Observation of Gap Phenomena and Development Processing Technology for ECDM of Sapphire

Abstract

The main purpose of this study was to develop observation techniques and processing technology for the electrochemical discharge machining (ECDM) of sapphire wafers. To measure the effect of gas-film thickness, discharge-spark conditions, and droplet sliding frequency on machining quality and efficiency in ECDM, this research utilized high-speed cameras to observe the gas film thickness and formation of the gas film during ECDM. Additionally, this study observed the machining-gap phenomena during ECDM. The formation mechanism and machining characteristics of the gas film were understood through experiments. The machining parameters included the liquid level, working voltage, rotation speed, and duty factor. This study analyzed and discussed the effect of each machining parameter on the gas-film thickness, current, electrode consumption, and droplet sliding frequency. Moreover, this study aimed to obtain optimized machining parameters to overcome the difficulty of machining sapphire. The experimental results indicated that utilizing a high-speed camera to capture the phenomena between electrodes during electrochemical discharge could effectively observe the gas-film thickness and the coverage of the gas film. A higher bubble coalescence rate enhanced the machining capability and reduced the lateral discharge. Therefore, this study could obtain better machining-hole depths through observation and analysis to improve gas-film stability and machining capability. This study demonstrated that a liquid level of 700 µm, a working voltage of 48 V, a duty factor of 50%, and a tool electrode rotational speed of 200 rpm could achieve a hole depth of 86.7 µm and a hole diameter of 129.5 µm.

1. Introduction

Sapphire, a single-crystal aluminum oxide, is a preferred raw material for optoelectronic components because of its wide optical transparency band, hardness, high melting point (2045 °C), corrosion resistance, and high-temperature resistance. Sapphire is gradually replacing strengthened glass as a new favorite among mobile-phone suppliers, becoming increasingly used in lenses, screens, and back panels. However, it is challenging to process sapphire with traditional mechanical methods due to its hard and brittle characteristics. Because of the great hardness of sapphire, conventional machining processes are nearly useless for delicate processing. Using high-hardness tools for this purpose results in high tool costs and other problems, such as workpiece fracture or quick tool wear. This results in increased processing costs, which is not conducive to industrial competitiveness.

This study investigated non-traditional electrochemical discharge machining (ECDM) as a non-contact material removal technique to process brittle and hard materials. Non-traditional machining methods used for processing non-conductive materials include electrochemical discharge machining, abrasive jet machining, water-jet machining, laser machining, and etching [1]. Compared to the aforementioned methods, ECDM boasts excellent processing speeds and minimal heat-affected zones for non-conductive materials. Therefore, ECDM is receiving increased attention as an emerging non-traditional machining technology. This method utilizes the energy generated by the discharge phenomenon to perform chemical etching and high-temperature melting, offering good processing flexibility and speed for non-conductive materials. Hence, it holds great potential for development.

A number of researchers have investigated the impact of gas-film thickness and stability on the reproducibility of ECDM [2,3,4]. Hajian et al. studied the influence of a magnetic field on the performance of ECDM, indicating that the magnetic field affects the movement direction of bubbles [5]. Appalanaidu et al. utilized electrostatic forces acting on bubbles to control the shape of the gas film, improving the processing results [6]. Yang et al. observed the experimental process of the electrochemical discharge machining of borosilicate glass using high-speed cameras and analyzed the mechanism of ECDM [7]. Mediliyegedara et al. use neural networks to classify pulses into five groups by observing voltage and current waveforms. They train the neural network to classify pulses with various functions and compare five different activation functions [8]. West et al. conducted performance experiments on ECDM with different polarities and studied the morphology after processing. They evaluated both anodic and cathodic methods. The results provide additional insights into the complexities and polarity-related mechanisms involved in the ECDM process. The anodic method generates unique spherical cavity microstructures, offering a new perspective for glass microfabrication [9]. Peng et al. proposed travel wire electrochemical discharge machining (TW-ECDM) as a technique for cutting small-sized optical glass and quartz rods. They discussed the energy release intensity and its physical phenomena under different wire diameters, power modulation methods, and electrolyte supply modes, and used a pulsed direct-current power supply to demonstrate the better spark stability and a higher proportion of spark energy release compared to a constant direct-current power supply. Finally, they summarized the suitable cutting conditions for quartz and borosilicate optical glass materials [10]. In order to enhance the quality of micro-holes in ECDM, Zheng et al. designed flat sidewall tool electrodes to reduce the conicity phenomenon caused by sidewall discharge. Additionally, they applied pulse voltage to improve the thermal impact zone in the rectified DC voltage. The experimental results demonstrate that the combination of flat sidewall tool electrodes and pulse voltage significantly improves machining accuracy [11]. Zheng et al. proposed a novel pulse voltage configuration design called offset pulse voltage and applied it to ECDM to enhance gas-film stability. The experimental results demonstrate that, compared to traditional ECDM generated by pulses, the configuration design of offset voltage during pulse-off time produces a more stable discharge [12]. Kim et al. applied ECDM to the micro-drilling of glass. One of the drawbacks of ECDM is the residual heat-affected zone (HAZ) left on the surface of micro-drilled holes. In order to reduce HAZ, a series of rectangular voltage pulses were applied in that study. The experimental investigation focused on the influence of voltage pulse frequency and duty cycle on glass ECDM. The experimental results showed that the thermal damage of micro-drilled holes decreased with the increasing frequency and decreasing duty factor [13]. Han et al. utilized ultrasonic vibration in ECDM to ensure sufficient electrolyte flow, which helps maintain stable spark generation and increases the machining depth of ECDM drilling. Additionally, by employing side insulation electrodes and utilizing pulse power, they improved the stability of gas-film formation and the surface quality of the hole entrance [14]. Cao et al. studied the effects of electrolyte, duty factor, voltage, feed rate, rotational speed, and electrolyte concentration on the ECDM drilling and milling processes in order to improve the processing of glass 3D microstructures and achieve good surface microstructures [15]. Tang et al. used high-speed cameras to study the evolution of the gas film around the tool electrode during ECDM glass machining and investigated the influence of current pulses on the gas-film state [16]. Jiang et al. introduced an analysis model for a gas film and used high-speed cameras to capture the gas film generated on the tool electrode during ECDM glass machining, finding good consistency between the experimentally obtained gas-film thickness and the model-estimated thickness [17]. Singh et al. discussed various aspects of ECDM, including the workpiece, electrolyte, tool electrode, gas-film formation, machining quality, and recent advancements in the ECDM process, summarized the research trends aimed at improving the efficiency of ECDM in the future, and addressed the challenges encountered during machining [18]. Sabahi et al. found that the wettability of the tool electrode, the surface tension of the bubble, and the electrolyte composition influence the thickness of the gas film through a theoretical analysis of single bubbles formed on the surfaces of tool electrodes [19]. Liu et al. investigated the effects of the applied voltage, pulse frequency, electrode rotational speed, and side-gap feed rate on wire electrochemical discharge machining (WECDM) and utilized WECDM for slicing glass. The experimental results indicated that the side gap increases with increasing the applied voltage and discharge coefficient, while it decreases with increasing the pulse frequency, feed rate, and electrode rotational speed [20]. Sundaram et al. investigated the influence of electrolyte level and pulse current on the formation of gas films and machining quality in the pulse electrochemical discharge machining (PECDM) of glass-fiber-reinforced epoxy composite materials [21]. Bian et al. obtained the voltage range for machining by studying the mechanism of tool electrode wear under different machining parameters in ECDM. The experimental results indicated that tool electrode wear mainly occurs at the edge tip due to localized melting or vaporization of the material at high temperatures [22]. Rajput et al. developed a thermal model to compare the material removal rate (MRR) under different electrolytes (NaOH, KOH, and NaCl) in ECDM. They obtained temperature distribution diagrams below the spark discharge and further processed them to evaluate the MRR. The predicted results were consistent with the experimental findings [23].

