Powder mixed electrochemical discharge process for micro machining of C103 niobium alloy

Nildri Mndl ,Nitesh Kumr ,Alok Kumr Ds ,*

a Defence Research & Development Laboratory (DRDL) - DRDO,Hyderabad,India

b Department of Mechanical Engineering,Indian Institute of Technology (ISM),Dhanbad,India

Keywords:Micro-electrochemical discharge machining C103 niobium alloy Surface integrity Material removal rate Hybrid powder mixed ECDM

ABSTRACT This work demonstrates the viability of the powder-mixed micro-electrochemical discharge machining(PMECDM) process to fabricate micro-holes on C103 niobium-based alloy for high temperature applications.Three processes are involved simultaneously i.e.spark erosion,chemical etching,and abrasive grinding for removal of material while the classical electrochemical discharge machining process involves double actions i.e.spark erosion,and chemical etching.The powder-mixed electrolyte process resulted in rapid material removal along with a better surface finish as compared to the classical microelectrochemical discharge machining (MECDM).Further,the results are optimized through a multiobjective optimization approach and study of the surface topography of the hole wall surface obtained at optimized parameters.In the selected range of experimental parameters,PMECDM shows a higher material removal rate (MRR) and lower surface roughness (Ra) (MRR: 2.8 mg/min and Ra of 0.61 μm) as compared to the MECDM process (MRR: 2.01 mg/min and corresponding Ra of 1.11 μm).A detailed analysis of the results is presented in this paper.

1.Introduction

With the rapid developments in the aviation industries,the demand for advanced aerospace materials and their processing is constantly growing.Niobium-based alloys have proven their competency in the space industries for years due to their retention of strength at higher temperature,and corrosion resistance properties[1].C103 is one of the niobium-based alloys that is utilized in hightemperature applications,such as in propulsion systems,thrust augmenter flaps of jet engines,and rocket nozzles[2,3].This alloy is a combination of 89% niobium,10% hafnium and 1% titanium(Nb-10Hf-1Ti wt%).Traditional machining of this alloy is difficult due to the formation of a built-up edge.The difficulty level further multiplies while machining micro-features on this material.

In view of the above,the unconventional machining processes such as the micro-electrical discharge process (μEDM),microelectrochemical (μECM) and hybridisation of both (micro-ECDM)are the suitable methods for shaping this material [4,5].

The ECM suffers from the problems like poor dimensional accuracy of the machined features,passivation layer formation,debris deposition etc.[6];whereas the micro-EDM process suffers from low material removal rate,high tool wear,recast layer deposition,surface micro-cracks,residual stress,surface roughness and etc.[7].The features machined through the above methods need postprocessing operation which is more challenging than the machining operation [8].

Therefore,in-process removal of the deposited debris,reduction of micro-cracks,and control of residual stresses etc.are required.The MECDM process is one of the viable solutions to overcome the challenges.As this process is a combination of micro-EDM and micro-ECM,the limitations of both processes are narrowed down.The surface finish of the fabricated features would be better than the micro-EDM wherein the geometrical accuracy would be better than the micro-ECM [9,10].

To this aspect,Yuehong et al.[11]machined a groove on Ti6Al4V in the presence of 20 wt% NaCl solution and DC power supply with the ECDM process.They found that the surface roughness of the machined grove is approximately 1.49 μm without any recast layer deposition and stray corrosion.It is suggested by other researchers that the performance of the MECDM process can be further enhanced by modification of the process such as mixing of nano powders with the electrolyte [9],and application of ultrasonic vibration to the tool/workpiece/machining tank [12].For instance,Han et al.[9]added micro-sized graphite powder in NaOH solution to improve the surface integrity by reducing the tiny cracks formed over the machined surface and thereby achieving a surface finish up to 1.44 μm.Kuo et al.[10]used micro-sized SiC powder added electrolytes to boost the machining performance.Due to the grinding action by the abrasive particles,the debris particles are removed efficiently from the machined surface,thereby improving the surface finish by 80%.In this hybrid MECDM approach,a very small gap(by using micro-size abrasive particles) is maintained to supply the fresh electrolyte and remove the debris particles from the machined depth.Yang et al.[13]reveal that mixing the powder/abrasive particles with electrolytes solution interrupts the accumulation of hydrogen bubbles and hindered the formation of the gas film layer over the tool electrode.This tends to increase the critical voltage and reduces the discharge energy.Varghese et al.[14]observed multiple sparking actions while performing the machining of polypropylene using graphite powder mixed with Noah (30 wt%) electrolyte solution.They concluded that by using powder mixed electrolyte,the material removal rate increases up to certain concentrations and after that,it starts to deteriorate due to the restriction of free movement of the ions in the solution.

