Examining the surface roughness and kerf quality of micro-slots cut on the surfaces of Ti-B4C nanocomposites by WEDM: a desirability approach

Micro slots and textures are created on Titanium (Ti) composites to improve its surface characteristics. Micro-textured Ti composites are generally recommended for bio implants, automobile, and aerospace components. In the current research, Ti-B4C nanocomposites were prepared by powder metallurgical route. Micro slots were cut on the Ti-B4C surfaces by Wire Electrical Discharge Machining (WEDM) Technology by varying the current, pulse-ON time, and pulse-OFF time. Scanning electron microscopy and XRD analysis validates the uniform distribution and inclusion of B4C nanoparticles in Ti matrix. Response surface methodology was used to plan the experimental runs. Analysis of variance and desirability analysis were employed to identify the most suitable machining factors for obtaining the minimum surface roughness, lower kerf width and higher material removal rate (MRR). Increase in applied current and pulse-ON time, increases the MRR. Increase of pulse-OFF time from 50 μs to 60 μs gradually reduces the MRR and reduce the surface roughness of the cut slots. Contrastingly an increase in pulse-ON time increases the roughness due to an extensive melting and resolidification of Ti nanocomposites. The morphology of the WEDMed surface reveals the recast layer and localized melting zones on the cut surfaces.


Introduction
For the aerospace and bio medical industries, titanium (Ti) is a crucial metal since it is both strong and moderately light. Ti is also employed in the production of gas turbine parts, intake valves, connecting rods, and other elements for diesel engines [1,2].Ti alloys have a tensile strength of approximately 1000 MPa which is twice that of Al alloys [3]. For challenging applications where high strength and stiffness, creep resistance, and high corrosion resistance are the main concerns, Ti composites have been proposed [4,5]. Fibers, whiskers, and particles are the most prevalent reinforcement materials employed in the Ti matrix [6]. Recent trends is to reinforce the metal matrices by nano-reinforcements to get the extraordinary improvement in mechanical properties [7,8]. Particulate reinforcements perform better among the employed reinforcements when internal structure failures and high production costs associated with fiber and whisker-based composites are taken into considerations [9].
Boron carbide (B 4 C) is a crucial material for heavy armor and high temperature resistant parts due to its ability to maintain its hardness at temperatures around 1100 and 1300°C [10]. It appears to have good thermal and chemical stability, low density (2.52 g cm −3 ), and rigidity [11]. B 4 C, SiC, and TiC are the widely recommended ceramic reinforcements for metal matrix composites [12,13]. The inclusion of nanoparticles to the metal matrix results in a more refined grain structure and a non-basal slip system, which increases the ductility [14,15]. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
The increase in mechanical characteristics was caused by the size effect of the nanoparticles, which opened the door for other strengthening mechanisms [16][17][18]. Modern friction theory which considers the surface aspects such as hardness discrepancy and atom shifting between friction pairs [19]. The two most crucial powder metallurgy factors which guarantees the optimal performance of Ti composite are homogenous powder mixing and dispersion [20]. While machining Ti alloys, cutting-edge heat increases significantly which reduces the tool life [21]. Also, the feed rate radically affects cutting speed and feed force [22].
The principal failures of the tool during the conventional machining of Ti have been demonstrated to be flank and nose wear [23]. Compared to monolithic Ti, the inclusion of 2% TiB 2 and 2% B 4 C particles in Ti produced microhardness that was 59% and 41% higher, respectively [24]. From hardness testing, it was found that the sintered specimens hardness properties were higher than those of the parent material [25]. Although Ti alloys and composites can be machined with PCD cutting tools more effectively, their use is restricted by their high tooling costs [26]. One of the tested non-conventional machining techniques for cutting Ti components is WEDM [27]. The temperature of the spark rises between 8000°C and 10000°C in the regionalized zone, melting and evaporating a little amount of the substance from the work piece [28,29].
The findings of the WEDM experiments demonstrated that the dominant performance parameters are considerably current and pulse-on-time. Taguchi approach and an artificial neural network were utilized to optimize and forecast the surface roughness values [30]. It is reported that an ANOVA was done to get the prevalent input and output values [31]. Authors used the hybrid methodology by combining TOPSIS and analytical hierarchy procedure to identify the predominant variables over measured responses [32]. In EDM, the material erosion process converts electrical energy primarily into heat energy [33]. A smaller exposed area with a 0°taper angle experiences high erosion intensity because of high current levels which leads to significant particle erosion in the area and greater kerf widths [34]. Numerical models were developed to explore the interactions between several WEDM factors and responses while machining Inconel 601 [35]. According to the results of the experiments, the kerf length, power density, server voltage, and wire tension all increase as the pulse on time increases [36,37]. It is noted that researchers trying to generate surfaces with least roughness and low kerf width [38]. Ti tends to weld with the cutting tool due to their strong chemical reactivity, which can result in chipping and early tool failure [39]. The ideal settings were found using a Taguchi coupled with desirability analysis with the goal of maximizing MRR and minimizing Ra [40]. The highest value of composite desirability was used to arrive at the ideal level of WEDM parameters [41]. The best kerf quality for the material tool steel of grade DF-2 was obtained by WEDM using Taguchi quality design and analysis of variance [42]. RSM and advanced algorithms were employed for the multi-objective optimization of machining parameters in turning Ti alloys [43].
Generally micro slots and textures are produced on Ti alloy surfaces to enhance its surface characteristics. The mechanical strength of Ti alloys can be further improved by the addition of B4C nanoparticles. However, an introduction of B4C nanoparticles causes extensive tool wear while machining. The low thermal conductivity and work hardening ability of Tit reduces the employment of conventional machining. Micro slot cutting on Ti-B4C nanocomposites was performed using WEDM technology by varying the current, pulse on time, and pulse off time with an objective to obtain minimum surface roughness/Kerf and maximum MRR. ANOVA and desirability analysis was employed for this purpose.

