Influence of flushing holes on the machining performance of blasting erosion arc machining

Abstract

Blasting erosion arc machining is a novel electrical erosion process depending on the hydrodynamic arc-breaking mechanism to achieve a reliable high-efficiency machining. In blasting erosion arc machining, the high-velocity fluid field in the discharging gap is the precondition of the mechanism to control arc plasma to efficiently remove workpiece material. Therefore, this study mainly investigates the influence of flushing holes on the fluid field distribution directly and on the machining performance indirectly. Three multi-hole solid electrodes with different types of flushing holes are designed out according to the distributing principle. The influence of their flushing holes on the fluid field is conducted by a comparison fluid simulation which demonstrates that the electrode with flushing-hole diameters decreasing gradually from the inner to the outer in the radial direction attains the best flushing velocity distribution on the workpiece surface. Furthermore, the influence of their flushing holes on the blasting erosion arc machining performance is investigated by a comparison machining experiment in order to verify the comparison results of fluid field simulation. The experimental results illustrate that these electrodes have very different machining performance when machining nickel-based high-temperature alloy GH4169 (similar to Inconel 718) under the conditions of same discharge peak current and flushing inlet pressure. The electrode with the best flushing velocity distribution rather than with the highest velocity at a particular point achieves the best machining performance of the highest material removal rate, the least relative tool wear ratio and the least surface roughness (Ra), indicating an optimized design of flushing holes in the multi-hole solid electrode.

Introduction

In electrical discharge machining (EDM), flushing is an important factor that greatly influences the machining performance. ¹ Munz et al. ² found that strengthening flushing velocity in the gap could achieve a high material removal rate (MRR) with an appropriate electrical energy. Therefore, many approaches had been proposed to improve the flushing conditions in the gap to improve the EDM performance,^3,4 such as EDM milling with rotary cylindrical tool, ⁵ jet flushing ⁶ and high-speed tool jump. ⁷ Moreover, it seemed that forming holes in tool electrode was a more useful method to enable fluid flush into the discharge gap, such as a mono-hole or a multi-hole electrode. Using a mono-hole electrode, Han et al. ⁸ proposed a novel high-speed EDM milling method with moving electric arcs. However, a multi-hole electrode was increasingly employed to improve the flushing effect, such as a multi-channel ⁹ or a multi-function electrode. ¹⁰

Shibayama and Kunieda ¹¹ bonded two copper plates together with micro grooves on the interface to form a multi-hole EDM electrode. By jetting the dielectric fluid through the holes into the gap, it significantly shortened the processing time and improved the processing accuracy compared with a conventional solid tool electrode. Zhao et al. ¹² proposed a bundled electrode that bundled a number of tubular electrodes into a die-sinking electrode according to the ultimate profile of the workpiece, which was also considered as a form of multi-hole electrode. Their experiments showed that the distribution of flushing velocity had an important influence on the machining performance. Gu et al. ¹³ further pointed out that a multi-hole bundled electrode could attain a well-distributed flushing velocity in the narrow gap compared with a mono-hole electrode by the comparison simulation and experiment. However, due to limited discharge energy in EDM, the MRR of existing multi-hole electrodes is still restricted for bulk material removal. Recently, arc discharge machining with a multi-hole electrode rather than spark discharge machining with a mono-hole or a solid electrode is suggested to remove the bulk material and attain a higher MRR.