Although previous studies [15]has demonstrated the effectiveness of the powder mixed MECDM to machine non-conductive materials such as glass and ceramics,but,this process has not been explored for the exotic materials including the niobium-based C103 alloy.Based on the research gap,this article has aimed to investigate the performance of powder-mixed MECDM (PMECDM) process for the machining of niobium-based C103 alloy and the performance comparison with the classical MECDM process.

Niobium-based C103 alloy shows its prominent presence in the aviation sector as an application in rocket nozzles [16].Hence,the machining capability of this material through both classical and hybrid ECDM approaches has been highlighted by the drilling of micro holes.Powder mixed MECDM (PMECDM) is a unique approach for the micro-machining of exotic materials with acceptable dimensional accuracy and surface finish.It also eliminates the need for post-processing operations to remove the debris from the wall surface of the drilled micro holes.The micro-sized Al2O3powder is the most frequently used abrasive powder as it is thermodynamically stable,hard in nature,high melting point,and chemical inertness as well as easily available [17].Hence,for the present investigation,micro-sized Al2O3powder is used.

1.1.Principle of powder mixed micro-ECDM process

The powder-mixed MECDM process works on the same principle as classical MECDM.The stages in the MECDM process are illustrated in Fig.1.Initially,voltage is applied across the tool and auxiliary (workpiece in this study) electrode placed in the machining tank which contains powder-mixed electrolyte [15].When the voltage exceeds the threshold limit,the electrochemical reaction takes place,thereby evolving hydrogen gas which is converted to bubbles (Fig.1(a)).The coalescence of the said bubbles leads to the formation of a passivation gas film [18](Fig.1(b)) and thus the electrochemical reaction slows down [19].With further increase in voltage,the sudden breakdown of the gas layer takes place and sparks (discharge) are established (Fig.1(c)) where the gas layer thickness is lowest(current always flows through the least resistive path)[20].Material removal takes place due to two actions simultaneously such as,the heat energy developed through discharge and heat-assisted chemical etching [21].

Fig.1.A schematic diagram of micro electrochemical discharge machining of conductive material.

The mixing of abrasive powders in small quantities to the electrolyte has added advantages that promote grinding action(Fig.2(b)) to remove the debris,thus incorporating compressive residual stress in the machined surface and also contributing to enhancing the material removal mechanism and surface finish[9,10].Suitable abrasive powder must be selected for better performance of the process.Fig.2(a)represents the mechanism of material removal through the classical MECDM process whereas Fig.2(b)depicts it through the powder suspended MECDM process.In Fig.2(b),the suspended powders (in the present study Al2O3)contribute grinding action on the machined surface [9,22].

Fig.2.Mechanisms of material removal in (a) classical (b) powder mixed MECDM processes.

2.Experimentation

2.1.Experimental methodology

The work was accomplished in three phases.The first phase deals with the fabrication of the micro-holes over C103 alloy sheet through the classical and powder mixed MECDM processes with different input parameter settings.In the second phase,the responses i.e.material removal rate and surface roughness were measured and analysed.In the third phase,optimization of the input parameters was carried out for maximization of material removal rate and minimization of surface roughness.In addition,confirmation experiments were conducted to validate the results.A flowchart of the present study is shown in Fig.3.

Fig.3.Flowchart of the experimental methodology.

2.2.Setup configuration

A customized micro-ECDM system(Fig.4(a))is used to conduct all the experiments.It comprises of three motorized stages(i.e.,X,Y,andZwith a resolution of 1 μm and a travelling range of 300 mm in each axis).Siemens CNC 828D controller has been used to control all the operations.A specially designed spindle having a maximum rotational speed of 2000 rpm is mounted onZ-axis.The spindle is facilitated with collet to hold micro tool of diameter ranging from 0.5 to 1.5 mm.A machining tank of dimension 300 × 300 × 100 mm3was fabricated from Perspex sheet of thickness 5 mm.This tank has a workpiece clamping facility capable to hold workpieces of sizes ranging from 10 × 10 × 1 to 200 × 200 × 100 mm3.A pulse DC power unit is used as a power source capable to vary three input machining parameters i.e.applied voltage(10-150 V),duty factor(from 5% to 70%)and pulse frequency (from 100 to 10 kHz).A digital microscope (Dino-lite microscope) is also integrated with the set-up for online monitoring of the machining process.Figs.4(b) and 4(c) depict the schematic diagram of the classical and powder mixed MECDM process.

