Effect of machining parameter on the surface roughness of AISI 304 in silicon carbide powder mixed EDM

Article history: Received September 3, 2016 Received in revised format: October 22, 2016 Accepted December 15, 2016 Available online December 15 2016 Powder mixed electro discharge machining (PMEDM) is a hybrid machining process where electrically conductive powder is suspended into a dielectric medium, for enhancing the material removal as well as the surface finish. In this investigation, electro discharge machining (EDM) has been performed for the machining of AISI 304 stainless steel by using the tungsten carbide electrode, when silicon carbide (SiC) powder is suspended into kerosene dielectric medium. Peak current, pulse on time, gap voltage, duty cycle and powder concentration are considered as process parameter while the surface roughness (Ra) is the only response. The effect of significant process parameters on the response has been studied. A regression analysis has been performed to describe the correlation of data between the machining parameter, and the response. Microstructural analysis has been done for the PMEDMed surface. The result shows that peak current is the most influential parameter for surface roughness. Surface roughness decreases with the increase of powder concentration. Growing Science Ltd. All rights reserved. 7 © 201


Introduction
Electric discharge machining (EDM) is the most popular among all the nonconventional machining process.Sometimes this type of machining is also called spark erosion machining because the material removed from the workpiece occurs by means of erosion attributed by electric spark (Choudhary et al., 2013).Conventional EDM has some advantages as well as some disadvantages.To overcome the disadvantages, the conventional EDM is replaced by hybrid EDM.Powder mixed electro discharge machining (PMEDM) is a popular hybrid EDM technique to enhance the material removal as well as the surface finish (Rajagopal et al., 2013;Pandey et al., 2010).Bhaumik and Maity (2014) investigated the effect of tungsten carbide electrode on the EDM performance while material removal rate (MRR), tool wear rate (TWR) and surface roughness (Ra) are taken as responses during machining of AISI 304 stainless steel.Mohanty et al. (2014) studied the influence of process parameter on the various EDM performances such as MRR, surface roughness (SR), radial overcut (ROC) and surface crack density (SCD) during machining of Inconel 625 and found that peak current was the most influencing parameter for all the performances.Pradhan (2013) determined the optimal setting for the machining of AISI D2 tool steel using the combination of response surface methodology (RSM) and grey relational analysis (GRA) coupled with principal component analysis (PCA) where they considered MRR, tool wear rate (TWR), ROC as response.Luis et al. (2005) studied the MRR and electrode wear (EW) for the siliconised or reaction-bonded silicon carbide (SiSiC) during electro discharge machining.Habib et al. (2009) developed mathematical model for the EDM performances viz.MRR, electrode wear ratio (EWR), gap size (GS), Ra for metal matrix composite Al/SiC.For improving the process performance the semi conductive powder particles are suspended into the dielectric fluid during electro discharge machining (Kumar et al., 2010;Singh et al., 2011).Kansal et al. (2005Kansal et al. ( , 2007) ) studied the machining performance of AISI D2 and EN31 steel when silicon powder was suspended in the kerosene dielectric and reported a significant improvement in terms of material removal and surface finish.Bhattacharya et al. (2013) studied the surface properties viz.surface finish and microhardness of die steel using PMEDM.They used graphite, silicon and tungsten powder for this investigation and reported that microhardness and surface finish both are affected by the powder concentration.Kumar and Batra (2012) reported the material transfer during EDM for die steels (OHNS die steel, H13 die steel and D2 die steel) when tungsten carbide powder are added in the dielectric fluid.They concluded that the significant amount of material migrated from the dielectric medium to the work material.Singh et al. (2015) have investigated the enhancement of material removal rate when tungsten powder added into the kerosene dielectric during machining of aluminum alloy 6061/10% SiC composite in electro discharge machining.They reported that material removal was enhanced by 48.43% during PMEDM.
From the above literature, it is evident that most of the researchers have reported about the enhancement of material removal and surface characteristics of the EDMed surface.In this present investigation, a face centered central composite design (FCCCD) based response surface methodology (RSM) has been adopted to design the experimental layout.SiC powder is mixed in the kerosene dielectric for the machining of AISI 304 stainless steel.The main objective of this study is to investigate the effect of significant process parameter on the surface roughness of machined AISI 304 stainless steel.Microstructural analysis has also been done for the machined surface.

Materials and methods
This investigation was carried out using ELECTRONICA-ELECTRAPLUS PS 50 ZNC (die-sinking type) with a servo head (constant gap) EDM machine.For this investigation, AISI 304 stainless steel and tungsten carbide electrode having diameter of 10 mm was taken as workpiece and electrode material respectively.SiC powder having a size of ~20 µm mixed into the kerosene oil in the electro discharge machining.The chemical composition of the AISI 304 stainless steel is shown in Table 1.The face centered central composite design (FCCCD) based response surface methodology (RSM) has been adopted for design the experimental layout.RSM is a group of mathematical and statistical techniques which are used for modeling and analysis of problems where the response is controlled by the input variables, and the objective is to develop a relationship amongst them (Montgomery, 2001).33 experiments have been performed for this investigation.The process parameters are i) peak current (Ip), ii) pulse on time (Ton), iii) gap voltage (Vg), iv) duty cycle (ґ), v) powder concentration (PC).The output parameter considered is surface roughness (Ra) µm.The input parameters and their levels are listed in Table 2.

Experimental procedure
The experimental layout and their results are tabulated in Table 3. Experiments are performed according to the designed matrix.The surface roughness of machined surface is measured by Talysurf (Model: Taylor Hobson, Surtronic 3+).Roughness is measured, in the transverse direction on the machined surface.The process is repeated three times and the average of three readings are noted as surface roughness value.