Some research works had found that flushing-assisted arc discharge could principally improve the machining efficiency. Meshcheriakov et al. ¹⁴ found that the hydrodynamic pressure of working fluid in the gap influenced the physical features of arc discharging and then proposed an arc dimensional machining (ADM) method, initiating a possibility of using flushing to control arc discharging. Suzhou Electrical Machining Research Institute in China achieved an MRR of up to 3000 m³/min when removing high-temperature alloy Inconel 718 with a mono-hole electrode, which was also a form of arc discharge machining controlled by flushing and mechanical motion. ¹⁵ Zhao et al. ¹⁶ proposed a blasting erosion arc machining (BEAM) with a multi-hole electrode based on the hydrodynamic arc-breaking mechanism. When machining nickel-based high-temperature alloy, BEAM achieved a MRR of 14,000 mm³/min (current: 500 A), and the minimum tool electrode wear ratio (TWR) was less than 1%, and the specify energy removal efficiency was 28 mm³/A min. Wang et al. ¹⁷ achieved a high MRR of 15,062 mm³/min (current: 920 A) using compound arc and spark power with high current density and a mono-hole electrode to machine Inconel 718. Moreover, using the same arc machining approach to the difficult-to-cut materials, Wang et al. ¹⁸ achieved a maximum MRR of 21,494 mm³/min at a TWR of 1.7% while machining titanium alloy (current: 700 A) and a maximum MRR of 12,688 mm³/min at a TWR of 2.3% while machining mold steel (current: 700 A). ¹⁹ Recently, Zhang et al. ²⁰ found that the processing factors including discharge medium, tool polarity, tool material, voltage and rotational speed have obvious effects on MRR and TWR in electro-arc machining. Zhao et al. ²¹ illustrated that it was possible to machine with high efficiency and a better control of the profile and surface quality of the workpiece by combining negative and positive tool polarities of BEAM processes together. Xu et al. ²² also revealed that by optimizing the combination of the negative and the positive BEAM processes, it was possible to attain favorable machining performances of high MRR and finer Ra while machining the nickel-based high-temperature alloy GH4169 (similar to Inconel 718) with a bundled electrode or a multi-hole solid electrode. By utilizing the polarity effects, Chen et al. ²³ had machined out a three-dimensional (3D) cavity of SiC-Al composites material by a multi-hole solid electrode with milling method, which showed that the machining allowance for the subsequent semi-finishing processes such as cutting could be further reduced. It is concluded that a high MRR is more likely to be attained when setting flushing holes in the tool electrode for all the methods above. However, the effects of flushing holes on the fluid field distribution as well as the machining performance have not been systematically reported.

In BEAM, the flushing is performed based on the flushing holes of the multi-hole electrode. However, as a bundled electrode, the distribution of its flushing holes is so simple that it will cause uneven fluid field distribution in the gap. The simulation of flushing velocity distribution is carried out in order to explain the disadvantageous conditions, as shown in Figure 1. The simulation results present that flushing velocity at the peripheral of bundled electrode is very high (maximum 55.4 m/s) when the flushing inlet pressure is 1.2 MPa. But it decreases to about 3 m/s on the central area where the ability of arc breaking is insufficient to remove the molten material, resulting in a nonuniform distribution of removal material, irregular discharge craters and rugged workpiece surface in the flushing direction according to the machining experiment. In that case, it will result in an unfavorable Ra and machining allowance for BEAM. Therefore, it is necessary to optimize the distribution of flushing holes in order to improve the flushing effects first and the machining performance ultimately.

Figure 1.

BEAM with a bundled electrode: (a) geometric model and (b) fluid field distribution in the gap.

In this study, three multi-hole solid electrodes with different types of flushing-hole distributions are designed out according to the principle of flushing-hole distribution in section “Designing of flushing holes for a multi-hole solid electrode.” In section “Comparison simulation and discussions of the fluid field distribution,” the comparison fluid field simulations are carried out to analyze the flushing performance of their flushing-hole distribution. The comparison experiments corresponding to the fluid field simulations are introduced in section “Setup and conditions of the comparison experiment,” and the influence of flushing holes on the BEAM performance is comparatively discussed in section “Results and discussions of the comparison experiment.”

Designing of flushing holes for a multi-hole solid electrode

Selecting a multi-hole solid electrode as the tool electrode

According to preliminary experimental results, the key to advance the BEAM performance is to improve the inner flushing in the gap which is related to the flushing-hole distribution directly. In BEAM, the main characteristic of the electrode is its multi-hole structure. Therefore, designing multiple and opening distribution of flushing holes could form different types of multi-hole electrodes to machine different 3D cavities, such as bundled electrode, laminated electrode or multi-hole solid electrode. ¹⁶ In this study, a comparison analysis between a bundled electrode and a multi-hole solid electrode is conducted in order to introduce their respective advantages and disadvantages, as shown in Figure 2. It is concluded that BEAM with a bundled electrode is able to machine an approximate 3D workpiece in sinking mode simply and rapidly. However, it leaves much machining allowance for the ideally machined surface because of its irregular endface. In addition, it is also difficult to bundle tubular electrodes with different diameters together to form a sinking electrode, resulting in inflexibility of machining different workpiece profiles. In contrast, BEAM with a multi-hole solid electrode leaves less machining allowance to decrease the semi-finish machining time. In addition, the flushing-hole diameters as well as the electrode dimension can be adjusted conveniently according to the workpiece profile, which makes it be suitable to machine more irregular 3D cavities. This study aims to provide a basic method of arranging flushing holes to optimize the flushing-hole distribution and then improve BEAM performance. Therefore, the investigations mainly focus on the flushing-hole distribution on a circular endface of the electrodes because of its perfect symmetry in geometry.