Fig.4.(a) Experimental setup and schematic diagram for (b) classical and (c) powder-mixed MECDM process.

A small submersible pump is used for the circulation of Al2O3powder into the NaOH electrolyte.For each experiment,a freshly prepared electrolyte solution is used to eliminate the experimental error.

2.3.Materials specification

In the experiment,Niobium-based alloy (89Nb-10Hf-1Ti wt%)is used as the workpiece with dimension 50 × 50 × 0.5 mm3.The commercially available tungsten rod (Ø1 mm,length: 25 mm) of purity 99% is used as a cathode (tool electrode).The physical properties of the workpiece and tool are presented in Table 1 and Table 2.The Energy-Dispersive X-ray Spectroscopy (EDS) is also carried out (Fig.5) to check the chemical composition of the workpiece (to be added in EDS),tool &powder materials.The average particle size of the suspended alumina powder (Al2O3) in the electrolyte is 5 μm.

Table 1The physical properties of the workpiece.

Table 2The physical properties of the tool electrode [23,24].

Fig.5.EDS plots of the (a) workpiece,(b) tool material.

2.4.Preparation of powder suspended electrolyte

The sodium hydroxide pallets(NaOH)of the required weight are mixed with the deionized water and stirred for 10 min for proper dissolution of the pallets in the water.Then,alumina powder of desired weight is mixed with the prepared electrolyte solution and stirred properly with help of a magnetic stirrer for 30 min.Thereafter,ultra-sonication of powder mixed electrolyte solution is done for 10 min to agitate the coagulated powders in the solution.Finally,the solution is further mixed by the magnetic stirrer for 10 min.The schematic diagram for the preparation of powder-mixed electrolyte solution is depicted in Fig.6.

Fig.6.Schematic diagram of the preparation of powder mixed electrolyte solution.

2.5.Experimental procedure

Micro-hole drilling on the workpiece was performed by varying the different input machining parameters (selected based on the literature survey) i.e.applied voltage,duty factor,and electrolyte concentration.The levels of the input parameters are selected by performing pilot experiments.The cylindrical tool was mounted on the spindle and positioned over the workpiece (mounted on the fixture) in such a way that the top surface of the workpiece is perpendicular to the tool electrode.Initially,the tool was made to touch the work surface and short-circuit was checked using a multi-meter subsequently by moving the Z-stage,then the tool is retracted by 30 μm to provide inter-electrode gap.The workpiece was made positive while the tool electrode was connected to the negative terminal of the pulse DC power supply.

The prepared powder mixed NaOH electrolyte was poured into the machining tank up to 2 mm [25]above the surface to be machined.To avoid settling down and coagulation of the alumina powder particles,a submersible pump (chemical pump) is integrated with the electrolyte tank.This helps to circulate the electrolyte solution,resist the coalescence of the powder and uniformly distribute the powder particles over/near the workpiece surface[26].

As per the results from pilot experimentation,the suitable range of machining variables is selected and presented in Table 3.The experimental runs were planned as per the response surface methodology (RSM) based Box-Behnken design [26].Total 15 sets of experiments were designed for each classical and hybrid ECDM process (as shown in Table 4).Further,each experiment was repeated to assure the repeatability of the results.In addition,experiments were performed in random order to eliminate the error caused by unknown factors.

Table 3Experimental condition for classical and hybrid MECDM approach.

Table 4Experimental run and their output responses.

Alumina (Al2O3) powder of concentration 3 wt% (based on the trial experiments) was mixed with alkaline electrolyte(NaOH)solution to prepare the electrolyte.A controlled velocity tool feed(decided post-trial experiments) was chosen for both the machining process.After each experiment,an acetone-filled ultrasonic bath was used for 15 min to properly clean the workpiece,removing the grime (debris and Al2O3particles) that had partially accumulated on the surface [27].The moisture was then removed from the workpiece by allowing it to dry for 10 min in an oven set at temperature of 80°C.Subsequently,MRR for both processes was calculated by the weight loss method of the workpiece,using equation (i)

whereWiis the initial weight;Wfis the final weight of the workpiece andtis the machining time in minutes.The same workpiece was used for the next experiments.

