Overview of the Ultrasonic Machining

The potential of the sound waves with high frequency (about 70 kHz) for machining was first observed by R. W. Wood and A. L. Loomis in 1927 (McGeough 1988). Subsequently, the first patent on using ultrasound for machining was granted to L. Balamuth in 1945 (Balamuth 1945). In the investigation on the ultrasonic grinding of abrasive powders, the surface of a container filled with abrasive slurry was found to disintegrate in the areas close to the vibrating tip of an ultrasonic transducer. In addition, the shape of the generated cavity was exactly the same as that of the transducer tip (McGeough 1988). Consequently, the capabilities of this technique for machining applications were quickly recognized by industrial users. The production of ultrasonic machine tools then started in the early 1950s by mounting the first USM tools on the bodies of drilling and milling machines. By 1960, various USM tools were commercialized and utilized by manufacturers in regular production. Various terms have been coined for the USM such as ultrasonic drilling, slurry drilling, and ultrasonic abrasive machining. However, it is more commonly referred as ultrasonic impact grinding (USIG), ultrasonic-assisted lapping (USAL), and ultrasonic machining (USM).

Basic Elements and Working Principles of USM

Figure 14 shows the basic elements of an ultrasonic machining system. High-frequency electrical signal produced by an ultrasonic power generator is converted into mechanical vibration via a transducer-booster combination. The booster may increase, decrease, or retain the vibration amplitude received from the transducer. The mechanical vibration is then transmitted to a horn-tool assembly (known as sonotrode) which acts as an energy-focusing device. This results in vibration of the tool in the axial direction, typically at a frequency and amplitude ranging from 20 to 40 kHz and 5 to 50 μm, respectively. The power employed in the USM process usually ranges from 50 to 3,000 W, and it may reach up to 4 kW in some USM systems (Thoe et al. 1998).

Figure 15 illustrates the working principle of the USM process. As illustrated, abrasive slurry which consists of a powder dispersed in a liquid medium (water or oil) is fed into the machining zone between the tool and workpiece. Abrasive materials such as boron carbide, boron nitride, aluminum oxide, silicon carbide, and polycrystalline diamond with particle size of 50–300 μm are commonly used. A downward controlled force (also known as static load or machining force) is applied on the vibrating tool to maintain a certain gap distance between the tool and workpiece (known as machining gap). The vibrating tool causes the abrasive particles within the machining zone to impact the workpiece surface causing indentation, microcracks, and eventually material removal. This process continues leading to the formation of a cavity whose geometry is similar to that of the tool tip.
Commonly used materials for booster and horn are Monel, titanium alloys (e.g., TiAlV64), stainless steel, and aluminum alloys (e.g., AlCuMg2) (McGeough 1988). Titanium and aluminum alloys have high fatigue strengths, and hence they can be used for high vibration amplitudes of up to 40 μm at 20 kHz vibration frequency. The tools used in USM process are usually made of materials with high wear resistance and fatigue strength properties (Thoe et al. 1998). Tool materials such as tungsten carbide, silver steel, pure tungsten, and copper are commonly used for machining of the materials with low fracture toughness such as glass, while chromium-nickel steel is recommended for machining of the materials with higher toughness such as sintered carbides (McGeough 1988).

Process Parameters and Performance Measures

Ultrasonic machining is a free abrasive machining process where abrasive particles are in interaction with both the tool and workpiece, thereby causing some levels of intricacy in the process, which is further influenced by USM process parameters. These parameters are generally controllable input factors affecting the machining conditions. The machining conditions in turn determine the outcome of the machining process assessed by performance measures. The process parameters are presented in Fig. 16 using a cause-and-effect diagram.
Performance measures in USM process include surface integrity, material removal rate (MRR), and tool wear rate (TWR). Surface integrity, as a crucial performance measure in components made of hard and brittle materials, comprises surface roughness, subsurface damage, and heat-affected zone (HAZ). The material removal in USM process involves crack initiation and propagation followed by chip breakage resulting in a rather coarse surface with shallow pits and voids and with subsurface microcracks which are generally not acceptable for delicate structural components. In contrast, HAZ is of lesser concern in USM since the process is considered as a nonthermal process which generates negligible amount of the heat. Thus, the machined surface is generally free from thermally damaged zone or residual stress (Thoe et al. 1998).