Materials for sintering
Owing to the remarkable corrosion resistance and structural properties Ti was chosen as matrix material. For the synthesis of Ti-B 4 C nanocomposites, Ti powder (99% purity) and nano B 4 C powders (99.5% purity) were employed as matrix and strengthening constituents, respectively. The size of the nanoparticles of the Ti and B 4 C powders employed was 10 μm and 50 nm, respectively. Table 1 illustrates the mechanical properties of Ti and B 4 C.
The original powders (99.5 wt% and 0.5 B 4 C wt%) were mixed in Ti ball mills in a vacuum assisted atmosphere as the vacuum atmosphere reduces the oxidization of Ti. Zinc stearate was used as binder material. The process took about one hour to perfectly blend the powders; these blended powders were compacted using hydraulic press with 80 bar pressure where the load applied was uniaxially distributed over the surface. The die was made of die steel with 12 mm diameter. Totally 20 specimens were compacted with dimensions 12 mm diameter and 15 mm height. The specimens after compaction were subjected to sintering process in vacuum assisted tubular furnace. Rate of heating and cooling was fixed as 10°C per minute and the sintering temperature was maintained at 1300°C. Time taken for sintering was about 5 h and 20 min including a dwell period of 2 h. At a temperature 1300°C, a small portion of Ti has fused. Different stages of fusion are shown in figure 1(a). The sintered samples used for WEDM experimentations is depicted in figure 1(b).

Microstructure of Ti-B 4 C nanocomposites
The sintered nanocomposites were assessed via scanning electron microscopy (SEM) to investigate the dispersion pattern of nano-B 4 C in Ti. The SEM images were depicted in figure 2. Individual B 4 C nanoparticles are spotted (highlighted with circles in figure 2(a)) in the SEM micrographs. It was obvious that fusion of Ti took place at 1300°C, which helped to form a bond between titanium powders and boron carbide nanoparticles. This phenomenon is indicated in figure 2(b) whereas B 4 Cpowders are not melt like Ti powders as its melting point is high, around 2300°C.