2.6.Characterization methods

The surface topography of the respective holes was analysed by using the Field Emission Scanning Electron Microscope (FESEM)and Energy-dispersive X-ray Spectroscopy(EDS).The drilled micro holes were cleaned and sectioned using a Wire-EDM along their diameters for surface roughness analysis on the periphery walls using a non-contact type surface profiler.

3.Response surface methodology

Response surface methodology is a combination of mathematical and statistical approaches that are frequently used in experiment design,regression model development,and response attribute optimization with minimum experiments.It utilises quantitative data from the linked experiment to assess the regression model and optimise an output response that is impacted by numerous independent input variables [22,26].In the present case,the regression model is developed for MRR andRausing MINITAB 17 statistical tool software.Eqs.(2) to (5) is presented in uncoded form for MRR(material removal rate for MECDM),PMMR(MRR obtained for powder mixed ECDM),Ra(average surface roughness for MECDM) and PRa(Raobtained for powder mixed ECDM).

The analysis of variance (ANOVA) has been carried out at 95% confidence level in order to comprehend the effect of each input machining variable on the experimental investigation.Tables 4 and 5 illustrate the ANOVA outcomes of MRR and Ra for the classical and hybrid machining approaches.When,p＜0.05,the parameters or their combinations are statistically significant otherwise it is insignificant.The developed regression model is significant if the Pvalue for the lack-of-fit is greater than 0.05.Consequently,a higher"F" value reflects a strong impact of the input variables on the efficiency of the machining.Furthermore,the R2and adjusted R2is above 90% which indicates the good fitness of the model to predict the responses [22,26].

Using the statistical software Minitab 17,the normal probability graphs for MRR and Ra were also drawn for both the machining process (as shown in Fig.7).The plot serves as a means of correlating the experimental finding with points from the standard normal distribution.With this technique,data is plotted against a hypothetical normal distribution so that the points form a straight line or depart from normality [28].From Fig.7,it is noted that the observed data and the projected response values' variation from the baseline approximately follow the same trend.Responses are thus fitted to the baseline in both situations and demonstrate a higher likelihood of accepting the developed models for the MRR and surface roughness.

Fig.7.Normal probability plots for the actual and predicted values of ((a),(b)) MRR and ((c),(d)) surface roughness for ((a),(c)) classical and ((b),(d)) hybrid process.

4.Results and discussion

The experiments were conducted and measured results are shown in Table 4.The samples were cleaned and taken for different characterization processes.The analysis which is carried out during and after the experiments are described in the succeeding sections.

4.1.Analysis of voltage-current waveform

An oscilloscope(Tektronix TDS2012C)was connected across the tool and the workpiece to capture the voltage and current signals.Figs.8(a)and 8(b)depicts the waveform to represent the variation of voltage and current signal with respect to the time in PMECDM and ECDM (at machining voltage 40 V,duty factor 30% and electrolyte concentration 15wt%) process.

In the classical ECDM process (Fig.8(b)),initially the electrochemical reaction occurs in the machining cell,which generates hydrogen bubbles near the tool electrode.These bubbles collapse and forms a gas film(insulating layer)and thus create resistance to the flow of current in the circuit ((a) and (b)) [29-31].Due to this resistance,ohmic heating of the electrolyte occurs near the cathode and thus increases the conductivity of the electrolyte and the flow of current in the circuit ((b) and (c)) [30].At this situation,if the applied voltage is increased gradually and exceeds the critical voltage,the insulating layer breaks down and electric sparks are produced((c)and(d))in the inter-electrode gap.As soon as spark is over,the gas bubbles again formed around the cathode and the cycle is repeated [29].A minor deviation in the voltage-current graph is observed in the PMECDM process (Fig.8(a)) as compared to the classical method.The initiation of the gas bubble formation occurs ((a) and (b)),followed by increasing the current density in the machining cell((b)-(e)).However,the presence of the powder particles burst some of these bubbles ((b)-(d)) and causes the formation of a sharp current graph as compared to the classical method while attending to critical machining voltage((d)and(e)).At last,spark discharge occurs ((e)-(h)) in a discrete manner((f)-(h))due to the spreading of sparks by the powder particles[9].In the PMECDM,the current peaks were observed to be sharp and multiple discharge peaks were noticed.This leads to a decline in the mean current value from 270 mA to 262 mA.This concludes that the presence of Al2O3powder in the solution disperses the discharge energy and reduces its intensity,contributing for the enhancement of the surface quality.However,the nature of the plot may vary depending on the conductivity of the powder particles mixed in the electrolyte solution.

