Characterization of electric discharge machining of titanium alloy utilizing MEIOT technique for orthopedic implants

In this research work, medical grade titanium alloy Ti4Al6V was electric discharge machined with an objective of attaining mirror finish for orthopedic implants. Experiments were conducted by varying tool materials, discharge current, pulse on time and pulse off time whereas the responses chronicled are material removal rate, electrode wear rate and surface roughness. The aluminium (A), copper (C) and aluminium alloy reinforced with graphite particles of various weight percentage (5-A1,10-A2,15-A3) were used as tool materials. The composites were fabricated using stir casting technique. The findings showed that the titanium alloy machined with A1 composite tool offers the highest MRR, the C tool has the lowest EWR, and the A2 composite tool results in good surface finish. The surface of specimens produced using A1 tool exhibits poor surface quality owing to the eczema surface. Specimens machined with the C tool have a remelted layer, pockmarks, and an uneven fusion structure, which were not present in specimens machined with the A2 tool. MOORA-ELECTRE Integrated Optimization Technique (MEIOT) was applied to select the best parametric combination and the best electrode material.


Introduction
The materials predominantly used as medical implants are nitinol, stainless steel, and cobalt-chromium alloys. The strength, elasticity and stiffness are the basic properties for materials in human implants [1][2][3]. Because of its harmonizing factor, Ti4Al6V, a medical grade titanium alloy, was recommended as orthopedic implants [4]. The substrate scuff caused by conventional processes results in certain electrochemical responses [5]. In general, the lot of literature incorporates research on mechanization and surface experiments for industrial titaniumbased alloys using conventional machining methods [6][7][8].
Suitable methods of machining must be selected to produce implants of good surface [9]. The Electric Discharge Machining (EDM) was suitable to manufacture medical implants with a high degree of accuracy [10,11]. The major process parameters influencing machining efficiency are current, pulse on time, pulse off time, voltage, gap width and dielectric medium [12][13][14][15]. From the literature survey, it was revealed that current and pulse on time are the most dominant factors influences the characteristic of the machining [16,17].
Selecting suitable combination of electrode materials and workpiece yields better results. Fly ash reinforced AA6061 composite was EDM'ed utilizing copper and brass electrodes. When brass was used as an electrode, maximum MRR was achieved, and less EWR was registered for copper electrode [18]. In case of SiC reinforced AA6061 composites copper and brass electrode showed better MRR and R a value respectively [19]. The same results were obtained when machining the Nimonic 75 alloy [20]. The graphite, tungsten, brass and copper tool were used for the machining of INCONEL 825 super alloy. As opposed to unflushed machining, flushing of machined debris improved MRR. The findings showed that thermal conductivity of the electrode influences the MRR and R a [21].

Materials and methods
About 1kg of AA6061 was kept in a graphite crucible and its chemical composition was depicted in the table 1. The crucible was heated to the temperature of 700°C and preheated graphite (Gr) particles of average particle size 5 μm was added to the charge. The mixture was stirred for 180 s using a mechanical stirrer at a speed of 1000 RPM. An equal weight proportion of Potassium Titanium Fluoride (K 2 TIF 6 ) was added as a flux to improve the wettability of reinforcing particles. The mixture was stirred for 120 s after the flux was added. The charge was then poured into the preheated mould made of die steel. The same procedure was followed for the manufacturing of composites with different weight percentage. The AA6061/5%Gr (A1), AA6061/10%Gr (A2), AA6061/15%Gr (A3), Aluminium (A) and Copper (C) was used as an electrode material. The EDM experiments were carried out on medical graded titanium alloy (TI6Al4V) by varying electrode materials, Pulse on time, Pulse off time and discharge current. The selected parameters were varied at 5 level as shown in table 2 and experimental runs were designed using L25 Taguchi orthogonal design. The MRR and EWR were determined by measuring the ratio of weight difference before and after machining to the product of machining time and density, as shown in equations (1) and (2). The specimens were machined for the time span of 10 min. The surface roughness was measured using the SJ210 testing machine manufactured by Mitutoyo. The value was measured on ten different points on the specimen surface and average was recorded as surface roughness. Scanning Electron microscope was used to examine the surface morphology.