Figure 2.

BEAM with (a) bundled and (b) multi-hole solid electrodes.

Principle of flushing-hole distribution

Distributing flushing holes in electrode endface should follow some universal rules such as well-distributed holes to make uniform fluid field distribution and standardized flushing-hole diameters to be machined rapidly. In order to quantificationally express the principle of flushing-hole distribution, this study introduces a dependent variable effective-area ratio (EAR) for all the multi-hole electrodes, which has a dual property of flushing effect and electrode effective area. ¹² An EAR (ψ) is defined as the ratio of the effective area (S_e) occupied by a solid material to the nominal contour area (S) of the whole electrode, as presented in equation (1). It reflects a certain geometrical configuration of a multi-hole electrode such as the endface area, the number and diameters of flushing holes. For a certain multi-hole electrode, its EAR is constant. Therefore, a smaller EAR means a relatively larger flushing-hole area and a narrower space between flushing holes. The results of the preliminary experiments show that when the EAR decreases sharply, the flushing velocity in the gap increases rapidly. Therefore, the machining system can employ a higher peak current and a higher feedrate, which will aggravate the electrode wear or even damage it because of the nonuniform discharging in the gap, resulting in a higher TWR. In contrast, an excessively high EAR will bright out a low flushing effect. The machining system has to employ a lower current peak and a lower feedrate, which will reduce the capability of removing the workpiece material, resulting in a lower MRR. In summary, a moderate EAR value should be first considered to design and optimize the flushing holes. The preliminary experiments of the multi-hole solid electrode show that the optimized value range of the EAR is from 80% to 85%, which corresponds to the bundled electrode in EDM.^24,25 The EAR in circular endface can be calculated as

$$\begin{array}{c}\psi =\frac{S_e}{S}=1-\frac{\sum _{i=1}^n{S}_i}{S}=1-\frac{\sum _{i=1}^n{d}_i^2}{D^2}\hfill \\ \psi \in [80\%,85\%]\hfill \end{array}$$

(1)

where S_i is the area of the flushing hole i, n is the number of flushing holes, d_i is the diameter of the flushing hole i and D is the diameter of the circular endface.

From equation (1), it is concluded that the principle of flushing-hole distribution can be first quantified to determine the maximum hole number n, the hole diameter d_i and the endface diameter D. Similar to a certain multi-hole solid electrode, its endface diameter D is constant. Therefore, the principle can be simplified to determine the number n and the diameter d_i. However, there is a direct relation between the endface diameter D and the number n. By increasing the size D, the required number n should be increased in order to ensure enough flushing effect, relating to one of the EAR properties. But the diameter d_i is not directly influenced by the endface diameter D, which is mostly determined by its machining capability, relating to the other one of the EAR properties.

According to the structure of the bundled electrode presented in Figure 1(a), its distribution of flushing holes has well geometric symmetry and flushing effect in the circumferential direction. Similarly, the distribution of flushing holes on the circular endface of the multi-hole solid electrode follows the same rule in the circumferential direction, as shown in Figure 3. Generally, it first selects a geometry symmetrical line as the baseline L, also shown in Figure 3. A flushing hole on the baseline L can be mapped to each hole on the same ring by rotating a certain angle in the circumferential direction. Therefore, the flushing-hole diameter can be set as the same value in the circumferential direction on the same ring. Second, the number n and the diameter d_i in equation (1) can be quantified to the maximum ring number k, the hole diameter d_j and the hole number n_j on the ring j, as expressed in equation (2)

$$\psi =1-\frac{\sum _{i=1}^n{d}_i^2}{D^2}=1-\frac{d_0+\sum _{j=1}^k{n}_j{d}_j^2}{D^2}$$

(2)

where d₀ is the innermost hole diameter.