Fig.8.Voltage-Current signal plot for (a) Powder mixed (b) Classical MECDM process.

Fig.9.Surface plots (a),(c),(e) of MRR and (b),(d),(f) PMRR at different input machining parameters.

Fig.10.Single parameter optimization for MRR at (a) Classical and (b) hybrid ECDM process.

4.2.Analysis of material removal rate

Figs.9(a)-9(f) shows the comparative surface plots between classical and powder mixed ECDM (PMECDM) processes for different input machining parameters.The graph reveals that PMECDM has a higher material removal rate than classical ECDM process.A similar observation is reported by Varghese et al.[14].In the classical method,the material is removed by thermal erosion and vaporization followed by thermally promoted chemical etching[32].It is noticed that the voltage,duty factor,and electrolyte concentration show an increasing trend of MRR in both processes.The hybrid approach shows maximum improvement of MRR up to 67% as compared to the classical one.Elhami et al.[23]also reported a similar observations that is enhancement of MRR nearly by 18% compared to the classical method while performing the machining on glass with nano Al2O3powder mixed electrolyte.In the hybrid process,MRR varies from 1.35 to 3.43 mg/min,whereas for the classical method it ranges from 0.79 to 2.80 mg/min.It has been noticed that the application of powder improves the quality and frequency of sparks.Further,the grinding action of abrasive particles helps to remove the partially melted material from the machining zone and thus enhances the machining performance with respect to the classical process [10].

Figs.9(a)-9(f) depicts that by increasing the voltage (V),duty factor(D)and electrolyte concentration(C),MRR considerably rises for both the processes.Among the input machining parameters,the variation in voltage shows a prominent impact on MRR in comparison to the other two parameters(Figs.9(a)and 9(b))in both the machining approach.This can be elaborated as with the rise of machining voltage,bubbles formation increases due to faster electro-chemical reaction inside the cell and resulted in the formation of intense sparks.A similar observation is reported by Zhao et al.[21].On the other hand,with the increasing duty factor,the duration of the spark increases thereby delivering high thermal energy for a longer period,thus enhancing the MRR[15].Similarly,with an increase in electrolyte concentration,ions mobility improves in the solution which contributes to the chemical etching followed by the generation of intense sparks [8].

Table 5 depicts the analysis of variance (ANOVA) for both the machining approaches.From the F-test results,it is confirmed that voltage (V) is the most significant parameter in both machining approaches.However,electrolyte concentration (C) followed by duty factor (D) is significant in MECDM whereas duty factor followed by electrolyte concentration in the PMECDM process[13-15].In the PMECDM process,the presence of powder particles resulted in low mobility of the ions,thereby the influence of concentration (C) is reduced as compared to the duty factor [9,13].Further,the R2value is more than 90% for both the machining approaches which indicates the established relationships among the process parameters and output responses are well fitted[24].

Table 5ANOVA for MRR.

Figs.10(a) and 10(b) shows single parameter optimization of MRR for both the machining approaches at maximum output conditions.The obtained optimum MRR for classical and hybrid approaches are 2.95 and 3.53 mg/min respectively which are obtained at parameter setting of applied voltage: 45 V,duty factor:35%,and electrolyte concentration: 20 wt% in both the machining approaches.For both conditions,desirability value is nearly 1 which indicates the suitability of the model.

4.3.Effect of parameters on surface roughness

Figs.11(a)-11(f)shows surface roughness plots for classical(Ra)and PMECDM approach (PRa) at different input machining parameters.From the graphs,lower surface roughness was observed in PMECDM as compared to MECDM process.The lower surface roughness is due to the grinding action abrasive powder mixed in the electrolyte [10],which improves the quality of the micro hole wall surface.In PMECDM process,the PRa varies from 0.495 μm to 1.568 μm,whereas for the classical Ra it ranges from 0.972 μm to 1.975 μm.Similar observation was reported by Kuo et al.[10]while machining of Quartz through WECDM methods at powder (SiC)concentration varying from 0 wt% to 5 wt% mixed in the 5 M KOH solution.