While some studies have used high-speed cameras to examine gas-film development, there is a lack of simultaneous observation of the droplet sliding frequency, gas-film thickness, and discharge-spark conditions in sapphire ECDM. Relevant studies on the influence of gas-film thickness, discharge-spark conditions, and droplet sliding frequency on the quality characteristics of sapphire ECDM drilling are still not clear. As a result, the goal of this study was to develop and build experimental equipment for the ECDM drilling of sapphire. High-speed cameras were utilized to observe various droplet sliding frequencies, discharge-spark conditions, and gas-film thicknesses. Through observation and analysis, this study investigated why certain key processing factors affect the quality features of sapphire ECDM drilling and electrode wear. Moreover, it also looked at approaches to enhance its processing capabilities and achieve deeper holes.

2. Experimental Principle

The basic principles of electrochemical discharge machining mainly involve spark discharge, chemical etching, and electrochemical reaction phenomena. The schematic diagram of ECDM is shown in Figure 1.

2.1. The Discharge-Spark Generation Mechanism of ECDM [24]

By observing the I-V curve shown in Figure 2, it could be noted that, with varying applied working voltages, the current characteristics curve of ECDM could be divided into five different regions:

• Region 1 (0–A): the electrolytic current approaches zero, and there are no electrolytic bubbles generated on the electrode surface;
• Region 2 (A–B): the electrolytic bubble generation rate and current increase linearly as the voltage increases;
• Region 3 (B–C): A relatively non-dense gas film is formed on the surface of the tool electrode, which increases the resistance between the electrolyte interface and the tool electrode. At this time, the current is stable and does not increase as the voltage increases;
• Region 4 (C–D): A dense and fine gas film forms on the surface of the tool electrode, creating high electrical resistance between the electrolyte interface and the tool electrode. This causes the current of the original electrochemical circuit to approach zero, and the voltage at this point is called the critical voltage (Vcrit), as shown in Figure 2;
• Region 5 (D–E): the energy contained in the gas film collapses through the film as the voltage rises over the critical voltage, resulting in insulation breakdown and the spark-discharge phenomenon.

2.2. The Material Removal Mechanism of Electrochemical Discharge Machining

2.2.1. The Electrochemical Discharge Machining Mechanism [24]

When a direct-current power supply is applied between the two electrodes, hydrogen gas bubbles are generated at the cathode due to the electrolytic reaction in the electrolyte. The chemical reaction is as follows:

$$2{\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{H}}_2+{\mathrm{O}}_2$$

(1)

Larger bubbles that have dispersed from the electrode surface tend to form when the quantity of bubbles is significant. With an increase in processing voltage, the quantity of bubbles grows significantly. The negative electrode surface becomes enveloped by a large number of bubbles, forming an insulating gas film. This insulating gas film creates high resistance, causing the current between the two poles to approach zero. The critical voltage refers to the working voltage at this point. Spark discharge is produced and insulation breaks down, as the voltage rises over the critical voltage as the energy in the gas film breaks through. The high temperature and energy produced by the spark discharge melt the workpiece surface, resulting in material removal.

2.2.2. Chemical Etching [25]

Sapphire is a single-crystal aluminum oxide. In this experiment, we used a relatively safe potassium hydroxide (KOH) alkaline electrolyte to conduct experiments on the sapphire substrate. The chemical reaction is as follows:

$$2\mathrm{K}\mathrm{O}\mathrm{H}+{\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3\to 2\mathrm{K}\mathrm{A}\mathrm{l}{\mathrm{O}}_2$$

(2)

If we look at the relationship between the etching reaction-rate constant (v) and temperature, the reaction rate of sapphire material in an alkaline solution at room temperature is extremely slow. The relationship is expressed by the following equation:

$$v=R\mathrm{e}\mathrm{x}\mathrm{p}(-E/K_{g}T)$$

(3)

where v is the etching rate, R is a frequency factor independent of the etching temperature T, E is the apparent activation energy, and K_g is the Boltzmann constant ($1.38\times {10}^{-23}{\mathrm{m}}^2\mathrm{k}\mathrm{g}{\mathrm{s}}^{-2}{\mathrm{K}}^{-1}$). The above formula shows that the high temperature around the ECDM area will accelerate the chemical etching rate of the sapphire substrate in the KOH solution. The etching rate will exhibit an exponential increase.