Figure 3.

Designing parameters of flushing holes.

According to the presentation in Figure 3, the principle of flushing-hole distribution gives out a mathematical expression to determine the maximum hole number n, which is simplified to determine the hole number n_j on the ring j and the maximum ring number k in equation (3). In order to attain a suitable EAR value, the flushing-hole number is set differently in the radial direction in different rings, which will increase proportionally to every increase of ring from inner to outer because of increasing perimeter. Thus, the mathematical relation between n_j and j can be given out in equation (4)

$$\begin{array}{c}n=n_0+n_1+\cdots +n_j+\cdots +n_k\hfill \\ n_0=1,k\in [1,2,3,\dots ]\hfill \end{array}$$

(3)

$$n_j=1\times 6\times j,j\in [1,2,3,\dots ,k]$$

(4)

The principle of flushing-hole distribution then gives out a mathematical expression to determine the hole diameter d_i in equation (1), which involves determining the hole diameter d_j, the spacing T_j and the maximum ring number k on the baseline L.

1. Determining the diameter d_j on the baseline L depends on its machining capability. The diameter should match with the standard diameter of drill tool in order to tooling holes economically and efficiently. However, a small hole cannot easily be drilled out because of its easy damage to the tool. Instead, a big hole will leave a large nubble on the workpiece surface during discharge machining, which obviously blocks fluid to flow into the gap.

2. Determining the hole spacing T_j between the two holes on the baseline L depends on the machining capability and flushing effect of the flushing holes. The spacing should be greater than 3 mm to avoid the damage to drilled holes because of the concentration of drill force. Relatively, it should not be greater than 5 mm to ensure well-distributed flushing velocity on the spacing region between the two holes in the radial direction according to the preliminary comparison experiments on the flushing-hole spacing. Therefore, it usually selects a constant (4 mm) to ensure uniform machinability and flushing effect. In addition, the spacing between the electrode periphery and the outermost hole can be selected as a constant (2 mm) in order to prevent the electrode edge to be broken when discharge machining.

3. Determining the maximum ring number k depends on the electrode diameter D, the diameter d_j and the spacing T_j (4 mm). The mathematical expression to determine hole diameter d_j and the maximum ring number k based on the designing parameters on the baseline L is defined as

$$\begin{array}{c}D=d_0+2(\sum _{j=1}^k{d}_j+4k+2)\hfill \\ D=2\times R\hfill \end{array}$$

(5)

where R is the radius of the circular endface as shown in Figure 3.

Finally, the principle of flushing-hole distribution should check out the flushing-hole design by solving the total number n_j in equation (4), the diameter d_j and the maximum ring number k in equation (5). The solution values are substituted in equation (1) to calculate the EAR(ψ), which should be an optimum solution to the distributing principle.

Tooling three multi-hole solid electrodes with different flushing-hole distributions

According to the principle of flushing-hole distribution above, this study designs out three multi-hole solid electrodes as shown in Figure 4(a), named electrode E1, E2 and E3, respectively. The parameters of these electrodes are listed in Table 1. It is concluded that they have the same flushing-hole number determined in equation (4). It is also concluded that the only difference among them is their different hole diameters on the baseline determined in equation (5). Compared to the diameters in E1, the diameters in E2 on the inner ring are increased in order to decrease the EAR and increase the flushing effects on the central area. Compared to the diameters in E2, the diameters in E3 on the outer ring are decreased in order to increase the EAR on the peripheral area. The diameter of the innermost hole is selected as a constant 2 mm for all the electrodes in order to ensure a moderate EAR on the innermost area. Totally, the calculated EAR from the three electrode parameters conforms to the optimum range in equation (1).

Figure 4.

Three electrodes with different flushing-hole diameters on the baseline: (a) designed and (b) tooling electrodes.

Table 1.

Designing parameters of the three electrodes.