MRR is considered as a primary process measure which reflects the efficiency of the machining operation. Besides the volumetric machining rate, drilling speed which is the penetration speed of the tool into the workpiece may also be used to measure the MRR in USM process. MRR might not be constant throughout the machining process due to the variation in the machining zone conditions such as distribution and speed of the free moving particles, indentation force introduced to the abrasive particles, intensity of the cavitation bubble collapse, interparticle collisions, and flushing of the accumulated debris.
Tool wear in the USM process consists of longitudinal wear (WL), side wear (WD), and cavitation wear (Adithan 1974). The length and weight of the tool decrease as a result of the tool wear, thereby causing the tool to vibrate out of the resonant frequency which results in a mismatch with the output frequency of the generator. This leads to reducing amplitude and subsequently a decreased MRR. Furthermore, tool profile may change as a result of tool wear, thereby affecting the dimensional accuracy of the machined features. TWR in USM is influenced by parameters such as particles’ type and size, workpiece and tool material, tool size, and rate and method of slurry delivery to the machining gap (Adithan 1974).

Capabilities and Applications of USM

USM processes are capable of machining a wide range of workpiece materials regardless of their electrical or chemical characteristics. Materials can be categorized into three groups in view of their machinability in USM process. Type I materials such as glasses are very brittle and hence machined easily by the USM process. The material is removed as a result of the propagation and integration of the minute cracks originated from the interactions of abrasive particles with surface material. Type II materials, like hardened steels, exhibit some plastic deformation before fracture, and they can be machined by USM process but with a lower efficiency and MRR as compared to Type I materials. Type III materials such as copper and soft steel are ductile with a high fracture toughness, and hence they are unsuitable for machining by USM technique. The applications of the USM process include:

• Various machining operations such as drilling, die sinking, and contour machining of features with circular and noncircular geometries as well as machining of the complex shapes and 3-D contours
• Machining of ceramics, glasses, silicons, germaniums, quartzes, sapphires, ferrites, optical fibers, and sintered carbides (Type I) as well as hardened stainless steels, hard carbon alloys, and nickel-titanium alloys (Type II)
• Application of USM in micromachining (i.e., termed micro-USM) for microhole drilling, slot machining, and -D microcavities milling

The advantages of the USM process are in the following:

• USM is an efficient and cost-effective technique for precision machining of hard, brittle, and fragile materials especially in medium or small quantities.
• USM is an environment-friendly process, suitable for machining both conductive and nonconductive materials.
• Since actual machining is carried out by abrasive particles, tool materials used can be softer than that of the workpiece material.
• Unlike other thermal-induced processes, USM does not create heat-affected zones, which may induce residual stresses in the machined surface.

Micro-Ultrasonic Machining (Micro-USM)

Micro-USM technique was introduced by Masusawa’s group in the mid-1990s (Sun et al. 1996a). Since then, a number of research works have been carried out to enhance this process in both technological and fundamental aspects. The capabilities of micro-USM technique have been demonstrated for various micromachining operations such as drilling of microholes, machining of straight and spiral grooves, machining of complex-shaped (3-D) microstructures, and generating of microfeatures in MEMS components (Medis and Henderson 2005). These machining operations have been carried out by implementing two modes of micro-USM, namely, die-sinking mode and contouring (scanning) mode.
In die-sinking mode, micro-tool is fabricated with a face containing the patterns similar to features required in workpiece. During the machining, micro-tool is driven toward the abrasive slurry and workpiece, whereby the patterns can be transferred from the tool onto the workpiece. This mode of micro-USM is capable of generating either a single complex microstructure or a pattern of simple microfeatures in one pass of the tool penetration toward the workpiece. In contouring mode, typically a micro-tool with simple shape such as a cylindrical rod is used. The desired microfeature can be machined by guiding the micro-tool along a predetermined tool path. Figure 17 depicts the microfeatures machined on different materials using die-sinking and contouring mode of micro-USM.

Since the inception of micro-USM in the mid 1996s, there have been some developments in the machine system design and configuration including tool preparation and tooling system, tool-workpiece interface monitoring, force measurement and monitoring, workpiece holding and horn design. These advancements have resulted in improvements to process efficiency and machined surface quality in micro-USM.

Comparison of USM and Micro-USM

The process conditions under which material removal takes place in USM and micro-USM are distinctively different due to the diverged process parameter magnitudes employed for micromachining. In order to satisfy the micromachining requirements, the mechanical stress causing crack initiation, propagation, and eventually fracture of the workpiece material is applied to a very small area or volume of the workpiece to reduce the unit removal and to attain low surface roughness (Masuzawa 2000). This is typically achieved by downscaling the particle size with magnitudes of a few hundreds of micrometers used in USM to micrometer or even submicrometer range suitable for micromachining (Egashira and Masuzawa 1999). Another factor that contributes to reduction of the unit removal for micro-USM applications is introducing smaller energies into the slurry medium and consequently to each single particle in the machining zone as compared to that of USM. This leads to a reduced indentation depth and eventually a smaller removal volume per particle impingement in the process. This requirement is achieved by lowering the vibration amplitude to the range of a few micrometers (Zhang et al. 2005). Machining conditions and magnitudes of the process parameters in USM and micro-USM are presented in Table 1.