Wire cut electrical discharge machine (WEDM) Setup
Sintered specimens were machined using WEDM process. The WEDM (Smartcut make) setup used in cutting process is shown in figure 3. The WEDM setup includes the dielectric chamber, wire tool holder, and the vice for positioning the workpiece. In this figure, workpiece is positioned in the vice with micro-slot cut on it. Numerous factors can be considered and controlled in WEDM. In this work, different values of current, pulse on time, and pulse off time were utilized. After initial trail runs, other machining factors were maintained as constant for all the experiments. The sliding of tables in X-Y coordinates is controlled by DC servo motors. The brass wire electrode moves continuously over the workpiece. The electrode is normally held in between a pair of wire guides. The pulse generator in the WEDM setup creates a series of electrical pulses. This induces a thermal flux between the brass wire and Ti-B 4 C nanocomposite work sample. The induced thermal flux augments the electro erosion on the work material. Machined debris were continuously removed from the work surface by high   Brass wire tool with a size of 250μm was employed in WEDM process. Micro-slots were cut on the Ti-B 4 C nanocomposites using WEDM process by varying the parameters for every operation. Kerf values were measured by 'METZER-M' tool maker's microscope which has the least count of 0.001 mm. The kerf values measured are generally higher than that of wire diameter used during experimentation. The surface roughness (Ra) was measured by SURFTEST SJ-210 and the least count for the tester is 0.001 mm. Response surface methodology (RSM) was used to design the experimental runs for cutting micro-slots in Ti-B 4 Cnanocomposites by WEDM process.

Design of experiments
RSM intended for figuring out the relationship among both response variables and input factors [13]. The purpose of RSM design of experiments is to compute the least possible numbers of experiments for obtaining the better output measures [14]. The chosen input variables for RSM method are presented in table 2. By using these input variables, 20 combinations of factors/experiments were formulated.

XRD analysis
The XRD graph of samples (Pure Ti and Ti-B 4 C) was obtained using analytical x-ray powder diffractometer in step mode. The selected diffraction analysis was performed with the following settings: diffraction angle 2θ, step size 0.05, measurement time 5s, temperature 14.3°C, and standard as Titanium powder. Figure 4(a) represents the XRD peaks for pure Ti. Based on the peak orientations, the observed material is Ti. In figure 4(b), in addition to Ti reflections, there are other minor peaks showing B 4 C nanoparticles. The martensite hexagonal h phase is illustrated by broad peaks that are slightly lower in position than the hexagonal h phases. The presence of B 4 C phase is confirmed by the shorter peaks in this figure at 2θ = 34°, 58°, and 71°. The formation of TiC ( figure 4(b)) is attributed to the reaction between Ti from the substrate and the carbon from the dielectric fluid. Carbide formation may be attributed due to the phenomenon of dielectric cracking during electric discharge. The dielectric fluid being a hydrocarbon, carbon atoms are generated during pyrolysis of the same. These carbon  atoms react with Titanium and thus a carbide layer forms on the machined surface. The formation this carbide phase is highly unfavorable to EDM cutting operation.

SEM Characterizations
The SEM images of Ti-B 4 C nanocomposites were used to investigate the surface morphology of the cut slots. WEDMed surfaces are largely entails of many layers like recast layer, heat affected zone (HAZ) and unaltered base material. Recast layer is the uppermost layer which directly affect the surface properties of the WEDMed components. Figures 5(a), (b) depicts the morphology of WEDMed surface with the following parameter settings (Ton = 125 μs, Toff = 51 μs and current = 180A). Resolidified layer with incomplete distribution of melted material causes visible cracks, in addition to the thermally induced cracks ( figure 5(a)). Moreover, many cracks and patches of local melting were observed in magnified image of melted surface as shown in figure 5(b). At the maximum range of input process parameters, there is a presence of many craters and black patches on the WEDMed surfaces. The presence of non-conductive and ceramic particles projecting from the cut area caused the creation of craters on the surface. As a result of cutting process, black patches are formed over the machined surface during machining. Due to high temperature, the material gets melted this is known as localized melting which is indicated in figure 5(a). During the spark produced in machining, materials get evaporated and formed like layers. This is called recast layers and formed like steps ( figure 5(b)). The recast layer is the uppermost layer of the machined surface. The recast layer is formed due to the re-deposition of molten Ti particles on the machined surfaces of nanocomposites. As the recast layer have direct contact with dielectric medium, dissipation of heat is quicker and consistent. As evident from the figure 5(b), the recast layer is non-uniform. The zones of recast layer and HAZ tends to have higher amount heat accumulation due to the low thermal conductivity of Ti.