	E1	E2	E3
Electrode diameter, D (mm)	30	30	30
Electrode axial height, H (mm)	50	50	50
Flushing-hole diameter, dj (mm)	2, 2, 2, 2	2, 3, 2, 2	2, 3, 2, 1.5
Total number, nj	1, 6, 12, 18	1, 6, 12, 18	1, 6, 12, 18
Projection area, S (mm2)	706.9	706.9	706.9
Effective-area ratio, ψ (%)	83.56	80.22	83.73

Another important stage of tooling multi-hole solid electrode is to machine out the electrode. The most suitable tool material for BEAM is graphite because of its good machinability, high melting point and low linear expansion coefficient, resulting in low tool wear. The process of machining multi-hole solid electrode consists of drilling holes and milling complex electrode profile, which is completed on a specified graphite electrode machining tool. The process time of drilling a hole in these electrodes is less than 1.5 min. Finally, the machined three electrodes are shown on Figure 4(b).

Comparison simulation and discussions of the fluid field distribution

The simulation of the flushing velocity distribution in the gap for the multi-hole solid electrodes is conducted with the commercial computational fluid dynamics (CFD) software (Fluent). The effects of flushing holes on the fluid field distribution are analyzed and compared to clarify the distributing principle.

3D geometric model

As presented in Figure 3, the distribution of flushing holes on the electrode endface is discrete symmetry in the circumferential direction. Therefore, the geometric model of fluid equipment is chosen as 3D model, as shown in Figure 5(a). The model has been simplified in order to analyze a partly systemic fluid model with a finite element method (FEM), as shown in Figure 5(b). Figure 5(b) also shows that the electrode endface in the gap is 1 mm far away from the workpiece surface, resulting in complicated fluid field distribution in the gap. During machining, the working fluid initially flows from the inlet, passes through the flushing holes into the discharging gap, then flows through the lateral gap and finally ejects into the surrounding environment from the gap near the workpiece surface.

Figure 5.

Geometric model of simulation: (a) outside view drawing and (b) sectional drawing.

The measured experiments on these electrodes illustrate that the gap distances in BEAM with negative polarity of tool electrode are all about 400 µm, which is much higher than that in bundled EDM. Because of the higher discharge peak current and pulse duration in BEAM, the arc plasma column has been developed and elongated completely before reaching to the workpiece surface. Therefore, the machining depth can be calculated as 1.4 mm used in Figure 5, which has slight influence on the comparison results of the flushing velocity distribution for the three electrodes. Similar to a same open voltage, the comparison simulation adopts a same discharge gap distance for all the electrodes. The experiments also show that the lateral gap distance between tool and workpiece could reach up to 1 mm, having nothing to do with the electrode endface, peak current and flushing inlet pressure.

In order to further simplify the simulation model, some conditions and assumptions are given as follow:

1. The working fluid is a noncompressible and water-based liquid with 1 × 10³ kg/m³ density and 1 × 10⁻⁶ m²/s viscosity.

2. The inlet and outlet boundary conditions are set to flushing inlet pressure and outlet pressure. The flushing inlet pressure of each flushing pipe is 1.2 MPa. The flushing outlet pressure is the pressure of the atmospheric environment (1.01 × 10⁻¹ MPa).

3. Due to the high-velocity flushing, both the bubbles and the discharge-erosion particles generated into the discharge gap are flushed out in time. Therefore, the simulation of fluid field distribution only deals with the effect of fluid, regardless of the solid and gas which involve in multiphase flow. The Euler–Lagrange model is then selected as the CFD method for the simulation.

4. The calculated Reynolds number of the fluid flow model is 40,000, which is much greater than 2000. Therefore, it is better to choose the k − ε turbulent model.

Simulation results of the fluid field distribution

The flushing effect near the workpiece surface is the critical factor that influences the capability of arc breaking. Therefore, section A-A near the workpiece surface as presented in Figure 5(b) is selected to analyze and compare flushing ability for the electrodes. The simulation results of the selected plane presented in Figure 6 show that the flushing velocity increases dramatically from the center to the periphery in the radial direction for all the three electrodes. An explanation is that the working fluid from the fixture cavity cannot fully pass through all the holes into the gap with a constant velocity because of the small gap between the electrode and workpiece. Therefore, the flushing holes on the outermost ring give a priority to pass working fluid through compared to those on the innermost ring. At last, the flushing velocity will be accumulative from the inner to the outer in the radial direction.