3. Experimental Setup

3.1. Experimental Material

3.1.1. Sapphire Wafer Workpiece

The sapphire wafer used in this experiment had a diameter of two inches and a thickness of 200 µm. Its primary chemical composition was made up of Al₂O₃, and its crystal orientation was in the C-plane (0001), as illustrated in Figure 3.

3.1.2. Cylindrical Tool Electrode

The tool electrode material used in this study was tungsten carbide. Figure 4 shows the dimensions of the tool electrode, and

Table 1. Dimensions of each item of cylindrical tool electrode.

Items	Dimensions
Tool electrode holder diameter (mm)	3.175
Tool electrode diameter (mm)	0.05
Tool electrode length (mm)	0.9

provides the dimensions of the cylindrical tool electrode.

3.2. Experimental Setup

This study utilized a high-speed camera to capture and observe the generation of bubbles, the formation of gas films during electrochemical discharge, and the phenomena between electrodes. Through experimentation, the goal was to understand the mechanism of gas-film formation and its impact on machining characteristics. The schematic diagram of the machining equipment is shown in Figure 5. The tool electrode was installed on the spindle of a CNC micro-EDM machine, and the spindle was connected to a power supply and a signal generator to generate an offset pulsed voltage for ECDM. To create the offset pulsed voltage, an N-type metal–oxide–semiconductor field-effect transistor (MOSFET) was installed. Using this transistor, a square wave offset pulsed voltage of 10 V was generated, as shown in Figure 6. The blue line represents the voltage square wave with an offset pulsed voltage of 10 V, while the yellow line represents the voltage waveform without using the offset pulsed voltage. When the offset pulsed voltage was applied, the bubbles on the tool electrode did not completely dissipate during the pulse-off time. This was because the working voltage was maintained near the critical voltage during the pulse-off time, which continuously generated bubbles and prevented them from completely dissolving. This helped maintain the stability of the gas film.

This experiment employed the single-factor experimental method to investigate various parameters. The single-factor parameters included the liquid level (above the workpiece), working voltage, rotation speed, and duty factor. The machining conditions were observed using a high-speed camera, and the electrolyte was added up to the set liquid level. The liquid level of the electrolyte was continuously observed using the high-speed camera during ECDM, and the electrolyte was replenished as needed to maintain a constant electrolyte level and electrode contact area. The initial gap between the workpiece and the tool electrode was set at 5 μm, and the total machining stroke was 125 μm. Data measurements were conducted when the working depth reached 5, 45, 85, and 125 μm. The schematic diagram of the machining stroke is shown in Figure 7. The measurement procedure involved the use of an oblique microscope to record the droplet sliding frequency, an oscilloscope to measure the current waveforms and current magnitudes, and a high-speed camera to record the gas-film thickness. A microscopic image-measuring device was then used to evaluate the electrode wear and micro-hole diameter once the data had been examined. The morphology and depth of the machined micro-holes were observed using a confocal laser microscope.

3.3. Experimental Parameters

This study utilized tungsten carbide tool electrodes for the observation of inter-electrode phenomena and the investigation of ECDM techniques on sapphire. The single-factor experimental parameters included the liquid level, working voltage, rotation speed, and duty factor. The study explored the characteristics of the gas-film thickness, droplet sliding frequency, current, electrode wear, hole diameter, and hole depth. Observations of the microstructure of the sample surface were conducted using a confocal laser microscope, a microscopic image-measuring instrument, and scanning electron microscopy (SEM). The experimental parameters and their settings are presented in

Table 2. The variation parameters and setting values of the experiments.

Parameter	Description
Liquid level (µm)	600, 650, 700, 750, 800
Working voltage (V)	42, 44, 46, 48, 50
Rotation speed (rpm)	50, 100, 150, 200, 250
Duty factor (%)	30, 40, 50, 60, 70

, while

Table 3. The constant parameters and setting values of the experiments.

Parameter	Setting Value
Total feeding stroke (µm)	125
Feed rate (µm/min)	8
Electrode diameter (µm)	50
Electrolyte	KOH
Electrolyte concentration (M)	7
Pulse duration (µs)	10
Auxiliary electrode	Graphite
Tool electrode	Tungsten carbide
Initial gap (µm)	5
Workpiece thickness (µm)	200
Distance between Auxiliary electrode and tool electrode (mm)	80

outlines the fixed factors and their parameter settings in this study.

3.4. Experimental Measurement and Observation

3.4.1. Droplet Sliding Frequency Measurement Method

In this experiment, an oblique microscope was used to record electrochemical discharge sparks and bubbles. The actual droplet formation process from 0 s to 55 s is depicted in Figure 8. Due to the splashing of the electrolyte caused by electrochemical discharge sparks, large liquid–gas mixed droplets were generated above the electrode over time, as shown by the circle in Figure 8. The droplet slid down when it became large enough to destroy the gas film. The gas-film destruction process lasted from 0.06 s to 0.11 s, as depicted in Figure 9.

3.4.2. Tool Electrode Wear Measurement Method

An optical microscopic image-measuring instrument was utilized to measure the processed tool electrode. In this experiment, electrode diameter wear was defined as the diameter wear at the front 100 µm of the tool electrode, as illustrated in Figure 10 and Equation (4).

$$D=D_0-D_1$$

(4)

where D represents the diameter loss of the tool electrode, D₀ is the original diameter of the tool electrode, and D₁ is the diameter at a necking point 100 µm from the front end of the tool electrode.