Experimental results
MRR was calculated by using the data given in equation (1). Table 3 shows the RSM design layout and experimental results. The calculated values for the responses like MRR, Ra and kerf width is indicated in this table. High MRR and minimum Ra/kerf width is always considered as the good machining conditions. Ra and kerf are measured at three different locations of each micro-slot and then the mean value is considered for calculations.

MRR
The amount of material extracted during machining time taken 1 ( )

Impact of machining variables on output responses
Impact of current, pulse ON time and pulse OFF time on MRR, Ra and kerf quality is represented in figures 6(a)-(c). Increasing the current from 180A to 200A tends to augment the MRR, Ra and kerf width. Increasing the current enhances the spark activation phenomenon in the discharge zone between the workpiece (Ti-B 4 C) and tool (brass), hence more material is removed. Material is removed from the Ti-B 4 C surface by melting and vaporization created by the high energy electrical discharge. Enhancement incurrent and pulse ON time triggers a rise in energy surface flux density and subsequently surface of the workpiece gets heated. At a peak current of 200A used in this work, the maximum MRR value was obtained. Augmenting the pulse ON time, rises the kerf width value, which is indicated in figure 6(b). At the same time, pulse ON time has mixed impacts on MRR and Ra. In WEDM process, worn brass tool/Ti-B 4 C elements amalgamates with the fragmented products of dielectric medium radically modify the material removal mechanism involving to the various levels of removal process, such as breakdown, discharge, and erosion of elements.
The curve 6(a-b) confirms that as the current and pulse ON time value rises, the Ra value also rises. Maximum Ra is seen at the highest peak current of 200A. At 200A, when compared to other two current values employed in this study more energy is supplied to the micro-slot cut zone resulting in deeper and larger craters on the Ti-B 4 C surface, which worsens the surface quality. As indicated in figure 6(c) that the Ra value diminishes (better Ra) when there is rise in pulse OFF time from 50 μs to 60 μs. Since the time gap is higher between ON and OFF, debris can be easily flushed away during this time from the area between the Ti-B 4 C workpiece and brass tool by the pressurized dielectric medium.

Surface plots of MRR
MRR values obtained by machining Ti-B 4 C nanocomposites using WEDM is indicated in the figure 7. The MRR is the primary response for any process which uses non-conventional machining is maximum in the region where current and pulse ON time are at peak value. Maximum MRR occurs at higher current (>199A) and pulse In fact, when the voltage gap widens, more electrically charged particles enter the cutting zone of Ti-B 4 C and can travel farther before they breach the functioning gap. These highly charged particles augment the productivity rate.

Surface plots of surface roughness
The work piece's smoothness is determined by the surface roughness. Ra value declines as pulse off time rises. Highly conducting electrode materials like brass produces more rough surfaces [39]. The surface plot in figure 8 illustrates the impact of various factors on Ra. It is noted from this plot that minimum roughness value occurs at the values of pulse OFF time (>52 μs) while current (<190A) and pulse ON time (<126 μs) remains low.

