, wire electrode, a power supply, a servo control system, and dielectric system of supply.An electrolytic copper wire having diameter 0.25 mm is used as an electrode.The brass electrode wire converts the electrical energy to the thermal energy for the cutting materials like conductive ceramics, alloy steel and aerospace materials without the consideration of toughness and hardness of the material [24].Deionised water acts as the medium of dielectrics, which possess rapid cooling rate with low viscosity.While machining work piece temperature rises to greater value that is greater than the work material melting point.This erodes material apart from both the work piece and wire by melting and vaporisation and results as a dominant thermal erosion process [6].With less pulse off time value, there is greater discharge quantity in a specified period of time, ending up with the increase in sparking efficiency [7].Discharge current is represented as I A and it is the greater range of the current passing through the electrode during pulse on time.In WEDM, MRR should be as high as possible so as to obtain least machine sequence time for increased productivity.The raise in current value raises the pulse discharge energy that better rate of cutting.For greater value of the current, gap conditions might become unstable with improper grouping of settings in pulse off time and pulse on time.It's measured in mm 3 /min calculated by the formula (1) given below.( In equation (1) I w indicates the work piece weight before the process of machining (g), F w provides the work piece weight after the machining process (g), the density of the material is indicated as ρ (g/mm 3 ) and t indicates the machining time (min).The portable surface roughness profilometer SJ301 with tip radius of 5 μm has been utilized to measure texture on material surface.An electronic balance with the accuracy of 0.001 g was utilized in the measurement of wire electrode weight.For the minimization of error in measurement the average of three weight measurements were utilized.
Adapting RSM with a Box-Behnken design for four variables and three levels total of 27 machining operations have been performed [12].All the factors including their responses are given in table 1.The average number of tests performed for machining process parameter and its measured MRR and SR are also presented in table 2. The selection of specific input parameters for the WEDM study involved a combination of factors including initial tests, review of relevant literature, and insights received from prior research.According to findings reported in the literature, based on trial experimentation the levels of four parameters are selected.Statistical Minitab software was used in performing the analysis of variance of MRR and Sr ANOVA is used because it is a conceptually simple and powerful statistical testing method for experiments repeated measures test.Based on the response values ANOVA analysis was done for regression test and multi objective desirability optimization was performed.

MRR and SR response model for AA7075%-9% PAC composite
The machining parameters effect (I A , T on , T off and SS) on the response variables are MRR and SR for AA7075%-9% PAC composite was analyzed through carrying out the experiments as mentioned in table 3. Identifying the correlation among the input parameters and output responses of MRR and SR, design expert software is used.
The values of thecoefficient of determination (R 2 ) and adjusted R 2 statistic (R 2 adj ) are compared that is summarized in table 4 for MRR and table 5 for SR, to choose the regression model.The full quadratic model for MRR and SR is listed, where R 2 = 98.02% for MRR and R 2 = 99.32% for SR indicates that in responses are described by analytics in the model [10].However, R 2 adj is 95.96% for MRR and 98.18% for SR, that contributes too many of analyses in the model describe the importance of the affiliation [12].Tables 4 and 5 indicates the regression coefficients in coded units for the model.For the testing of adequacy of the model, with the confidence level of 95%, the p-value for the statistically significant term must be lowers than 0.05.Multiregression analysis is carried out to produce a quadratic response surface model for MRR and SR and the thus obtained equation in un-coded unit is given in equations (2) and (3).The effects of these variables and the interaction are represented as: The regression equation for MRR and SR of WEDM process of composite is    The F-value for the Model is 13.57 that are more than the significant P value and has only a 0.01% of probability for the F-value to behigh in occurring noise this indicates the model issignificant.Also B, C and D are significant terms of model whosevalues are less than 0.1000.The Lack of Fit for is 2.82 and has 16.25% of probability that error could occur due to noise.The Predicted R-Square is 98.02% and agrees with the Adj R-Square of 95.96% the signal to noise ratio.The F-Ratio shows a sufficient signal as the Adeq precision ratio isgreater than 4. Current model could be utilized in navigating the design space.The equation in terms of actual factors is utilized in predicting MRR forgiven levels of each factor [10].The equation with respect to actual factors is utilized in making predictions about the response for each factor of given levels.The figure 5(a) shows that the surface plot of MRR which is high as the servo speed decreases and the discharge current increases are 1900 mA and servo speed 60 rpm.
The figure 5(b) shows that the contour plots of MRR which is high as discharge current and speed is low.Because of maximum energy created when discharge produces between the wire and the workpiece [11].Figure 6 shows the main effect plot of MRR which predicts the mid discharge current and high servo speed is the highest off set value.Thus, from the variation plots on MRR based on machining parameters, it is clear that the pulse on time, discharge current, servo speed are the major parameters that influences maximum MRR and pulse off time is the minor parameter influencing maximum MRR.Discharge current is the peak sensitivity performance for this WEDM process due to more discharge energy produces [12].Since the error prediction in the residuals, the strongest interactions among all input parameters are shown in figure 7.