Figure 6.

Simulation results of flushing velocity distribution on the section A-A: (a) E1, (b) E2 and (c) E3.

The flushing velocities of the central and peripheral points as well as of the average area on the section A-A are given and compared to estimate the flushing effect for the electrodes. The maximum flushing velocity of the electrodes E1, E2 and E3 is at the peripheral point, which is 46.4, 45.3 and 43.3 m/s, respectively. The minimum flushing velocity of the E1, E2 and E3 at the central point of workpiece surface is 3.67, 3.04 and 3.51 m/s, respectively. It can be concluded that the E1 has higher flushing velocity at the central and the peripheral points than the E2 or E3. However, the average flushing velocity (area-weighted average velocity in Fluent) of the E1, E2 and E3 on the section A-A is 17.61, 18.67 and 19.45 m/s, respectively. Thus, the velocity gradient in Figure 6 of the E3 changes slower than the E2 and much slower than the E1, indicating that the E3 has a more uniform flushing velocity distribution on the workpiece surface than the E2 and much more than the E1. In the next section, the comparison experiment is carried out to clarify the prime factor that influences the machining performance of BEAM, a higher flushing velocity at a particular point or a relatively uniform flushing velocity distribution on the workpiece surface.

Setup and conditions of the comparison experiment

An important part of BEAM system is the experiment setup which is mainly composed of the flushing apparatus, a tool fixture with a multi-hole solid electrode, as shown in Figure 7. The working fluid is provided by the flushing inlet from the pump whose maximum pressure is up to 1.2 MPa. The system usually uses high peak current and long pulse duration to obtain arc discharging conveniently. The maximum peak current of BEAM power is 500 A, and the pulse duration is 10,000 µs.

Figure 7.

Experimental setup with a multi-hole solid electrode.

The machining conditions of the comparison experiment for the multi-hole solid electrodes are listed in Table 2. During machining, the tool electrode feeds in a sinking mode accompany with a helical motion. The helical motion diameter is 2 or 3 mm according to the size of the maximum flushing-hole diameter, the helix angle is 5° and the feed depth is 3 mm. The main purpose of helical motion is to eliminate the nubbles left on the holes area which seriously block fluid passing through the holes into the discharge gap.

Table 2.

Experimental conditions.

Tool polarity	Negative
Tool material	Graphite
Workpiece material	Nickel-based alloy GH4169
Pulse duration, ton (µs)	6000
Pulse interval, toff (µs)	500
Peak current, Ip (A)	200, 300, 400, 500
Open voltage, U (V)	90
Flushing inlet pressure, p (MPa)	0.6, 0.9, 1.2
Machining depth, h (mm)	9

In rough machining, the MRR is the primary factor in machining performance to estimate the quality of flushing-hole distribution and then following by the TWR. A special Ra is selected as an estimating factor to determine the machinability for the next semi-finish machining because of the very coarser workpiece surface. The MRR is determined by calculating the removed volume of the workpiece material per minute according to the lost weight. The TWR is measured by taking the ratio of volumetric loss of the tool electrode material over the counterpart of the workpiece. The Ra is determined by the profile height curve in the perpendicular direction from the lowest point on the recast layer to a reference point on the machined surface, which is measured out by a surface profiler.

Results and discussions of the comparison experiment

Comparison experiment under different flushing inlet pressures

In order to verify the simulation results, the comparison experiment is first carried out with these multi-hole solid electrodes presented in Figure 4 under different flushing inlet pressures. The obtained results of MRR and TWR are shown in Figure 8. It is concluded that the MRR increases gradually with an increasing flushing inlet pressure for all the electrodes. The maximum MRR (mm³/min) of electrodes E1, E2 and E3 is up to 5463, 5629 and 6705, respectively, while the flushing inlet pressure is 1.2 MPa. Correspondingly, the maximum specify energy removal efficiency (mm³/A min) is about 13.7, 14.1 and 16.7, respectively. Increasing the flushing velocity can efficiently break arc plasma off and quickly flush out of the erosion debris and discharge heat, which promotes the effective rate of arc discharge in a single duration. In that case, a higher feedrate is adopted during the whole machine process. The MRR of E3 is 19.1% higher than E2 and 22.7% higher than E1. It is concluded that E3 achieved a better performance on MRR than E1 or E2 under a same flushing inlet pressure, which has the best flushing velocity distribution according to the simulation results above.