MOORA-electre integrated optimization technique (MEIOT)
The MEIOT method was used to determine the right parametric combination and best electrode material. The efficiency of the alternatives was compared to each other and the one with the most exaggerated performance was chosen as the best. The available alternatives for assessment in this dilemma are A, A, A1, A2, and A3. The procedure began with the formulation of a decision matrix, since the engineering problem had 25 experimental runs, which were denoted in the I element, and it was analysed with the aid of three responses, namely MRR, EWR and Ra, which were denoted in the j element, resulting in a decision matrix of 25×3 as shown in table 3. Following that the matrix was normalised utilizing the equation (3) [39]. Correlation between these normalised value and standard deviation of the experimental response was determined as per the equation (4). A matrix was generated by multiplying the correlation and standard deviation values. If the problem has 'n' responses, a matrix of 'nx1' was formed as depicted in equation (5); in this scenario, a matrix of '3×1' has been constructed [40]. The weight of each criterion was determined by the ratio of these matrix values to the summation of all the values in the matrix as shown in equation (6)   å - The next step consisted of forming a weighted normalised matrix generated from product of each criteria's weight and normalisation matrix as depicted in equations (7) & (8). The next step was the calculation of assessment value which is the difference between the weighted normalised value of beneficiary attribute to the non-beneficiary attribute as shown in equation (9). The experiment with highest assessment value was taken as best parametric combination [42] as shown in table 4.  In order to identify the best electrode material an alternative decision matrix (Qij) was formed according to the equation (10). It is the sum of all the values of the ith alternate's weighted normalised decision matrix as shown in table 5. Qij was normalised and multiplied with criterion weight to form the weighted alternative decision matrix (Rij) as shown in the equation (11).
The concordance element Cij was formed by comparing the performance measure of electrode with each other [43]. When the value of beneficiary response was higher than the alternative, the equivalent solution weight has been taken and the non-beneficiary criteria calculated vice versa as shown in equation (12). The alternative A1 was compared with A2 in such a way that R11>R12, R21>R22 and R31<R33. The value of C11 was 0.719. Similarly all other alternatives are compared with each other as depicted in equation (13) and its value was shown in table 6. Following Č was determined which was the average value of all non-zero element in the concordance set [44]. If the concordance element was higher than Č, the value was one or zero in concordance matrix Cij, as shown in table 7.  The ensuing stage was the calculation of the discordance element Dij as shown in equation (14). The discrepancy between the alternatives was computed and the value of discordance is the proportion between the most negative and the highest value [45]. For example to compute the element D 12 (R 11 -R 21 ), (R 12 -R 22 ) and (R 13 -R 23 ) was calculated. The ration of maximum negative value of these three elements to the maximum value among of these three elements was taken as D 12 as depicted in equation (15). Similarly Ď was the average value of all non zero elements in the discordance set [46] as shown in table 6. If the discordance element was higher than Ď, the value was one or zero in discordance matrix Dij, as shown in table 7.
Max nij Max ni nj   [49]. Owing to this the materials removed from the work piece as well as from the electrode. The industry demanded a low EWR and a high MRR, as well as an excellent surface quality. The MRR rises with increasing current as it facilitates the melting an evaporation [50] until the saddle point, beyond this limit it triggers energy destabilization [51] which reduces the MRR as shown in figure 1. Destabilization occurs at different point for distinct electrodes, for electrodes A and A1, it was 14A, and for electrodes A2, A3, and C, it was 21 A. Of the five different electrode A1 composite tool offers highest    MRR. As discussed earlier while machining some of the materials were removed from the electrode surface, owing to this the graphite particles get detached from the surface and enter inside the spark gap. When a voltage was applied to these particles, they become energized and travel in a zigzag pattern [52,53], resulting in the bridging effect. It generates multiple discharges in a single flash, which results in faster sparking and impoverishment of the workpiece surface [54,55]. With increase in graphite content the MRR reduces owing to the occurrence of short circuit. when machined with a standard C electrode, it has a lower MRR than A1 but a higher MRR than other electrodes used for experimentation. When machined with composite tool, the MRR decreases as the pulse on time increases, which contradicts the pattern found by other researchers [56,57]. Because of the bridging effect, higher discharge energy and spark intensification occur at lower pulse on time.
The plasma channel widens [58] as the pulse on time increases until it reaches a saddle point of 45 s, reduces MRR. For each additional increment above 45 s, plasma channel densification happens, which increases MRR. As C and A were used as electrodes, MRR increased until it reached a svec, after which it began to decrease, as observed by several researchers. The MRR decreases with increase in pulse off time owing to the reduction in discharge energy. The highest MRR was observed when the sparking time was 3.75 μs per cycle.

Influence of process parameters on EWR
The electrode with the highest melting temperature has the lowest EWR [59], which was well correlated with the results that the copper tool with a melting temperature of 1085°C has the lowest EWR as compared with aluminium electrodes as depicted in figure 2. Interestingly the EWR was reduced when graphite was added as reinforcement. Due to the tiny arcs that occur during production, the intensive movement of electrons will reverse the feed path to maintain an even greater chimney gap [60]. The EWR value could then be lowered quickly since most negative ions pass into the phase breakdown [61]. The EWR raises with increase in discharge current owing to the higher spark intensity. When A1 was used as an electrode, it had the lowest EWR with increasing current as compared to other composite tools due to complete heat dissipation because of increment in spark gap [62]. Owing to the short circuit, A2 and A3 have a higher EWR. The EWR decreases as the pulse on time increases until it reaches a tush point of 45 s. As previously mentioned, the materials extracted from the electrode were relegated as a result of the decrease in spark power [63]. The plasma densification evaporated the tool metal beyond the tush limit, increasing EWR. The Electrode machined with a pulse off time 2 μs posses highest EWR as it declines sharply when the parametric value was set at 4 μs thereafter it increases gradually with increase in pulse off. The least EWR was achieved when the off time was kept at 7.5 μs/cycle.