Influence of process parameters in case of surface roughness
Influence of process parameters on surface roughness is discussed below.Table 4 shows that with the increase in discharge current and pulse on time, the SR increases.Foreseen and investigated SR values are graphically exhibited.The high pulse on time results in high velocity and SR of wire EDM.Similarly, SR is less with the servo speed of 150 rpm and the T on of 75 μs.With continues increase in experimental data, the surface roughness decreases with the increase in on time and off time [11].However, the material roughness reduces with the on time of 5 μs and the discharge current of 2000 mA.It's studied that the foremost influencing parameters pulse on, pulse off time to most SR and therefore the minor influencing parameters discharge current and servo speed to contribute high MRR [12].The normal plot for the residuals of SR is shown in figure 8, where the errors are distributed around the straight line.
The F-value for the Model is 21.15 that are more than the significant P value and only a 0.01% of possibility is available for the F-value to be greater for occurring noise.This indicates the model is significant [17].Also A, B  and D are significant terms of model whose values are less than 0.1000.The Lack of Fit is 1.11 and has 16.25% of probability that error could occur due to noise.The predicted R-Square is 99.32% and agrees with the Adj R-Square of 98.18% the SNR.The F-Ratio shows a sufficient signal as Adeq precision ratio is greater than 4. Present model could be utilized in navigation to the design space [17].
From figures 9(a) and (b), it is obvious that, with the increase pulse time, the discharge current increases from 1750 mA to 2000 mA.When, the on time of pulse decreases, the off time of pulse increases to 10 μs. in addition, the roughness of the surface declines with servo speed and off time of pulse.Figure 10 shows the main effect plot of the parameters on pulse on time and pulse off time which are increasing the influence on material roughness.The other two parameters are discharge current and servo speed decrease the influencing on surface roughness.The effective interactions among all parameters of input are noticed from figure 11.
The Pred R-Squared of 0.9932 seems in sensible agreement with Adj R-Squared of 0.9818.Adeq Precision estimates the signal to noise ratio (SNR), and the value of SNR greater than 4 is acceptable.F-ratio of 21.15 illustrates a satisfactory signal while the current model could be utilized in design space navigation [15].

Adequacy test for the developed models
The statistical results obtained using the designed regression models are depicted in table 6.The forecasted values preferably match with the consistent experimental results with the value of R 2 being one.Since, the R 2 value obtained for MRR is 0.98 and SR is 0.99, the developed models are seemed acceptable.The adequacy of the established regression models were tested with the variance analysis and the results are depicted.From table 6 it is found that the tabulated error values are found at 95% confidence level [17].

Validation of the regression models
In the idea of checking the accuracy measure of the developed regression models, five experiment runs were conducted on the AA7075/PAC composite with various values of discharge current, servo speed, pulse on time and pulse off time which are not used in design matrix.From table 6, it is obvious that the model designed for the prediction of MRR produces an above 98.27% and the model designed for the prediction of SR is above 97.17%.The error percentage among the experimental and predicted MRR and SR were estimated and illustrated in table 7.