Further to above conditions desired to satisfy micromachining requirements, and from the standpoint of the feature size magnitudes in micromachining, a tool with dimensions in the micrometer range is required in micro-USM process. This requirement confines using methods such as jet flow and/or suction flushing which are normally employed in USM, and instead it favours the use of tool rotation to facilitate the slurry circulation and debris removal in the machining zone (Egashira and Masuzawa 1999). The flushing conditions in USM are affected by flow rate and flow pressure while the flushing conditions in micro-USM are controlled by the centrifugal forces resulted from the tool rotational speed. Therefore, these factors give rise to variations in conditions of the tool-workpiece interface between USM and micro-USM.
Nevertheless, since micro-USM is an adoption of the conventional USM, both processes still share the same principle for machining. Similar to USM, the material removal in micro-USM is performed bymechanical action of the abrasive particles driven directly or indirectly by either vibrated micro-tool or workpiece as well as by cavitation in the slurry fluid causing the particle indentation, microcrack initiation, propagation, and fracture breakage in the impacted material (Egashira and Masuzawa 1999). This implies that certain process performance results are common for the two processes. Material removal is implemented by multiple particle indentations and impacts, which cause the machined surface to be covered by numerous minute craters with randomly distributed positions. This effectively limits the minimum surface roughness achievable in the process. Also, the machined workpiece surface may contain a layer with many microcracks which is not acceptable for the majority of microstructures and microcomponents (Brinksmeier et al. 1998). Although these surface quality issues are encountered in both USM and micro-USM, their influence is more significant in micro-USM due to relatively smaller size difference between surface defects and machined microfeatures. Moreover, the surface roughness should be reduced in proportion to the feature size as an important factor for micromachining. The ability to control and predict the micro-USM process performance is also more crucial due to miniature tools, small particles, and features size involved in the process.
In addition, there is a need to improve the accuracy and reliability of the micro-USM system, which is considered as an important factor to perform micromachining. Thus, improvements to the micro-USM technique require advancements that provide better understanding of material removal mechanisms and conditions, better control and prediction of process performance, as well as better precision and reliability of the machining system.

Configuration of the Machine System

Two major configurations have been proposed for micro-USM system, namely, tool vibration and workpiece vibration methods. In the former method, the tool is attached to ultrasonic horn and it is oscillated in a vertical direction, while workpiece is fixed onto the XY stage (Sun et al. 1996b). Also, a rotational movement is applied to the horn and attached tool through a spindle system. While workpiece clamping is simple in this type of micro-USM system, implementing the vibration unit into the tooling system to some extent limits the rotational accuracy of the spindle system which is undesirable from the standpoint of the requirements for a high-precision micromachining process (Masuzawa 2000).
In the workpiece vibration method, the oscillation of the tool is removed; thus tooling system is only responsible for guiding and rotating the micro-tool. Applying the vibration to workpiece rather than tool is an attractive method due to its simplicity in the design and operation of the spindle system, precision of the tooling system, effective agitation of the slurry, and efficient delivery of the particles into the machining zone (Egashira and Masuzawa 1999). Furthermore, unlike the micro-USM with tool vibration method, tool wear has no influence on the vibration amplitude of tool tip in workpiece vibration method. Despite these advantages, issues such as reliable workpiece clamping and consistent transmission of the vibrations from horn to the workpiece need to be addressed to improve the process performance.

Tool Preparation and Tooling System

Miniaturization of the tool imposes some constraints on implementation of the micro-USM process, particularly in the fabrication of tiny tools and configuration of tooling system (Zhang et al. 2005). Mounting and aligning of the micro-tools are key issues, which have a direct impact on process productivity and part accuracy. Various tooling strategies and tool preparation techniques such as on-machine tool preparation using wire electrical discharge grinding (WEDG) and ultrasonic-assisted grinding, utilizing the sintered diamond tools, and fabrication of the multi-tools have been proposed to enhance the capability of the micro-USM process. Also, rotation of the micro-tool has been reported to reduce the form error and to improve the machining rate (Egashira and Masuzawa 1999). However, tool rotation has the adverse effect of enlargement of the microhole diameter mainly due to the spindle runout. While improvements are made to rotational and linear movements of the micro-tool, the flexibility of the tooling system also should be considered to accommodate the tool preparation techniques.