At similar machining parameters i.e.applied voltage:40 V,duty factor: 30% and electrolyte concentration: 15wt%,the PMECDM process produces a minimum surface roughness of 0.49 μm which is approximately 49.02% less compared to the MECDM process.The application of powder particles in the electrolyte lead to grinding action that help to remove the partially melted material from the machined surface in addition to their abrasion and polishing actions [10,23].

The surface roughness decreases with increase in voltage and electrolyte concentration(Figs.11(a)-11(d)).It was noticed that Ra decreases approximately by 0.3 μm (MECDM) and 0.5 μm(PMECDM) when voltage changes from 35 V to 40 V and then increases approximately by 0.8 μm(MECDM)and 0.7 μm(PMECDM)with rising in voltage from 40 V to 45 V.A similar observation was reported by Kuo et al.[10].At low voltage,sporadic discharges occur due to the formation of unstable gas film over tool surface.Above 40 V,the gas film was unstable and abnormal discharge was observed resulting in a bigger-sized crater and the surface roughness increased [4,15].

A similar trend was also observed with the rise in the concentration of the electrolyte solution.At 15 wt% concentration(Figs.11(c)-11(f)),the minimum surface roughness was observed.Since at low concentration,electro-chemical etching is minimum.Irrespective of the high rate of chemical etching that occurs at higher concentration,the formation of intense sparks nullify the impact of the chemical etching and increases the surface roughness[15].

A decreasing trend of Ra was noticed with increasing the duty factor in both the machining approaches(Figs.11(a)and 11(b)and Figs.11(e) and 11(f)).At 15 wt% concentration,duty factor ranging from 25% to 35%,the surface roughness significantly declines.It can be concluded that with an increase in the spark duration,the temperature of the electrolyte near the sparking zone increases which promotes the chemical etching and improves the surface finish [24].

Table 6 depicts the analysis of variance(ANOVA).From the F-test results,it is confirmed that the voltage is the most significant parameter for both machining approaches,followed by duty factor(D) and electrolyte concentration (C) for MECDM;whereas electrolyte concentration and duty factor for the PMECDM.Further,for both machining techniques,the R2values are greater than 90%,demonstrating the suitability of the established correlations between the process parameters and output responses [33].

Table 6ANOVA for Ra.

Figs.12(a) and 12(b) depicts the single parameter optimization having desirability tends to 1.It was applied to know the suitability of input machining parameters for obtaining minimum micro hole wall surface roughness [33].It was observed that minimum Ra for classical and hybrid ECDM approaches are 0.92 μm and 0.42 μm respectively at optimum machining conditions,i.e.,applied voltage:40 V (approx.),duty factor: 35% (approx.),and electrolyte concentration:14 wt% (approx.).

4.4.Analysis of surface topography and morphology

The FESEM images of the cross-sectional view of the drilled micro holes were analysed to understand the uniformity of the micro hole along their axes.Fig.13 depicts the cross-sectional view of micro holes fabricated with((a)-(f))classical ECDM and((g)-(l))PMECDM approach.As observed from the images,the classical ECDM process shows a rough wall surface as compared to PMECDM,which indicates that the mixed abrasive powder performed grinding action and help to obtain a smooth surface.But,the presence of powder leads to the enlargement of the hole diameter due to dispersion of spark along with grinding action at the edges of the micro holes [23].The combination of low voltage(35 V)and electrolyte concentration of 10wt% and 15wt%(for both the machining approach) produces inverted cone shaped through holes(Figs.13(a)-13(c)and Figs.13(g)-13(i))at duty factors of 25,30 and 35% respectively.Since at low energy input,the machining operation became slow which directly impacts the top surface of the hole,as the duration of chemical etching and spark discharge is more [24].This leads to a larger diameter at the top surface,whereas with an increase in depth,the sparking action and electrochemical dissolution rate deteriorated significantly,leading to tapered hole [34].

Fig.13.FESEM image of a cross-sectional view of micro-hole fabricated with ((a)-(f)) classical and ((g)-(l)) hybrid ECDM processes at different input machining parameters.

