Keywords

1 Introduction

Technology is in the path of exponential growth with new and more innovations happening in various fields all around the world. This advancement in technology has to be assisted with modern capabilities for the manufacturing of the evolved products. There lies the significance of manufacturing technology in the development and progress of the world. In this scenario where competition is very tight, manufacturing a component is just not enough, but it has to be made within the strict industrial specifications and at a higher production rate. Therefore, the terms efficiency and productivity assume higher importance compared to what was present a few decades ago. With the advent of new and improved materials, it has become more difficult to machine them to the desired shapes and size using the conventional manufacturing techniques. Electrical discharge machining (EDM) is one of the technologies widely used for machining of difficult to machine materials in the present-day industry. Two Russian scientists, Lazarenko and Lazarenko, developed EDM process in 1943 [1]. The major applications of EDM technology are in the field of tool-and-die making and in mould manufacturing for machining of tool steels and heat-treated steels. This technology is also used for processing advanced materials such as metal matrix composites, super-alloys and ceramic materials, for variety of applications. In EDM, electrode and workpiece do not make direct contact with each other during the machining process. This eliminates mechanical stresses, vibration and chatter that occur during conventional machining process [2]. The electrode is moved progressively towards the workpiece, till the gap is very small for the applied voltage, to ionise the dielectric fluid and help in developing a sparking phenomenon [3]. Any material that conducts electricity can be machined with EDM irrespective of their hardness [4].

2 Literature Review

In the study conducted by Yilmaz et al. [5], a comparative investigation of fast hole drilling of aerospace alloys using EDM was performed in order to explore the influence of electrode type and material. Pham et al. [6] studied the electrode wear behaviour of tube and rod electrodes, which are the two main kinds of electrodes for EDM drilling. Yan and Wang [7] investigated the machining characteristics of Al2O3/6061Al composite using rotary electrical discharge machining with a copper tube electrode. Different flushing methods were studied and were found that flushing pressure and electrode rotation speed were having minor effects on MRR, TWR and Ra. Lonardo and Bruzzone [8] investigated the machining of Cr, Mo, V steel and observed the importance of electrode material, injection flushing and geometry of cutting on material removal rate, electrode wear and surface quality. Influence of dielectrics on the material removal rate was studied by Zhang et al. [9]. In the experiments conducted by Wong et al. [10], the influence of flushing on the efficiency and stability of machining conditions in EDM has been extensively investigated, including the effects of flushing configuration on the wear of the tool and the profile of the workpiece. Later on, the process of micro-EDM emerged and few people like Zahiruddin [11] studied their energy and removal efficiencies. Other significant experimental studies were also conducted to characterise material removal [12] and crater formation [13] during the EDM process. Dielectric medium and its properties also affect the performance of EDM. Zou et al. [14] studied the effect of using nitrogen plasma as the dielectric medium while machining brass H62 using tungsten electrode. They pointed out a significant improvement in discharge distance, machining efficiency and surface quality. Apart from the dielectric medium tool, electrode material also has a predominant effect in EDM performance. Recently, Koyano et al. [15] investigated the effect of tool electrodes with high electrical resistivity (single-crystal silicon) in micro-electrical discharge machining of micro-rods. Their experimental results showed a reduction in the average diameter of discharge craters by 0.4 μm. Similar work has been carried out by Pilligrin et al. to find the effect of tool electrode materials in terms of tool wear rate, taper angle, overcut and surface roughness. From their experiments, it was observed that at higher discharge energy, MRR and TWR are more. The surface finish was observed to be good at low discharge energy. It is evident that EDM is one of the fast-evolving machining processes for machining of difficult to machine materials. The flushing efficiency plays a major role in the material removal rate as well as in the surface quality. As there are different flushing methods possible for the machining process, an attempt has been made in the present work to study the effectiveness of injection flushing technique during drilling hole using different electrodes in EDM.