3.4.3. Gas-Film Thickness Measurement Method

In this experiment, the gas-film thickness of the tool electrode at a position 100 µm above the surface of the sapphire sample was measured, as illustrated in Figure 11.

4. Result and Discussion

4.1. Effect of Liquid Level on Sapphire ECDM

This study focused on the observation of inter-electrode phenomena and the investigation of ECDM techniques on sapphire. To assess the influence of the liquid level on ECDM, comparative experiments were conducted, with liquid levels set at 600 µm, 650 µm, 700 µm, 750 µm, and 800 µm. The experiments used a combination of parameters with a working voltage of 48 V, a pulse duration of 10 µs, a duty factor of 50%, and a rotation speed of 50 rpm. Except for the tool electrode contacting the workpiece for processing and the Z-axis of the electrode having no feed, all machining parameters followed the fixed factor parameter settings shown in Table 3. The tool electrode stayed at the workpiece surface for one minute to observe the droplet sliding frequency and measure the gas-film thickness, average current, machining-hole depth, and micro-hole volume. By comparing the volume of the micro-holes produced through ECDM, this study assessed the machining capabilities at different liquid levels and compared the depths of micro-hole machining. Figure 12 depicts the volume of the micro-holes produced through ECDM at different liquid levels. It could be observed that, as the liquid level increased from 600 µm to 700 µm, the volume of the machined micro-holes increased. However, the volume of the machined micro-holes decreased when the liquid level increased from 700 µm to 800 µm. This could be explained by examining the machining-current performance and the spark using high-speed photography. The higher machining current caused rapid electrode wear as the liquid level increased from 700 µm to 800 µm, leading to a decrease in machining efficiency. Figure 13 presents the average current at different liquid levels. The average current was obtained from the oscilloscope. Figure 14 shows the current waveforms at different liquid levels. It could be observed that, at a liquid level of 600 µm, more electrolysis phenomena occurred, as indicated by the circle. The differences in the remaining current diagrams are not significant. The sharply rising current waveform represented the electrolysis of water. This was because the distance between the workpiece surface and the liquid surface was short, resulting in a small space for bubble accommodation. The small buoyancy of the bubbles made them less likely to dissipate, leading to the accumulation of more and larger bubbles. This increased the resistance of the electrolyte, causing the machining current to be much smaller than that at other liquid levels. As the liquid level increased from 650 µm to 800 µm, the machining current increased with the liquid level and approached a linear growth. This occurred because the contact area between the electrode and the electrolyte increased with the increase in the liquid level, leading to an increase in the current.

Figure 15 shows the current density at different liquid levels, calculated by dividing the average current by the electrode surface contact area with the electrolyte. It could be observed that, at a liquid level of 600 µm, the current density was relatively low due to the small average current. As the liquid level increased from 600 µm to 700 µm, the current density increased with the liquid level increase. However, the current density decreased when the liquid level exceeded 700 µm. It could be observed that the workpiece surface was farther away from the liquid surface after a liquid level of 700 µm, resulting in a larger space for bubble accommodation, as shown in Figure 16. The increased buoyancy of the bubbles caused them to disperse upwards, preventing the rapid coalescence of bubble particles. As a result, the average bubble size became smaller with the higher liquid level, and the gas film became thinner and denser. The resistance of the electrolyte increased, leading to a decreasing trend in current density from 700 µm to 800 µm. The current density curve followed a similar trend to the micro-hole volume curve at different liquid levels, as both increased with the liquid level from 600 µm to 700 µm and then decreased after the liquid level exceeded 700 µm.

Figure 17 shows the droplet sliding frequency under different liquid levels. It could be observed that, as the liquid level increased, the rate of droplet formation and sliding from the upper surface increased. And, the gas film became denser with the higher liquid level. Therefore, the gas film was less dense at a liquid level of 600 µm and 650 µm, leading to a higher occurrence of electrolysis in the electrochemical reaction and more bubble generation. This resulted in a faster rate of droplet formation and sliding from the upper surface. When a droplet fell into the electrolyte, it disrupted the gas film formed on the electrode surface. This phenomenon also affected the processing capability at lower liquid levels.

Figure 18 shows the morphologies of the deepest and shallowest holes obtained under different liquid levels. From the SEM images, it could be observed that the insulating gas film was thinnest at the edges and corners of the tool electrode. As a result, the discharge sparks concentrated on the outer side of the circle. This could be seen in the SEM images, where the machining started from the outer side, forming a donut-shaped hole. It was evident that the micro-hole depth at a liquid level of 700 µm was significantly deeper compared to a liquid level of 800 µm. At a liquid level of 700 µm, the higher current density resulted in more efficient machining of the sapphire. The inner circle of the donut was smaller at this liquid level. However, excessive lateral discharge energy caused rapid electrode wear, as the liquid level increased from 700 µm to 800 µm, leading to a decrease in machining efficiency. Furthermore, the bottom of the micro-hole received a higher discharge energy that caused the workpiece to fracture, as illustrated by the circle in the figure.

Figure 19 illustrates the sidewall morphology of the micro-holes under different liquid levels. It could be observed that, at a liquid level of 700 µm, the maximum micro-hole machining depth could be achieved.

According to a comprehensive analysis of the experiment’s findings, smaller micro-hole volumes and lower machining capabilities were produced at a liquid level of 600 µm and 650 µm due to the reduced formation of dense insulating film, relatively large and persistent bubbles, and faster droplet formation and sliding frequency. On the other hand, the denser insulating coating resulted in a reduced current density and quick electrode wear due to excessive lateral discharge energy at a liquid level of 750 µm and 800 µm. These factors also affected the machining capabilities and decreased the volume of the micro-holes. However, the maximum micro-hole machining volume and depth were achieved at a liquid level of 700 µm. Therefore, a liquid level of 700 µm was chosen for the subsequent parameter experiments.