On the other hand,increasing the voltage(45 V)and keeping the electrolyte concentration similar i.e.10wt% and 15wt% shows that fabricated through micro hole is almost parallel geometry(Figs.13(d)-13(e) and Figs.13(j) and 13(k)) at duty factor 30 and 35% respectively.The edges of these holes also depict some burrs and debris.Furthermore,at high voltage (45 V) and electrolyte concentrations (20wt%) and a duty factor of 30% shows almost straight holes with edges have significant burrs or debris in both processes(Figs.13(f)and 13(l)).This can be due to the high energy input,the discharge action is more prominent as compared to the electrochemical dissolution action.Hence,drilling through a micro hole was attained at a minimum time and a lesser duration was available for electrochemical action [35,36].

The magnified images of the surface were captured(as shown in Figs.14(a)-14(d)) to compare the surface morphology.The three areas i.e.near the hole entrance,the middle and bottom part of the hole were considered for the analysis.Two sets of experiments(from both processes having similar machining parameters) were selected for this investigation.The experiments performed at high and low levels are favourable to understanding the quality of the processes.Hence,the chosen machining parameters are applied voltage:35 V and 45 V,electrolyte concentration:10wt% and 20wt%,at duty factor: 30% respectively.Figs.14(a) and 14(b) shows the surface obtained by machining with MECDM whereas Figs.14(c)and 14(d) for PMECDM.It is observed that PMECDM produces a smooth and clean surface with less debris and particles.It is also noticed that MECDM has developed micro-cracks at the middle part (Fig.14(b)) of the hole due to the increasing energy input which is absent in other processes.However,the presence of blow holes,craters and recast layer was noticed in both processes but the density was more in the PMECDM process.This can be elaborated as in the conventional approach the surface is covered by the debris particles hence these defects are not easily visible.However,the grinding action by the abrasive particles helps to remove these debris particles and exposed the surface more clearly.Further,at low energy input,the PMECDM shows a cleaner surface with respect to the classical approach.However,this input energy results in a low material removal rate.To overcome this problem,multiobjective optimization was carried out to identify the optimal parameters suitable to obtain high surface quality with more material removal rate.

Fig.14.Enlarged view at different locations of the micro holes along their axes ((a) and (b)) MECDM and ((c) and (d)) PMECDM process.

4.5.Multi-objective optimization

The output parameters viz.material removal rate (MRR) and surface roughness (Ra) are considered the response parameters in the present work.The desirability function approach is used for multi-objective optimization.This approach transforms the response into a dimensionless quantity that ranges from 0 (undesirable characteristic) to 1 (desirable characteristic) [33].The composite desirability has been calculated by considering the geometric mean of all the individual desirability values [37].The optimization of output responses is carried out by considering the maximum MRR and minimum SR for both machining approach(Fig.15).

Fig.15.Multi-objective optimization of MRR and SR for classical and hybrid ECDM processes.

The multi-objective optimum parameters(at approx.42 V,35%,and 16wt%) for maximum MRR and minimumRafor the MECDM process are 2.01 mg/min and 1.11 μm,respectively.On the other hand,for PMECDM the MRR andRavalues are 2.77 mg/min and 0.61 μm respectively,at the utmost desirability of 0.75.The optimal response value for the parameter settings is depicted in Table 7.For the model's validation,the confirmation test was conducted at the optimum parameter settings.In Table 7,the outcomes of the confirmation experiments,predicted values,and errors are displayed.The error percentage among the predicted and experimental results reveals an error within 5% for both MRR andRawhich indicates the outcomes that were obtained closely match those that were predicted.

Table 7Comparison of predicted and experimental results.

4.6.Surface topography analysis at optimized parameters

The morphology of the hole wall surface obtained at optimized parameters i.e.V＝42 V,D＝35%,C＝16wt% was analysed through a non-contact type surface profiler and FESEM for both the machining approach (Figs.16-18).Figs.16(a) and 16(b)) indicates the 3D profile meter images of the sectioned holes captured for both MECDM and PMECDM processes,respectively.The distortions in the images are due to the asymmetric sectioning of the samples.The surface produced through the PMECDM process appears to be clean (Fig.16(b)) as compared to the MECDM process (Fig.16(a)).

Fig.16.Topographical images captured for (a) MECDM and (b) PMECDM process.