Multi response optimization using desirability function
Desirability is an independent function which changesbetween zero outside of limits and one at goal [21].An optimization, which is numerical, brings up a point that makes the most of the desirability function.The desirability goal's characteristics may be transformed by altering the importance or weight [22].Table 8 illustrates the level of input parameters and the output response for desirability.
The overall desirability functions of the responses in shown in bar graph of figure 12.The optimal region had an overall desirability of 0.8597 that represents the nearness of the target.Design expert statistical software was used for conduct the desirable and optimum settings which meet necessary goals are achieved.A set consisting of 11 optimal solutions areobtained for the design space for individual response characteristic namely, pulse on time, discharge current, servo speed and pulse off time.Certain criterion providing greatest value of desirability is chosen as the optimum situation for anticipated response.3D plots of desirability are createdwith input parameters within range and responses at minimum.Table 9 shows the data set of optimal solutions for high desirability combinations.The data is used to predict optimum responses to carry out surface morphology and metallurgical analysis.10 presents the concluding optimum level set of numerous parameters involved in process and the foreseen values of various response characteristics [21].Figure 14 indicated by each ramp function for desirability characteristics for parameters.A linear ramp function is produced between among the high value and the goal or the low value and the goal as each parameter was predefined with the weight value of one [25].

Analysis of WEDM machined surfaces
The surface morphology and metallurgical analysis on the predicted optimum responses performed based on table 9. Microstructure image observation is shown in Figures 15(a)-(b).The EBS analysis for the optimum responses is shown in figures 16(a)-(b).Figure 15(a) shows the heavy flow of material in cutting side and it causes flaw between tool and material.It also shows interface between the materials reaches high temperature near to melting point [26].Due to generation of high temperature during WEDM process plasticized tool material layers are extruded and mechanically mixed with composite layers.Figure 15(b) shows there will be no voids or cracks present in the machined surface so it is concluded that flow of material is uniform.It also evident there is smooth flow of material behind the input parameters [27].
Figures 16(a)-(b) reveals that EBSD image of WEDM processed composites which show its controlled reinforcement fractions and color scheme reflects its grains orientation from the surface [27].The images extent the confirmation on no deformation is found using morphology interface area by the green, yellow and red colors represent progressively increasing levels of internal orientation across the composites.
Figures 17(a)-(d) shows Atomic Force Microscopy (AFM) images of RSM predicted machined composite samples and figures 18(a)-(d) shows AFM images of desirability predicted machined samples.AFM images are used to examine the adhesion and mechanical behaviour of composites at very micron scale levels.Figure 17(a) shows the topography lattices of machined samples which indicate the presence of metallic crystals.This reveals surface lattices but local molecular scale defects are measured.From figure 17(b) well-ordered arrays of folds were observed in height mode AFM image of surfaces [28].Figure 17(c)-(d) shows the difference between

Conclusion
This paper investigates the use of WEDM process on AA7075 reinforced with 9% powdered activated carbon composite.It evaluates the machining responses focusing on surface roughness and material removal rate.ANOVA analysis was employed to statistically validate the mathematical models developed using response surface methodology.Optimal parameters were predicted using RSM technique and desirability function methods.Validation tests were performed based on these optimal parameters.The research work leads to the following conclusions: • Discharge current and servo speed rate are the significant parameters for MRR and SR in WEDM process.
• The desirability predicted optimal values and the RSM predicted results are close to each other with low percentage of error.
• The optimal combination of discharge parameters with respect to the corresponding highest desirability, was obtained as Ip: 2000mA; Ton: 8.9 μs; Toff: 25 μs and SS: 150 rpm for the precision machining for improving MRR and minimising of Sr • The optimum response parameters are found as MRR 10.46 mm 3 min −1 and SR 3.32 μm.Results shows that the model designed for the prediction of MRR produces an above 98.27% and the prediction of SR is above 97.17%.
• Validation tests are conducted using the optimal parameters to confirm the reproducibility of experimental conclusions.EBSD study and AFM topography presented that optimum parameters are interface clear surface qualities with no deformations after WEDM machining.

Figure 4 .
Figure 4. Normal probability plot of residuals material removal rate.