4.2. Effects of Different Parameters on Sapphire ECDM

This study used a tungsten carbide cylindrical tool with a diameter of 50 μm to perform offset pulsed voltage-assisted ECDM drilling on a 200 μm sapphire substrate. Except for the feed rate set at 8 µm/min, all parameters were set according to the fixed factor parameter values shown in Table 3. This study observed, evaluated, and discussed the effects of several machining parameters, such as the rotation speed, working voltage, and duty factor, on the quality features of sapphire and the machining phenomena through single-factor experiments. The observed machining phenomena included the insulating gas-film thickness, current, and droplet sliding frequency. The quality characteristics included the micro-hole diameter, micro-hole depth, and tool electrode wear.

4.2.1. Effect of Different Working Voltage on Sapphire ECDM

The working voltage is a crucial parameter in ECDM, and the effect of different working voltages on the machining-hole diameter and depth at various electrode speeds is illustrated in Figure 20 and Figure 21. Observations under different working voltages revealed that, with the increasing working voltage, the machining-hole diameter and depth also increased. However, the machining-hole depth significantly decreased at a working voltage of 50 V. This phenomenon could be attributed to the fact that insufficient discharge energy leads to a machining speed that cannot keep up with the tool electrode feed rate at lower working voltages, resulting in tool electrode breakage and a shallower hole depth. The discharge energy and spark size increased as the working voltage increased, leading to a larger machining-hole diameter and depth. The deepest machining-hole depth was achieved at a working voltage of 48 V, and the tool electrode remained intact. However, the excessive discharge energy and sparks caused rapid wear of the tool electrode’s front end when the working voltage reached 50 V. When the tool electrode was fed down to −20 µm, the front end experienced significant wear due to the initially excessive discharge energy during the initial machining. This resulted in a larger gap between the tool electrode’s front end and the workpiece surface during subsequent machining. The reduced discharge energy at the tool electrode’s front end led to insufficient machining capabilities, causing a significant reduction in machining-hole depth.

Figure 22 presents the thickness of the insulating gas film under different working voltages. Figure 23 displays the high-speed camera images captured under various working voltages. Figure 24 provides a schematic diagram of the insulating film under different working depths. Figure 25 compares the insulating gas film under different working voltages using an oblique microscope. When the tool electrode was fed down to the 0 µm position and the working voltage was in the range of 42 V to 48 V, the appropriate discharge gap between the tool electrode tip and the workpiece surface could not be achieved. As a result, there were fewer discharge events at the tool electrode tip. A significant number of bubbles formed at the tool electrode tip and accumulated on the specimen’s surface when there were fewer discharge events at the tool electrode tip, as could be seen in the thickening of the insulating gas film. Under a voltage of 50 V, due to the higher voltage and increased discharge energy, the discharge gap was larger compared to that at lower voltages, resulting in noticeable discharge sparks. And, due to the higher working voltage and greater discharge energy at 50 V, more discharge sparks were visible than at lower voltages because of the bigger discharge gap. When the tool electrode was fed down to the position of −40 µm, the rotation of the tool electrode allowed for a smooth supply of electrolyte within the machining zone and contributed to the formation of the most stable and relatively thin insulating gas film on the tool electrode surface. This condition resulted in bright and stable spark discharges, creating the optimal processing capability compared to other positions. However, the tool electrode tip experienced significant wear at a working voltage of 50 V due to the overly large discharge sparks, which caused a huge discharge gap in the latter stages of processing. Consequently, the machining ability of the micro-hole was significantly reduced. When the discharge gap was not appropriate, the generated insulating gas film tended to be much thicker compared to the conditions where the discharge was stable. When the tool electrode was fed down to the position of −80 µm, the disturbance of the insulating gas film increased. This caused the insulting gas film to become thicker. The morphology of the insulating film became unstable, and the machining sparks were relatively less noticeable compared to the −40 µm position. The supply of the electrolyte became challenging as the working depth increased, resulting in a decrease in machining capability. The tool electrode was also prone to lateral discharge due to bending, leading to poorer machining capability compared to that at −40 µm. The disturbance of the insulating gas film became even more severe when the tool electrode was fed down to the position of −120 µm, and the insulating gas-film thickness was thicker than at −80 µm. The brightness of the sparks was low, causing a further reduction in machining capability. Across different working voltages, it was observed that higher working voltages led to thicker insulating gas-film generation, as the higher working voltages produced more bubbles, larger discharge sparks, and higher machining capabilities. However, if the sparks and discharge energy were too intense, they would cause excessive wear on the tool electrode’s front end, adversely affecting the machining capability.

Figure 26 illustrates the average current under different working voltages. As the feed depth of the tool electrode increased, the average current decreased. This was attributed to the increasing difficulty in supplying the electrolyte, resulting in an unstable and less dense gas film. At a working voltage of 50 V, and when the tool electrode was fed down to −20 µm, the front end of the tool electrode experienced significant wear due to excessive discharge sparks. This resulted in a substantial reduction in the length of the tool electrode. In the later stages of machining, the discharge gap became excessively large, causing a rapid decrease in the average current.

Figure 27 depicts the droplet sliding frequency chart under different working voltages. It could be observed that, as the working voltage increased, the electrolytic reaction produced more bubbles. Consequently, the generation and sliding frequency of the droplets also increased. The discharge sparks became increasingly unstable with the increased working depth, leading to a higher quantity of electrolytic bubbles. As a result, the number of electrolytic bubbles combining with droplets also increased, causing a faster rate of droplet generation and sliding. This led to more disruptions in the formation of gas films, resulting in a decrease in the machining capability.