Fig.17.(a),(c)Surface topography of micro hole wall obtained at optimized parameters and(b),(d)contour curve of a respective hole machined through(a),(b)PMECDM and(c),(d) MECDM approach.

A random surface area was selected for the measurement of roughness profiles.Figs.17(a) and 17(c) represent the 2D surface profile whereas Figs.17(b) and 17(d) depict surface roughness(contour curve)profiles machined through PMECDM and MECDM,respectively.Shock waves are produced during the discharge process,which energises the abrasive particles and creates a rubbing action which is helpful for the in-situ removal of debris from the machined surface [38].Therefore,the PMECDM produced clean surfaces as compared to those obtained from MECDM.

Fig.18 represents the FESEM images of the surface topography of the hole walls obtained at optimum parameters.Figs.18(c) and 18(d)are magnified views of the micro hole wall surface.Along the depth of the hole,three regions were selected(Figs.18(e)and 18(f))i.e.top(T),middle(M)and bottom(B)for the hole surface analysis and comparison.It is observed that the presence of debris particles,blow holes,and micro-cracks on the hole wall were produced through the MECDM process(Figs.18(b),18(d)and 18(f)),whereas in the PMECDM process (Figs.18(a),18(c) and 18(e)),the above flaws were minimum,thereby leading to superior surface quality.Better surface integrity was achieved in the latter case due to the grinding action of the Al2O3particles present in the electrolyte.

Fig.18.FESEM image of sectioned and zoomed view of micro hole obtained at optimized parameters at(a)PMECDM(b)MECDM approach.(e)and(f)FESEM images of a different area of the hole wall surface fabricated with (a) PMECDM a (b) MECDM approach.

4.7.Energy-dispersive X-ray spectroscopy (EDS) analysis

The EDS analysis of the micro hole wall surface fabricated at optimum machining parameter settings for both the machining approach(Fig.19)indicates the transfer of tool material(tungsten)to the work surface due to the melting of the material during the discharge [4].However,relatively low material transfer was observed in PMECDM due to the dispersion of discharge energy by the micro-abrasive particles.The presence of aluminium in the PMECDM surface is due to the transfer of aluminium from the Al2O3powder.It is affirmed that the surface integrity of the machined surface in case of the PMECDM method is better than the MECDM process.

Fig.19.EDS analysis of the micro hole wall surface fabricated by (a) hybrid (b) classical MECDM approach.

5.Conclusions

Micro-machining operation on hard-to-machine C103 niobiumbased alloy was carried out using MECDM and PMECDM techniques.The investigation was focused on the enhancement of material removal rate and surface finish.To these aspects,a comparative study was carried out to evaluate the machining performance during deep hole drilling operations.Multi-objective optimization was performed to simplify the analysis process.The following conclusions are drawn from this study:

· The applied voltage (V) was the prominent factor among the used input parameters i.e.electrolyte concentration(C)and duty factor(D) in both the processes (MECDM and PMECDM).

· The PMECDM show multiple sparks due to the presence of powder in the electrolyte solution which help to enhance the surface quality of the machined surface as compared to MECDM process.

· The optimum MRR with minimum Ra was obtained at applied voltage ＝42 V,C＝16 wt % and DF ＝35% through multiobjective optimization method.The confirmation experiments depict an error within 5% of the predicted results.

· In all the parameter settings,the PMECDM shows higher MRR and lower surface roughness as compared to the MECDM process.The highest MRR of 2.8 mg/min and corresponding PRa of 0.61 μm were observed in PMECDM,whereas MRR of 2.01 mg/min and corresponding Ra of 1.11 μm were observed in the case of the MECDM process.

· The PMECDM process exhibits larger hole diameter as compared to MECDM process.

· The transfer of tool material (tungsten) onto the work surface was confirmed through EDS plots and the transfer of aluminium from Al2O3particles was also confirmed.However,these abrasive particles are beneficial in the removal of deposited debris from the hole wall.The efficiency of this PMECDM process may be improved further by introducing ultrasonic vibration either to the electrolyte solution or to the tool,which needs further exploration.

Author contributions

All authors contributed to the study conception and design.Material preparation,data collection and analysis were performed by Niladri Mandal.The first draft of the manuscript was written by Nitesh Kumar.Alok Kumar Das supervised the finding of this work.All authors read and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.