3 Thermal Modelling of EDM Process

Finite element modelling of the EDM process was done with the help of COMSOL Multiphysics software version 4.3b. The crater shape which is developed as a result of the application of heat on to the surface of the workpiece was used to estimate the volume removed in a single crater. The following assumptions were made during the modelling. Workpiece material is assumed to be isotropic in nature and the model developed is axisymmetric about Z-axis. Radius of the heat source was kept constant with time. The thermal properties of the workpiece material are independent of the temperature induced on to the surface. The shape of the elements remains constant during thermal expansion which occurs due to the heating of the workpiece. The workpiece is not affected by any kind of stress that may occur during the EDM process. Conduction and convection are the only modes of heat transfer during the machining process. Any effect of radiation is neglected. Thermal analysis is transient, and the heat source has Gaussian distribution of heat flux incident on the workpiece surface. The heat transfer in a solid module of the COMSOL Multiphysics software is used to model the EDM process. The time-dependent mode of the solution module is employed to solve the problem. The governing equation used for solving the model is given in Eq. (6.1).

$$\rho C_{p} \frac{\delta T}{\delta t} + \rho C_{p} u \cdot \nabla T = \nabla \cdot \left( {k\nabla T} \right) + Q$$
(6.1)

The heat applied has a Gaussian distribution which is dependent on the distance of the spark from the axis of the workpiece, unlike the uniform distribution of heat flux that was applied in the initial works done by several researchers. The relation for Gaussian distributed heat flux is given in Eq. (6.2)

$$Q^{{\prime \prime }} \left( r \right) = \frac{{4.57F_{c} {\text{VI}}}}{{\pi R^{2} }}{\text{e}}^{{ - 4.5\frac{{r^{2} }}{{R^{2} }}}}$$
(6.2)

At the start of the EDM process (t = 0), the workpiece is immersed in the electrolyte, and the temperature of the whole domain is assumed to be at room temperature (T0). The heat flux is applied on the top surface of the workpiece. The left-side vertical surface is taken as the axis of symmetry along which the 2D square can be rotated so that a cylindrical 3D object can be obtained. The other two surfaces, the bottom surface and the vertical surface on the right, are insulated thermally.

4 Experimental Details

The workpiece material used for conducting experiments in EDM is EN31 alloy steel. Major constituents of this alloy are iron, carbon and chromium. The mechanical properties of EN31 alloy steel are given in Table 6.1. Tool electrodes of two materials were used to conduct the experimental investigations. Copper and brass electrodes were used to machine the workpiece. Copper electrodes are known for their superior thermal conductivity. Relevant properties of copper and brass electrodes are shown in Table 6.2. The machine used to carry out the experimental investigations was Electronica Smart ZNC as shown in Fig. 6.1. The electrodes were having a geometry of hollow cylinders which was used to facilitate the flow of dielectric oil through the centre of the electrodes in order to implement the injection flushing configuration in the experiments. The electrodes were bought in the form of hollow geometry with a length of 300 mm and a diameter of 4 mm. Then, they were cut into small pieces using a hacksaw. After that, they were faced to get a very good finish and drilled to make the central hollow portion a perfect circle. After facing, the end surfaces were further finished by polishing using different grades of emery paper. Figure 6.2 shows one of the electrodes under preparation.

Table 6.1 Mechanical properties of EN31 alloy steel
Full size table
Table 6.2 Properties of copper and brass electrodes
Full size table
Fig. 6.1
figure 1

Electronica Smart ZNC EDM machine

Full size image
Fig. 6.2
figure 2

One of the tool electrodes under preparation

Full size image

Electronica Smart ZNC has six valves to provide dielectric flushing into the system. Two of these valves are connected to flow dividers which have nozzles at the ends to provide jet flushing for clearing away the debris formed during the EDM process. Therefore, special tubes and connectors are attached to the dielectric supply system to provide an alternate route to facilitate the injection flushing configuration in the system, as shown in Fig. 6.3. Out of the major EDM parameters, the peak current and the pulse-on-time are most influential on the output of the machining process. So these two parameters were selected for preparing the design of experiments and conducting the experiments. Table 6.3 represents the experimental conditions in which the machining is carried out.