Influence of process parameters on R a
The surface quality of the product worsens with increase in the input current as depicted in the graph. When the current was 7A, the products are manufactured with the average Ra value of 2.699 μm and it was drastically increased to 6.527 μm when the discharge current was tuned to 14A. The Ra value increased with discharge current regardless of the electrode used as shown in figure 3. This was attributed to the fact higher current produces high spark intensity [64] which creates deep craters and crack on the surface hence surface quality reduces [65]. The EDM process machined the metal with superior surface quality when the parametric value of pulse on time was set to 30 s. The transformation from good to bad surface happens when the unit is tweaked to 45 s, and it worsens with further increase in parametric value. This was ascribed by the fact multiple discharge in single flash creates cracks on the surface, many researchers reported the similar findings [66][67][68]. With increasing pulse off time, the Ra value increases until a saddle point of 4 μs further begins to decrease. The electrode A and C produces the material of higher Ra value, it was drastically reduced when graphite was added to aluminium. While machining the particles that present inside the eroding area increases the spark gap. Hence the melted material are completely flushed away from the surface which eliminates the formation of remelted layer on the surface [69]. With increase in weight percentage of graphite particles more foreign particles hang in the spark gap which hinders short circuit, hence Ra value increases [70]. Manufacturing titanium implants with the mirror surface finish was the ultimate objective of the work. When machined utilizing A2 tool at the discharge current of 7A a minimum surface roughness value of 0.847 μm was attained.

Surface topography
The surfaces of the titanium alloy machined using the tools A1, A2 and C having surface roughness of 1.37, 0.847 and 2.683 μm respectively was investigated using SEM. The surface topography of titanium alloy machined with copper tools showed black spots, craters and micro pits as shown in figure 4(a). The surface also displayed remelted layer which occurs because of incomplete flushing. At higher magnification the surface topography showed micro cracks, deeper craters and uneven fusion structures as shown in figure 4(b). It was evident that the heat was not completely dissipated when copper was used as an electrode. Pockmarks were observed on the surface which was formed due to the release of entrapped gas during the cooling phase.
The occurrence of eczema on the EDMed work surface obtained with the A1 composite tool electrode was a notable feature as shown in the figure 5(a). This eczema was white in colour and spherical in shape and spread all over the surface. Apart from that machined surface showed globules and scratches. At higher magnification this eczema appeared as the lot of tiny remelted layer as shown in figure 5(b). Owing to this A1 composite tool offers 61% worsen surface as compared with A2 tool. Micro pits and globules were also observed on the surface.
The surface topography of the titanium alloy machined with A2 composite alloy showed black spots and minute scratches as shown in figure 6(a). The black spots are formed owing to the deposition of carbon content from the dielectric fluid. Pits, craters and cracks were not formed on the surface which reveals that the heat was completely removed from the machined area owing to the increase in spark gap. The formation of remelted layer was controlled to a greater extent but was not completely eliminated. It showed some melted materials are redeposited over the surface during the colling phase. At higher magnification the surface texture showed globules, remelted layer and wrinkled shape as shown in figure 6(b). This wrinkled shape was formed due to the release of debris from the electrode material. As more flashes occurred in a single cycle which results in the formation of globules.

Conclusion
The aim of the work was to achieve the best surface quality of titanium implants, for which composites tools as well as traditional tools were used. Five electrodes namely copper, aluminium and aluminium reinforced with graphite particles of 5, 10 and 15 weight percentage fabricated using stir casting technique were employed for EDM. The impact of discharge current, pulse on time and pulse off time over MRR, EWR and R a were analysed and following conclusions were drawn.
(1) Maximum MRR was achieved due to the bridging effect when A1 composite tool was used as electrode; however, as the weight percentage of graphite increases, MRR decreases due to the occurrence of short circuit. The copper tool with the highest melting point among the chosen electrodes has the lowest EWR.
Owing to the increase in spark gap the R a value decreases when A2 composite tool was used as electrode.
(2) MRR improves with increase in current intensity until a saddle point, thereafter it declines because of energy destabilization. Higher pulse on time causes plasma channel expansion, which lowers MRR, while lower pulse on time causes plasma intensification that improves MRR. The discharge energy decreases as the pulse off time increases, resulting in a lower MRR.
(3) Because of the increased spark intensity, the EWR increases as the discharge current increases. EWR increases due to plasma densification with increased pulse on time. R a roughness increases with increase in discharge current and pulse on time. When composite tools are used, debris is totally flushed away due to the increase in spark gap, which prevents re-solidification of materials over the surface, which improves R a .
(4) The surface topography showed black spots, globules and redeposited particles which were absent on the surface machined using A2 composite tool. Eczema like surface was observed when A1 composite tool was used, at higher magnification it was revealed as the cluster of tiny remelted layer. A2 machined surface showed black spots and globules, formation of remelted was prevented owing to comprehensive heat removal which improves R a .
(5) MEIOT optimization technique was utilized to select the best parametric combination and electrode material. Titanium alloy machined with parametric value of 60 μs pulse on time, 8 μs pulse off time and 7A current machined using A2 composite electrode increases productivity.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).