Figure 4
Figure 4 depicts the normal residuals plot for MRR.Here the errors are normally distributed through straight line.The F-value for the Model is 13.57 that are more than the significant P value and has only a 0.01% of probability for the F-value to behigh in occurring noise this indicates the model issignificant.Also B, C and D are significant terms of model whosevalues are less than 0.1000.The Lack of Fit for is 2.82 and has 16.25% of probability that error could occur due to noise.The Predicted R-Square is 98.02% and agrees with the Adj R-Square of 95.96% the signal to noise ratio.The F-Ratio shows a sufficient signal as the Adeq precision ratio isgreater than 4. Current model could be utilized in navigating the design space.The equation in terms of actual factors is utilized in predicting MRR forgiven levels of each factor[10].The equation with respect to actual factors is utilized in making predictions about the response for each factor of given levels.The figure5(a)shows that the surface plot of MRR which is high as the servo speed decreases and the discharge current increases are 1900 mA and servo speed 60 rpm.The figure 5(b) shows that the contour plots of MRR which is high as discharge current and speed is low.Because of maximum energy created when discharge produces between the wire and the workpiece[11].Figure6shows the main effect plot of MRR which predicts the mid discharge current and high servo speed is the highest off set value.Thus, from the variation plots on MRR based on machining parameters, it is clear that the

Figure 5 .
Figure 5. (a) Surface plot interaction on process parameters on material removal rate.(b).Contour plot interaction on process parameters material removal rate.

Figure 6 .
Figure 6.Main effect plot for material removal rate.

Figure 7 .
Figure 7. Interactions plot for influence of process parameters on material removal rate.

Figure 8 .
Figure 8. Normal probability plot of residual on surface roughness.

Figure 9 .
Figure 9. (a) Response plot for interaction of process parameters on surface roughness.(b).Contour plot for interaction of process parameters for surface roughness.

Figure 10 .
Figure 10.Main effect plot for surface roughness.

Figure 11 .
Figure 11.Effect of process parameters interactions plot on surface roughness.

Figures 13 (
Figures13(a)-(b) shows surface and contour plot of each response for each factor using design expert solver.Table10presents the concluding optimum level set of numerous parameters involved in process and the foreseen values of various response characteristics[21].Figure14indicated by each ramp function for desirability characteristics for parameters.A linear ramp function is produced between among the high value and the goal or the low value and the goal as each parameter was predefined with the weight value of one[25].

Figure 12 .
Figure 12.Bar graph of desirability for response values.

Figure 13 .
Figure 13.Desirability response of process parameters (a) surface plot (b) contour plot.

Figure 14 .
Figure 14.Ramp function graph of desirability for parameters.
deflections in force versus distance curves measured on the material surface.The variation of hard deflection is caused by depression of eroded materials on the top layers.From figure 18(a) observed that surface topography is composed of molecules and crystals.Figure18(b) also shows visibly ordered fold arrays in height mode image.Figures18(c)-(d) reveals that topography deflection curves induced small deformation on the surfaces[29].
Figure 19(a) represent the 3D images of RSM predicted WEDM machined surface.This indicated that that the severely damaged surface with deep normal scratches

Figure 17 .
Figure 17.(a) AFM topography image of the surface of machined composite (RSM), (b) AFM height image on the surface of machined composite, (c) Topography curve and (d) Deflection-Force curves recorded on the surfaces.

Figure 18 .
Figure 18.(a) AFM topography image of the surface of machined composite (Desirability), (b) AFM height image on the surface of machined composite, (c) Topography and (d) Deflection-Force curves recorded on the surfaces.
Figure 19(b)  shows the 3D images of desirability predicted WEDM machined surface.This indicates the narrow ploughing trenches with no delamination and cracks[31].

Table 2 .
Variables and the levels.

Table 3 .
Process parameters and responses for WEDM.
3.2.Influence of process parameters on material removal rateThe most influencing process parameters that signify the responses are discussed in table 3. The MRR increases with the increase in discharge current and pulse on time.The experimental and predicted values of MRR are compared graphically and presented.When pulse on time is high the velocity is high which results in high MRR.

Table 4 .
ANOVA table for MRR estimated regression coefficients.

Table 5 .
ANOVA table for SR estimated regression coefficients.

Table 6 .
Statistical results in terms of developed regression models.

Table 7 .
Results of conformity tests in terms of developed regression models.

Table 8 .
Range of input parameters and responses for desirability.

Table 9 .
Set of optimal solutions for high desirability combinations.

Table 10 .
Optimal sets of parameters using desirability approach.