Figure 28 illustrates the morphology of the tool electrode under different working voltages. At working voltages ranging from 42 V to 46 V, the insufficient machining capability resulted in the tool electrode falling behind in feed speed. This led to the bending and eventual fracture of the tool electrode in the later stages of machining. The tool electrode experienced smaller length loss when the machining capability and feed speed were closer, at a working voltage of 48 V, and there was no occurrence of tool electrode fracture. The front end of the tool electrode also generally maintained a cylindrical shape. However, the excessive discharge sparks and energy caused significant wear on the front end of the tool electrode at a working voltage of 50 V. This resulted in an average reduction in the length of the tool electrode to 100 µm and transformed the front end of the tool electrode into a conical shape.

Figure 29 illustrates the sidewall morphology of the micro-holes with the deepest and shallowest depths under different working voltages. It could be observed from the SEM image that, at a working voltage of 48 V, the micro-hole depth was noticeably deeper compared to the depth achieved at 50 V. And, under a working voltage of 48 V, the surfaces of the machined micro-holes appeared smoother. However, the bottoms of the micro-holes were subjected to momentary discharge energy due to the larger discharge sparks and energy, resulting in the formation of more microcracks under a working voltage of 50 V, as indicated by the circles in the figure. Furthermore, this led to significant wear on the front end of the tool electrode. In the later stages of machining at a working voltage of 50 V, the discharge gap became excessively large. This caused a substantial decrease in the capability to machine micro-holes in the workpiece.

Considering the results from the aforementioned experiments, we observed that, at different working voltages, the aperture and depth of the machined holes tended to increase with higher voltage settings. However, an interesting phenomenon occurred when the working voltage reached 50 V, where there was a noticeable reduction in the depth of the machined holes. In contrast, we achieved the deepest hole depth at a working voltage of 48 V, and the tool electrode did not experience any fracture. Therefore, further parameter experiments used a working voltage of 48 V.

4.2.2. Effect of Different Rotation Speeds on Sapphire ECDM

The machined hole diameters and hole depths under different rotation speeds are illustrated in Figure 30 and Figure 31. It was observed that, with higher rotation speeds from 50 rpm to 200 rpm, both the hole diameter and hole depth increased. However, there was a significant decrease in the hole diameter and hole depth at a rotation speed of 250 rpm. The increase in rotation speed from 50 rpm to 200 rpm promoted the circulation and fluidity of the electrolyte, facilitating bubble generation. Moreover, higher rotation speeds led to the easier dissipation of bubbles due to centrifugal force, resulting in a thinner gas film. The thinner gas films were prone to discharge insulation breakdown, contributing to improved machining capability but also causing lateral discharges and an increase in hole diameter. The extremely thin and dense gas film, coupled with intense discharge sparks and the energy at a rotation speed of 250 rpm, led to excessive wear at the front end of the tool electrode. Additionally, higher rotation speeds contributed to increased wear on the tool electrode in the later stages of machining, resulting in a larger gap between the front end of the tool electrode and the workpiece surface. This led to reduced discharge energy at the front end of the tool electrode, significantly decreasing the machining capability and resulting in a substantial reduction in hole depth.

Figure 32 displays the gas-film thickness under different rotation speeds, Figure 33 provides a comparison under an oblique microscope at various rotation speeds, and Figure 34 presents high-speed camera images under different rotation speeds. Observations under varying rotation speeds revealed that, as the rotation speed increased, the gas-film thickness decreased. This occurred because the higher rotation speeds promoted the circulation and fluidity of the electrolyte, and bubbles were more likely to dissipate due to the centrifugal force. As a result, the gas film around the tool electrode became thinner and denser. However, the film became extremely thin and dense at 250 rpm, with discharge sparks and energy surpassing other rotation speeds. This led to rapid wear at the front end of the tool electrode, causing a larger gap between the tool electrode’s front end and the workpiece surface in later stages of machining. This larger gap resulted in reduced discharge energy at the front end of the tool electrode, significantly compromising the machining capability. Additionally, the generated film thickness when the discharge gap was improper was much thicker compared to that under stable discharge conditions. At the tool electrode’s downward feed position of −80 µm, a substantial amount of bubble generation occurred. The bubbles accumulated within the observation window, making it impossible to measure the film thickness at that particular position. When the rotation speed increased from 50 rpm to 200 rpm and the tool electrode was fed down to 0 µm, an insufficient discharge gap was achieved between the front end of the tool electrode and the workpiece surface, as shown in Figure 33 and Figure 34. This resulted in fewer discharges at the front end of the tool electrode. It could be observed that, when there were fewer discharges at the front end of the tool electrode, the gas-film thickness became significantly thicker. Furthermore, a large number of bubbles emerged and accumulated on the surface of the workpiece. When the tool electrode was fed down to the position of −40 µm, a stable rotation of the tool electrode and a smooth supply of electrolyte within the machining area were achieved. The tool electrode surface possessed the most stable and thin film at this point, resulting in bright and stable spark discharges. This position exhibited better machining capability compared to the other working depths. The disturbance in the gas film increased when the tool electrode was fed down to the position of −80 µm, and it started to thicken. The gas-film morphology became unstable, and the machining sparks were relatively less noticeable, compared to when the tool electrode was fed down to −40 µm. The supply of electrolyte became challenging as the working depth increased, leading to a decrease in machining capability. Additionally, the tool electrode was more prone to lateral discharges due to bending, resulting in poorer machining capability compared to when the tool electrode was fed down to −40 µm. The disturbance in the gas film became more severe at a working depth of −120 µm, and the film thickness was even greater than at −80 µm. This resulted in very low spark brightness and a further reduction in machining capability. However, the discharge sparks and energy were too intense, when the rotation speed was 250 rpm and the tool electrode was fed down to −80 µm. This caused substantial wear at the front end of the tool electrode, leading to a significantly larger discharge gap in the later stages of machining. As a result, the capability to machine micro-holes in the workpiece was greatly diminished. When the discharge gap was inappropriate, the generated gas-film thickness was much thicker compared to that under stable discharge conditions.