Fig. 6.3
figure 3

Special tube connected to one of the six valves for injection flushing-type dielectric supply in the Electronica Smart ZNC EDM machine

Full size image
Table 6.3 Experimental conditions for EDM machining
Full size table

5 Results and Discussion

5.1 Thermal Modelling of EDM

After running the simulation for the stipulated time, the required results were obtained under the result’s tab in the model tree. The gap voltage is kept constant for the entire simulation. The value of heat flux is calculated by substituting the values of current, voltage and spark radius into the Gaussian heat distribution function. The spark radius is calculated with the help of an empirical formula which is dependent on the values of peak current and the pulse-on-time, as given in Eq. (6.3).

$$R = 2.04 \times 10^{ - 3} I^{0.43} T_{\text{on}}^{0.44}$$
(6.3)

Figure 6.4 shows the temperature distribution in the volume of the workpiece after the model is solved. It can be seen that the maximum temperature has formed at the centre point where the Gaussian heat distribution value is the maximum. Figure 6.5 represents the isothermal contour after the first spark and the subsequent spark that has been given to the workpiece. It denotes the line along which the temperature formed during the application of heat is at a value of 2770 °C. Figure 6.6 shows the variation of temperature with respect to the r-coordinate under different intervals of time for the first spark and for subsequent spark. Here, it is found that the maximum temperature for subsequent spark no longer occurs at the centre because the spark has occurred at a different point where the distance between the workpiece and the tool is minimum.

Fig. 6.4
figure 4

Temperature distribution in Kelvin for 9A peak current and 50 µs pulse-on-time

Full size image
Fig. 6.5
figure 5

Isothermal contour of melting point after the application of the first spark in the centre of the workpiece (left), and after the application of the subsequent spark

Full size image
Fig. 6.6
figure 6

Variation of temperature with respect to the r-coordinate under different time values for first spark (left) and for subsequent spark (right)

Full size image

5.2 Machining with Hollow Copper and Brass Electrodes

Sixteen experiments each were conducted with hollow copper and brass electrodes by varying peak current and pulse-on-time in four steps. Graphs are plotted to find the effect of the input parameters on the outputs. The effect of peak current on MRR for different pulse-on-times for copper and brass electrodes is shown in Fig. 6.7. The MRR seems to be increasing steadily with an increase in peak current. This is due to the fact that the energy that has been discharged on to the workpiece increases with the current, as the current is directly proportional to discharge energy. This energy transfers more heat on to the workpiece which in turn melts away more material resulting in a higher material removal rate. This is in agreement with the Gaussian heat flux equation denoted earlier in the paper and with the findings of Govindan et al. [12]. It can also be observed that the values of MRR are comparatively low for brass electrodes with respect to copper electrodes.

Fig. 6.7
figure 7

Effect of peak current on material removal rate at different pulse-on-time values for copper electrodes (left) and brass electrodes (right)

Full size image

Tool wear rate also shows the same trend as MRR, having an increasing tendency with an increase in peak current. This is because some part of the energy formed in the plasma channel is utilised to melt the tool electrode. As it is already mentioned that the discharged energy is proportional to the peak current applied; the energy transferred to the tool, though not as high as that transferred to the workpiece, also keeps increasing. Thus, a higher tool wear rate is obtained for higher peak current values. The effect of peak current on tool wear rate for copper and brass electrodes at different pulse-on-time values is shown in Fig. 6.8. It can be observed that the magnitude of tool wear is also low for brass electrode compared to the copper electrode, but is not appreciably low.

Fig. 6.8
figure 8

Effect of peak current on tool wear rate at different pulse-on-time values for copper electrodes (left) and brass electrodes (right)

Full size image

Figure 6.9 shows the effect of peak current on surface roughness for both electrodes at different pulse-on-time values. Surface roughness tends to increase with an increase in peak current value. This trend occurs due to the bigger crater formed on the workpiece surface at higher currents. In simple mathematical terms, the crater formed can be modelled as hemispherical voids formed on the surface due to the bombardment of ions or electrons with high energy. Considering higher energy due to the high peak current, the diameter of the hemispherical crater formed is higher. This can be a direct indicator of a higher surface roughness value at higher peak current during machining.

Fig. 6.9
figure 9

Effect of peak current on Ra at different pulse-on-time values for copper electrodes (left) and brass electrodes (right)

Full size image

The effect of pulse-on-time on material removal rate for copper and brass electrodes at different peak current values is shown in Fig. 6.10. With respect to pulse-on-time, MRR initially shows an increasing trend for copper electrodes, but gradually tends to decrease when very high pulse-on-time is given. This is due to the fact that the plasma channel radius which is dynamic in nature tends to increase with time, and thus, the energy which has been given to the workpiece tends to get more and more distributed along the workpiece and thus reduces the concentration of power on the required diameter. This also points to the fact that keeping on increasing discharge energy does not guarantee a higher removal rate during the machining operation. It is also observed that the material removal rate has not much been affected by the change in pulse-on-time for brass electrodes as compared to copper electrodes.