Figure 8.

Machining performance of BEAM under different flushing inlet pressures (I_p = 400 A).

TWR presented in the Figure 8 gradually decreases with the increase in the flushing inlet pressure for all the electrodes. When the flushing inlet pressure is 1.2 MPa, the TWR (%) of E1, E2 and E3 is 4.9, 4.7 and 4.3, respectively. First, a graphite electrode has a high melting point and a low linear expansion coefficient to retain from discharge heating wear, resulting in a relatively stable tool wear in a single duration. Second, tool electrode set as negative polarity will receive a little part of the heat source produced by arc discharge according to the energy distribution coefficient on the cathode and anode of the arc plasma, which obviously decreases the electrode wear. Third, a high-speed fluid can quickly take the discharge heat away and then cool down the electrode surface to resist the thermal effect. All the three reasons reveal that increased volume of electrode wear is less than that of removal material, bringing out a decreased TWR. The TWR of E3 is 8.5% lower than E2 and 12.2% lower than E1. It is also concluded that E3 achieves a better performance on TWR than E1 or E2 under a same flushing inlet pressure.

Comparison experiment under different peak currents

In order to further verify the simulation results, the comparison experiment is then carried out with these electrodes under different peak currents. The MRR presented in Figure 9 increases linearly with an increasing peak current for all the electrodes. The maximum MRR (mm³/min) of E1, E2 and E3 is up to 5978, 6379 and 7407 under the maximum peak current of 500 A, respectively. Correspondingly, the specify energy removal efficiency (mm³/A min) is about 12, 12.8 and 14.8 respectively. In order to explain the higher energy density of arc discharging in BEAM with the increase in the peak current, the waveforms of gap voltage and discharging current are sampled by an oscilloscope, as shown in Figure 10. The arcs presented in this figure are disturbed by high-velocity working fluid, and consequently, the gap voltage varied significantly, but the current remained relatively stable at 500 A, which is set to the pulse power, and the arcs seldom broke during a pulse duration. Therefore, the arc plasma can stably develop to remove the molten martial because of its higher energy density, indicating more workpiece materials eroded in single duration. The MRR of E3 is 16.1% higher than E2 and 23.9% higher than E1. It can be concluded that E3 achieves a better performance on MRR than E1 or E2, which has the best flushing velocity distribution according to the simulation results above.

Figure 9.

Machining performance of BEAM under different peak currents (p = 1.2 MPa).

Figure 10.

Waveforms of gap voltage and discharge current in BEAM (I_p = 500 A, t_on = 8000 µs, t_off = 500 µs and p = 1.2 MPa).

The TWR presented in the Figure 9 gradually decreases with the increase in peak current for all the electrodes. The energy density of the arc plasma column is increased with the increase in peak discharge current, hence removing more materials from both the electrode and the workpiece in positive BEAM. Since the good heat resistance as well as thermal and electrical conductivity of the graphite material, the removed volume of the electrode material is much less than that of the workpiece while the arc plasma moving/breaking off on the cathode electrode surface with adequate flushing velocity in the gap. The results of decreasing tool wear are also found in EDM and explained of arc spots sliding on electrodes. ²⁶ When the peak current is 500 A, the TWR (%) of E1, E2 and E3 is 4.2, 4.0 and 3.7, respectively. The TWR of E3 is 7.5% lower than E2 and 11.9% lower than E1. Therefore, TWR of the E3 is the least under a same peak current.

From the machining performance on MRR and TWR for all the electrodes, it can be concluded that a larger peak current and flushing inlet pressure in BEAM are preferred, which not only increase the MRR but also decrease the TWR. The results of the comparison experiment indicate that E3 has a better machining performance than E1 or E2, which corresponds to the simulation results. It also demonstrates the relationship between machining performance and flushing velocity distribution that the performance of multi-hole electrode mainly depends on a relatively uniform flushing velocity distribution on the workpiece surface rather than the highest flushing velocity at a particular point in the gap.