Result and discussion
In Fig. 1 the mean response curve is presented showing the influence of input parameters IP, Ton, Vg, ґ and PC on the performance of Ra.In this case, IP, Ton, and г has a significant effect on Ra, and it is supported by Table 5.From the main effect plot, it is seen that when IP increases from 4A to 6A, Ra increases by 1.83 µm and 1.083 µm increases when IP increases from 6A to 8A.When Ton increases from 50 µs to 100 µs Ra increases upto 0.46 µm and then decreases upto 0.085 µm when Ton increases from 100 µs to 150 µs.Gap voltage increases from 45V to 65V, Ra decreases by 0.25 µm having a maximum Ra of 8.33 µm at 55V.Duty cycle increases from 55% to 65%, Ra drcreases up to 0.167 µm with a maximum Ra of 8.25 µm at 60%.When SiC powder is added up to 10g/l, Ra gradually decreases up to 0.83 µm.Table 4 shows the ANOVA for Ra with the percentage of contribution of each parameter and their interactions.It shows that the process parameter such as peak current, pulse on time, duty cycle, square term of peak current and interaction term of peak current×gap voltage, peak current×powder concentration, pulse on time×duty cycle, gap voltage×duty cycle, gap voltage×powder concentration have significant effects on the Ra.This table shows that peak current is the most efficient factor for the Ra.It has 80.08% contribution followed by powder concentration, pulse on time, gap voltage and duty cycle having a percentage of contribution of 6.11%, 1.27%, 0.36% and 0.12% respectively.The coefficient of determination (R 2 and adjusted (R 2 ) values are found to be 99.41% and 98.43%, respectively.Lack of fit is not significant for Ra.Table 5 shows the ANOVA table for Ra after eliminating all the insignificant terms.The truncated model possesses lower R 2 than the full quadratic model (R 2 = 99.15%).This exhibits the significance of relationship between the response and the machining parameters for Ra.The model acquired from the regression analysis for Ra shows the influence of linear, square and interaction terms as shown the Eq.(1).Ra = 32.2769+3.3209× IP + 0.0510×Ton -0.0740×Vg -1.2091× г +0.1650×PC -0.1416×IP 2 + 0.0093× г 2 -0.0157×IP × Vg -0.0161× IP × PC -0.0008×Ton × г -0.0029×Vg× г-0.0026×Vg×PC (1) Fig. 2 represents the contour and 3D response plot on the IP and Ton at a constant level of Vg, ґ, and PC.It is seen that Ra increases with the increases of IP and Ton.Ra is minimum at lower value of IP (4A) and Ton (50µs).It happens because with the increase of discharge current the discharge energy density and impulsive force increases leading to a deeper and larger crater, hence surface roughness increases.With the increase of Ton, Ra increases.It happens because an increase in pulse on time the plasma channel in between the tool-worlpiece interface expands which decreases the discharge energy density and impulsive force.For this molten metal cannot be removed properly from the channel resulting increase in Ra.It is seen that Ra increases with the increases of IP and decreases with the increase of powder concentration.Ra is minimum at lower value of IP (4A) and higher value of PC (10 g/l).A significant decrease in Ra is observed with the addition of powder concentration in the dielectric.When powder particle are mixed into the dielectric the plasma channel get enlarged and widened.So the discharge energy is distributed among the powder particle over a large area.As a result the large and shallow craters are generated on the workpiece surface (Fong and Chen, 2005).Along with this, the molten metal is not heavily compressed by the plasma channel and the gas babble.This condition reduces the entrapping of gas in the machining cavity.Hence the surface turn out to be less concave, smooth and uniform (Tzeng and Lee, 2001).

Microstructural analysis
The SEM image of the machined surface are taken with 500X magnification.Fig. 4, Fig. 5, Fig. 6 shows the SEM images of AISI 304 at various level of IP.It is seen from the figure that the existance of cracks are limited in the white layer.This layer consists of cracks, micro cracks, pores and globules.This layer is mainly composed of retained austenite, martensite, and some dissolved carbide.From the SEM images, it is seen that the machined surface has complex exterior such as craters, globules, pores because of the rapid heating and quenching.The formation of surface cracks is credited to the differentials of residual/ contraction stress within the white layer.When the contraction stress exceeds the material's ultimate tensile stress, the surface crack develops.It is seen that with the increase of peak current the surface irregularities increases leads increase in surface roughness.

Conclusions
In this study the influence of most significant parameter on the surface roughness has been studied for AISI 304 stainless steel.RSM based face centered central composite design (FCCCD) has been adopted to conduct the experiment.A full quadratic mathematical model of process parameters which have the significant effect on Ra has been developed.The input factor such as peak current, pulse on time, duty cycle and powder concentration have a significant effect on the Ra.Ra should keep as minimum as possible.For the best setting of Ra peak current of 4A, pulse on time of 50µs, gap voltage of 65 V, duty cycle of 65% and 10 g/l SiC powder concentration should be considered which yields the best value of Ra of 6.7µm.The developed mathematical model for Ra can be effectively employed for optimal selection of PMEDM process parameter to achieve good surface finish of AISI 304 stainless steel.In the SEM image surface cracks, globules, pores are observed.With the increases of peak current the surface irregularities increases.

Fig. 2 .Fig. 3 .
Fig. 2. Contour and response surface plot depicting the effect of IP and Ton on Ra clearly visible in the microstructures.

Table 1
Chemical Composition of AISI 304 stainless steel

Table 2
Level values of input factor

Table 3
Experimental layout and result of PMEDM performance

Table 4
ANOVA for Ra (before elimination)

Table 5
ANOVA table for R a (after elimination)