Fig. 6.10
figure 10

Effect of pulse-on-time on MRR at different peak current values for copper electrodes (left) and brass electrodes (right)

Full size image

Tool wear rate increases initially with an increase in the pulse-on-time but later tends to decrease slightly for copper electrodes. The reason can be attributed to the same factors that cause a decreasing trend in MRR due to higher pulse-on-time values. For brass electrodes, tool wear rate also seems to remain almost constant for a change in pulse-on-time values in spite of the appreciable increase in tool wear rate for higher current values in the order of 18 A. The effect of pulse-on-time on tool wear rate for copper and brass electrodes at different peak current values is given in Fig. 6.11.

Fig. 6.11
figure 11

Effect of pulse-on-time on tool wear rate at different peak current values for copper electrodes (left) and brass electrodes (right)

Full size image

5.3 Comparison Between Copper and Brass Electrodes

Figure 6.12 compares the values of material removal rate and tool wear rate obtained using copper and brass electrodes for the machining of EN31 alloy steel. It is clear that copper has better material removal rate than brass. Even though the tool wear rate of copper is higher than that of brass, the difference is not appreciable as what we see in the material removal rate graph. This is because copper has low electrical resistivity, high melting point and high thermal conductivity compared to brass. So it can be concluded that copper is a better material to use for machining EN31 alloy steel to obtain better material removal rate with the least tool wear rate.

Fig. 6.12
figure 12

Comparison between MRR (left) and TWR (right) for copper and brass electrodes

Full size image

5.4 Comparison of Thermal Model with Experimental Results

The simulation for electrical discharge machining is done for a gap voltage of 60 V and a pulse-on-time of 50 µs with single spark and multi-spark, and the results are compared with experimental data. The comparison is shown in Fig. 6.13. It can be observed that the simulation results of MRR are higher than the experimental values. This can be attributed to the fact that the simulation is nearer to an ideal scenario, whereas the experimental material removal rate is affected by various external factors. Also, multi-spark simulation is closer to the experimental data than the single spark simulation because, in multi-spark simulation, the overlap of craters is also taken care of during the modelling of the problem.

Fig. 6.13
figure 13

Comparison between experimental values, single spark simulation and multi-spark simulation

Full size image

6 Conclusions and Scope for Future Work

Finite element modelling of the EDM process was done with the help of COMSOL Multiphysics heat conduction in solid module to calculate the temperature distribution and the volume of the craters formed during sparking, thus calculating the MRR. Experimental investigations on the machining of EN31 steel alloy were done by varying the input parameters, and the results were compared with respect to the effect on material removal rate, tool wear rate and surface roughness in drilling annular holes on the workpiece. Material removal rate by using copper electrodes was found to be around 4.5 times than that of brass in the middle of the range of the experiments. This is because copper has low electrical resistivity, high melting point and high thermal conductivity compared to brass. Tool wear rate of copper is also high, and it is around 3.5 times that of brass electrodes. Although the tool wear rate is higher for copper, this can be compensated by the advantage of a high material removal rate. No appreciable difference was observed in the surface roughness measurements when the results from using the two different electrodes were compared. Therefore, copper is a better electrode than brass for machining shallow holes and fins on EN31 alloy steel with the help of injection flushing. The results obtained from the model were compared with the experimental data and were found that they are complementing each other. The variation between the experimental data and the simulation results may be due to the fact that the experimental data was obtained from real machining conditions and the simulation data had few assumptions taken.

An array of shallow holes are drilled on surfaces for weight reduction, and fabrication of micro-fins for better thermal management. Electrical discharge machining  carriedout with hollow electrodes and injection flushing is a promising solution for making such features. As there are no much literature available on the use of such electrodes and flushing techniques, it was important to perform this experiment to identify the best tool electrode material and processing condition for obtaining the best machining results. In future, the effect of different flushing pressures can be studied to find the optimum pressure of the dielectric. The modelling can be extended for performing large number of sparks in a single simulation using the COMSOL with MATLAB interface. If the FEA simulation is repeated several times, there will be no pattern observed, as the next spark in simulation strikes at the location that is closest to the tool electrode, and once the simulation completes few initial runs, the spark will occur at random locations on the workpiece surface for the spark to act. It may be noted that EDM process is a stochastic process.