Figure 35 illustrates the average current under different rotation speeds. As the rotation speed increased, the average current also increased. This was because higher rotation speeds resulted in a thinner film that facilitated discharge phenomena, leading to higher average currents. With an increase in working depth, the average current exhibited a decreasing trend. This was due to the difficulty in supplying electrolyte as the working depth deepened, causing instability and less density in the gas film. Additionally, the front end of the tool electrode experienced substantial wear at a rotation speed of 250 rpm and when the tool electrode was fed down to −80 µm. The discharge gap became excessively large at the later stages of machining, causing a rapid decrease in the average current.

Figure 36 depicts the droplet sliding frequency under different rotation speeds. It could be observed that, as the rotation speed increased, the droplet slide rate became faster. This was attributed to the larger discharge sparks and energy. The electrolyte splashed by the discharge sparks tended to adhere more easily to the tool electrode surface, leading to a faster rate of droplet generation and sliding. Additionally, the centrifugal force made it less likely for droplets to adhere to the tool electrode under higher rotation speeds, resulting in smaller droplets that slid more easily. When these droplets fell into the electrolyte, they disrupted the gas film formed on the tool electrode’s surface. This caused a decrease in machining capability. However, the front end of the tool electrode had already experienced substantial wear at a rotation speed of 250 rpm and when the tool electrode was fed down to −120 µm. Consequently, the phenomenon of splashing caused by spark discharge diminished, leading to a significant reduction in the speed of droplet generation and sliding.

Figure 37 illustrates the tool electrode wear under different rotation speeds. The tool electrode wear increased with higher rotation speeds. This was because the size of the discharge sparks and discharge energy increased with the tool electrode rotation speed. There was a drastic increase at a rotation speed of 250 rpm. This was because the faster rotation speed resulted in a thinner gas film, leading to larger discharge sparks and energy. This caused excessive tool wear at the front end of the tool electrode. The average wear length was 150 µm. Figure 38 provides images of the tool electrode under different rotation speeds. It could be observed that, with the increasing rotation speed, the front end of the tool electrode took on a more conical shape. This was attributed to the larger discharge energy, which caused wear at the front end of the tool electrode and resulted in a conical shape.

Figure 39 illustrates the sidewall morphology of the deepest and shallowest micro-holes after machining under different rotation speeds. From the SEM images, it was evident that the micro-hole depth at a rotation speed of 200 rpm was noticeably deeper compared to that at a speed of 50 rpm. This was because the faster rotation speed resulted in a thinner gas film and larger discharge sparks and energy, leading to a deeper hole depth. However, the larger discharge sparks and energy could cause unevenness at the bottom of the micro-hole. Under an SEM observation at a magnification of 1500, the sidewall at 50 rpm appeared flatter and smoother. This was attributed to the smaller discharge sparks and energy. The larger discharge sparks and energy resulted in irregular protrusions on the sidewall at 200 rpm, and the bottom of the micro-hole exhibited larger pits and a rough surface.

In summary of the experimental results, it was observed that the hole diameter increased with higher rotation speeds, and the working depth became deeper. However, there was a significant reduction in the hole depth at a tool electrode rotation speed of 250 rpm. The maximum working depth was achieved at a rotation speed of 200 rpm. The front end of the tool electrode did not experience severe wear at this rotation speed. Therefore, a rotation speed of 200 rpm was chosen for the subsequent parameter experiments.

4.2.3. Effect of Different Duty Factors on Sapphire ECDM

The hole diameter, hole depth, and hole morphology under different duty factors are shown in Figure 40 and Figure 41. It could be observed that, as the duty factor increased from 30% to 50%, both the hole diameter and hole depth increased. The deepest hole depth was achieved at a duty factor of 50%. However, the hole depth decreased when the duty factor increased from 50% to 70%, and the hole diameter reached its maximum at a duty factor of 70%. The shorter pulse-on time led to a reduction in bubble formation when the duty factor increased from 30% to 40%, resulting in a thinner gas film. The discharge sparks and energy were relatively smaller, leading to reduced machining capabilities. This could cause bending or even breakage of the tool electrode, resulting in a smaller hole diameter and hole depth compared to a duty factor of 50%. Stable and bright sparks could be observed at a duty factor of 50%, ensuring constant machining without excessive tool electrode wear. As the duty factor increased from 50% to 70%, a thicker gas film was obtained with a longer pulse-on time and faster bubble generation. The discharge sparks increased with the duty factor, leading to unstable gas film and extensive lateral discharge. This instability and sidewall wear on the tool electrode could cause the tool electrode to fracture before it was finished, in contrast to that of a duty factor of 50%, which would reduce the hole depth.