Surface plots for kerf
The kerf value should be as low as possible to attain the near net shape of the micro-slots cut on the surfaces of Ti-B 4 C nanocomposites. The effect of various factors on kerf is revealed in the figure 9. The kerf value decreases with decrease of current and pulse ON time. Minimum kerf appears at the lower values of current (<185A) and pulse ON time (<125.5 μs). Pulse ON time is the most predominant factor which affects the quality of kerf. Augmentation of pulse ON time leading to rise the amount of energy concentrated on the work surface. When the electrical discharge energy exceeds above certain threshold value, the kerf width increases immediately. The quantity of vapor produced by the materials in the electrodes rises with high discharge energy. As a result, the pyrolysis process intensifies which changes the characteristics of the dielectric liquid and potentially lowers processing productivity. This increases the electrode wear and the particle separated from the electrodes and found in the dielectric liquid. Table 4 illustrates the ANOVA for MRR. This table corroborates that all the factors considered for micro-slot cutting on Ti-B 4 C nanocomposites are substantial model terms, indicated by the P-value less than 0.05. Current is the most prevalent factor which affects the MRR with contributing factor equal to 95.77%. Pulse on and pulse

ANOVA for Kerf
The ANOVA table 6 highlights the contribution of each factor such for kerf width. Kerf is calculated from the quadratic equation (4). From table 6 it was concluded that the current (84.49%) is the most influential factor which affects the kerf width. In EDM process, the discharge energy is directly proportional to the current. Hence, with increase in applied current value, the discharge energy also increases. The higher values of current during machining helps in the effective melting of work piece, which leads to increased kerf width. The table 6 also validates that pulse on time is another major factor for kerf quality.

Optimal global solution
The response optimization was done to get the optimal parameters for achieving better responses. The arrived global solution is shown in figure 10 and table 7 gives the maximum MRR with minimum roughness while machining. These parameters were arrived by maintaining minimum kerf/roughness and maximum MRR for a typical process application. The table 7 reveals that the multi response optimization is of greater importance and weightage was assigned for getting higher desirability value. Figure 10 indicates the optimal setting values for confirmatory test and the higher composite desirability (0.89714) was chosen. The optimal setting of process parameters identified from the desirability analysis are Current = 195.556 A, T on = 125.242 μs and T off = 51.0606 μs.
The response optimization was performed to get the optimal factors for achieving better responses. The arrived global solution is shown in figure 10 and table 7 gives the maximum MRR with minimum roughness while machining. These parameters were arrived by maintaining minimum kerf/roughness and maximum MRR for a typical process application. The table 7 reveals that the multi response optimization is of greater importance and weightage was assigned for getting higher desirability value. Figure 10 indicates the optimal setting values for confirmatory test and the higher composite desirability (0.89714) was chosen. The ideal setting of machining factors identified from the desirability analysis are Current = 195.556 A, T on = 125.242 μs and T off = 51.0606 μs.

Verification test
Desirability analysis is intended to remodel the multi response optimization into distinctive response optimization. The verification test was done to substantiate the optimum input values for micro-slot cutting.
The input values with ideal conditions are specified in the table 8 delivers the good responses. The verification test was performed to check predicted values with actual values. Table 8 and figure 11 demonstrates the comparison of predicted values with the actual values. The calculated value of deviation was well below 5%, so the confirmatory test parameters (Current = 195.556 A, T on = 125.242 μs and T off = 51.0606 μs) were adequate for giving better machining rate in the process industries.

Conclusions
In this work, Ti/B 4 C nanocomposites were prepared using powder metallurgy route and then micro-slots were cut on the surfaces of Ti/B 4 C nanocomposites by WEDM. The following decisions were arrived.
✓ XRD peaks validates the incorporation of B 4 C nanoparticles in Ti matrix. Uniform dispersion of B 4 C nanoparticles in Ti matrix were validated by SEM images.
✓ From the desirability analysis, the ideal WEDM input settings for Ti-B4C nanocomposite were found to be current = 195.556 A, T on = 125.242 μs, and T off = 51.0606 μs. The optimized experimental values are in close agreement with the desirability values.
✓ The desirability results validates that that the existing mathematical model is sufficient for obtaining good cut quality on Ti-B 4 C nanocomposites. The essential recommendations and database for cutting titanium nanocomposite can be obtained by these findings in accordance with the quadratic model. The concerned industries will benefit by using this information to satisfy the demands for micro cutting.
✓ The confirmation test validates that the test values and optimized experiment values are in close agreement. The predicted and actual responses comparison showed that the error percentage is within the acceptable limit. The maximum error percentage is less than 3% for all the responses.

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