Influence of flushing holes on the Ra in BEAM

Due to the rather coarse surface in rough machining presented in Figure 11, BEAM selects Ra to preliminarily weigh the surface performance which is useful to next procedure such as positive tool polarity BEAM presented in the study by Xu et al. ²² The workpieces machined out under the optimized machining parameters are selected as the comparative objective. Figure 11 also shows that the maximum depth of discharge crater is usually found at the peripheral area where the flushing velocity is the highest. Therefore, a comparison analysis on Ra is carried out on this area to report a profile height curve. As presented in Figure 11, the maximum Ra (µm) of E1, E2 and E3 is 300, 210 and 130, respectively. In addition, the height gradient of E3 varies more slowly than E2 or E1. It is concluded that electrode E3 attains the best surface performance on Ra. An explanation is that E3 has the most reasonable flushing velocity distribution among the three electrodes, which brings out a uniform crater depth and minor height variation on the workpiece surface.

Figure 11.

Ra for the three electrodes measured by a surface profiler (I_p = 500 A and p = 1.2 MPa).

Comparative performance between a multi-hole solid electrode and a bundled electrode

Similar to a type of multi-hole electrode in BEAM, a multi-hole solid electrode attains the maximum MRR at 7407 mm³/min with the best flushing-hole distribution, flushing inlet pressure and peak current, which is 50% of that a bundled electrode under the same machining conditions. ¹⁶ An explanation is that a bundled electrode has a unique structure that there are many interspaces among flushing holes which can also pass the fluid through, as shown in Figure 1. Therefore, they will advance the fluid field distribution on the surrounding of a flushing hole and finally bright out a whirlpool effect on this area. In that case, the discharge arc plasma can be moved and broken off quickly, promoting to eject the molten material out and then produce deep discharge crater on the workpiece surface. However, the coupled property of the interspaces in bundled electrode will result in many residual protuberances on the interspaces area of the workpiece which will seriously increase the Ra. Therefore, a multi-hole solid electrode can attain a low Ra at 131 µm, which is 36% of that with a bundled electrode (more than 364 µm) under the same machining parameters. ¹⁶ As a result, a multi-hole solid electrode used in BEAM as rough machining can decrease the process time for the next semi-finish machining. Compared with a bundled electrode, it can efficiently decrease the machining allowance especially machining a complicated 3D cavity or an irregular cross-section profile.

Machining verification

A 3D cavity of similar tunnel in turbine disk is machined to verify the feasibility of applying a multi-hole solid electrode, which has utilized the flushing-hole distribution to optimize the electrode design. As shown in Figure 12(a), the diameter of the coarse electrode is of 50mm. The maximum length and width of the electrode working surface milled out by a graphite CNC machine are of 40mm and 20mm respectively. The workpiece material is the successive material GH4169. The complicated contour profile cavity is machined out by negative tool polarity BEAM in a sinking method, as shown in Figure 12(b). According to the comparison experiment, the prime machining process parameters are set as the highest peak discharge current of 500 A and flushing inlet pressure of 1.2 MPa to obtain high machining efficiency. The corresponding MRR is of 6300 mm³/min, and the TWR is of 5%. Therefore, it reveals a promising method of machining a complicated 3D cavity with a multi-hole solid electrode in a sinking mode.

Figure 12.

3D cavity machined by optimized multi-hole solid electrode in BEAM.

Conclusion

In order to investigate the influence of flushing holes on the machining performance of BEAM, the comparison simulation and experiment are carried out with three multi-hole solid electrodes:

1. The results of the comparison simulation indicate that a better flushing-hole distribution can significantly improve the fluid field distribution in the gap.

2. The results of the comparison experiment indicate that higher peak current and flushing inlet pressure are preferred for the multi-hole electrodes to attain a higher MRR and a lower TWR.

3. The combinative results of the comparison simulation and experiment illustrate that a flushing velocity distribution rather than the highest velocity at a particular point is the crucial factor that influences the machining performance in BEAM.

4. Furthermore, the combinative results also demonstrate that a multi-hole solid electrode with flushing-hole diameters decreasing gradually from the inner to the outer on the baseline in the radial direction attains the highest MRR, the lowest TWR and the least Ra. Therefore, the principle of flushing-hole distribution could be put forward to optimize the design of the multi-hole solid electrode or even other multi-hole electrodes.