Figure 42 displays the gas-film thickness under different duty factors, Figure 43 provides a comparison under an oblique microscope at various duty factors, and Figure 44 presents high-speed camera images under different duty factors. The gas-film thickness increased with the increased duty factor. However, the gas-film thickness no longer significantly increased when the duty factor exceeded 50%. This was because the gas film had already fully and stably enveloped the electrode, making it less prone to contacting the electrolyte and forming bubbles. Therefore, no further significant increase in gas-film thickness occurred. Moreover, the discharge sparks and energy were significantly greater compared to the other duty factors at a duty factor of 70%. This susceptibility led to rapid wear at the tool electrode tip, causing a larger gap between the tool electrode tip and the workpiece surface. Consequently, the tool electrode tip’s discharge energy became insufficient, resulting in an inadequate micro-hole machining capability and shallower hole depths. Furthermore, a significant formation of bubbles occurred when the tool electrode was fed down to a working depth of −120 µm, as shown in Figure 44. These bubbles accumulated inside the observation frame. This accumulation hindered the measurement of the gas-film thickness at that working depth. When the tool electrode was fed down to a working depth of 0 µm under various duty factors, the inadequate discharge gap between the tool electrode tip and the workpiece surface resulted in a significantly thicker gas film, as shown in Figure 43 and Figure 44. There was less occurrence of discharge at the tool electrode tip, and a large number of bubbles emerged and accumulated on the workpiece surface. When the tool electrode was fed down to a working depth of −40 µm, the smooth rotation of the tool electrode and the seamless supply of electrolyte within the machining region led to the most stable and relatively thin gas film on the tool electrode surface. This resulted in bright and stable spark discharges, making this working depth the most suitable for machining. However, the gas film thickened and became more disturbed when the tool electrode was fed down to a working depth of −80 µm, giving an appearance of instability. The machining sparks were less pronounced compared to the tool electrode fed down to a working depth of −40 µm because the supply of electrolyte became more challenging with the increasing working depth, leading to a decrease in machining capability. The tool electrode was also more prone to lateral discharge due to bending, resulting in poorer machining capability compared to the tool electrode fed down to −40 µm. The disturbances in the gas film became more severe at a working depth of −120 µm, and the gas-film thickness was even greater than at −80 µm. The spark brightness was very low, causing a further reduction in machining capability. However, the excessive discharge sparks and energy led to substantial wear at the tool electrode tip when the duty factor was 70% and the tool electrode was fed down to a working depth of −120 µm. This resulted in a significantly enlarged discharge gap in the later stages of machining, severely diminishing the ability to machine the workpiece.

Figure 45 depicts the average current under different duty factors. As the duty factor increased, the average currents also increased. This was because the pulse-on time was longer at a higher duty factor, leading to a longer period of spark discharges within the pulse duration. With the increase in working depth, the average current tended to decrease. This was attributed to the growing difficulty in electrolyte supply as the working depth increased. The gas film became less stable and less dense, leading to a reduction in the frequency of spark discharges.

Figure 46 illustrates the droplet sliding frequency under different duty factors. It could be observed that, as the duty factor increased, the upper droplets were generated and fell at a faster rate. This was because the current increased with the increase in the duty factor. This resulted in more splashing phenomena and electrolysis reactions caused by the spark discharge, leading to a higher production of reaction products. Consequently, the rate at which the upper droplets slid also increased, disrupting the formation of the gas film more frequently and causing a decrease in machining capability. However, the extensive wear at the tool electrode tip reduced the splashing phenomenon caused by spark discharge when the duty factor was 70% and the tool electrode was fed down to −120 µm. As a result, the speed of droplet generation and falling decreased significantly.

Figure 47 presents the tool electrode wear under different duty factors. When the duty factors were 30% and 40%, the tool electrode experienced significant wear due to insufficient machining capability. This inadequacy led to the tool electrode bending in the later stages of machining, causing fractures. The machining capability matched the feed rate when the duty factor was 50%, resulting in a decrease in tool electrode wear, as shown in Figure 48. However, the excessive discharge sparks caused severe lateral discharges at the tool electrode as the duty factor increased to 70%. This led to a substantial reduction in the tool electrode diameter, making it extremely fragile and prone to breakage. Consequently, it became impossible to measure the tool electrode wear under these conditions.

Figure 49 displays the sidewall morphology of the deepest and shallowest holes machined under different duty factors, as observed through the SEM images. It could be noted that the micro-hole machining depth at a duty factor of 50% was deeper compared to that at a duty factor of 30%. The pulse-on time was shorter at a 30% duty factor, resulting in an extremely thin gas film and smaller discharge sparks. This led to a reduced machining capability and the occurrence of cracks on the sidewall, as indicated by the circles. However, the discharge sparks were sufficient to maintain stable machining without causing excessive wear at the tool electrode tip at a duty factor of 50%. It was observed that the sidewall morphology was flat and smooth at a duty factor of 50%.

Based on the results of the above experiments, it was observed that the machined hole diameter and depth increased as the duty factor increased from 30% to 50%. The tool electrode experienced excessive wear due to excessive discharge sparks when the duty factor was further increased from 50% to 70%, leading to a reduction in the hole depth. The maximum depth of the micro-holes was achieved when the duty factor was 50%.

5. Conclusions

This study focused on observing the phenomena and research on machining techniques for the ECDM of sapphire and pioneered the incorporation of droplet sliding frequency to assess its effect on machining outcomes. The experiments were conducted using an offset pulse method, and the impact of single-factor processing parameters on machining characteristics was investigated. Based on the experimental results, the following conclusions were summarized.

• When the liquid level increased from 600 µm to 700 µm, the volume of the machined micro-holes increased with the rising liquid level. At this liquid level, the current density also increased with the liquid level. However, the current density began to decrease when the liquid level exceeded 700 µm, resulting in a reduction in the volume of the machined micro-holes with the increasing liquid level;
• When the working voltage was increased from 42 V to 48 V, the hole diameter and hole depth of the micro-holes increased with the higher working voltage. However, excessive wear at the tool electrode’s tip led to a significant decrease in machining efficiency when the working voltage was raised to 50 V, resulting in the shallowest hole depth;
• Under different rotational speeds, the hole diameter and hole depth of the micro-holes increased with the higher rotational speed. However, excessive wear at the tool electrode’s tip led to a significant decrease in the machining efficiency when the rotational speed was raised to 250 rpm, resulting in a decrease in the machining depth;
• As the duty factor increased from 30% to 50%, the hole diameter and hole depth also increased. However, excessive wear on the tool electrode occurred due to overly large discharge sparks when the duty factor was increased from 50% to 70%. As a result, the machining depth decreased with the increased duty factor.
• When the optimal machining parameters were used, including a liquid level of 700 µm, a working voltage of 48 V, a rotation speed of 200 rpm, and a duty factor of 50%, a hole diameter of 129.5 μm and an average machining depth of 86.7 μm were obtained.