Monitoring and Control of the Machining Force

During material removal in micro-USM process, the micro-tool progresses further toward the workpiece through abrasive slurry to maintain a certain machining gap and thus a stable machining condition. While a machining gap larger than an optimum level would cause the particles to lose their effective kinetic energy before striking the workpiece surface, using a very small machining gap might result in suppressing the abrasive particles by the micro-tool and hence leaving them with insufficient energy for impacting the workpiece surface. In both conditions, the material removal rate would be very low or even negligible. Furthermore, after the onset of the machining process, the interface boundary between tool and workpiece as well as the conditions of machining zone vary continuously, and hence it is difficult to maintain the machining quality (Zhang et al. 2005). The control of the machining gap in micro-USM is made possible through precision measurement and control of the static load (machining force) as the feedback in the process.
While in USM, static load and tool size are in the range of a few kilogram-force and larger than 1 mm, respectively; in micro-USM process the static load is in order of a few gram-force and tool size ranges within 500 μm (Zhang et al. 2005). Besides, mismatched tool-workpiece interactions not only adversely affect the stability of the micro-USM process, but they also lead to the breakage of microtools. About 60 % of the micro-tool breakages have been reported to occur due to mishandling of the initial engagement of tool-workpiece interaction (Zhang et al. 2006). Therefore, downscaling of the process parameters from USM to micro-USM necessitates the development of a reliable force sensing and monitoring system with high precision and rapid response in order to enhance the stability of the micromachining process and to avoid tool breakages.
Different strategies have been proposed to control the machining gap and contact force between micro-tool and abrasive slurry in micro-USM process. Acoustic emission (AE) technique was proposed and implemented to monitor the tool-workpiece interaction in order to reduce the probability of the tool breakage in the process (Moronuki and Brinksmeier 2002). However, the AE output signal has been reported to be influenced by the workpiece position on the machine table as well as the intrinsic characteristics of the ultrasonic generator used in the process, thereby causing an incorrect assessment on actual state of the tool-workpiece interaction and hence incomplete machining (Zhang et al. 2006).
Other methods such as using electronic balance and dynamometer also have been proposed for measuring the machining force in micro-USM. A digital balance with resolution of 10 mg and response time of 10 ms was proposed in (Sun et al. 1996a). Also, an electronic balance with a minimum index of 1 mg was installed in a micro-USM experimental setup acting as the sensor for feedback control (Egashira and Masuzawa 1999). However, low-frequency response and small sampling rate were identified as major limitations of this method. In another study, a precision force actuation and measurement system with the capability of sensing and control of forces in the low mN range was developed (Hoover and Kremer 2007). Nevertheless, the force measurement in this system only provides the desired accuracy when employed for measurements in the horizontal direction rather than vertical one. Thus, it may not be suitable for ultrasonic machining process.

Workpiece Holding Method

The machining accuracy is directly affected by the workpiece holding mechanism in micro-USM with workpiece vibration method. In such a setup, the workpiece must be attached firmly to the ultrasonic horn, which otherwise mechanical vibrations would not be transmitted to the workpiece consistently, causing unstable machining conditions and excessive heat generation in the horn-workpiece interface.
The use of double-sided adhesives to attach the workpiece to the transducer or horn has been attempted by researchers (Hu 2007). However, this method has two main drawbacks. The removal of workpiece after machining is difficult, especially in the case of thin and fragile workpiece materials. Also, the adhesive tape itself may absorb the ultrasonic power and alter the acoustic characteristics of the original mechanical vibration supplied into the ultrasonic horn. Therefore, a reliable clamping system is deemed necessary to attain a uniform transmission of the mechanical vibrations from ultrasonic horn to the workpiece.

Development of a Micro-USM System with New Techniques for Workpiece Clamping and Static Force Control

A new micro-USM system was designed and fabricated with the purpose of improving the functionality and flexibility of the machine system (Zarepour et al. 2011). Effective and well-proved techniques such as applying the vibration to workpiece are incorporated in developed system. Figure 18 illustrates the schematic diagram of the developed micro-USM system which consists of five main subsystems, namely, generation and transmission of ultrasonic vibrations, tooling system, workpiece holding, measurement and control of static load, and slurry delivery system.
The tool rotation has been adopted to reduce form errors and to provide flexibility in the tooling system. The whole tooling system is mounted on a three-axis stage. The micro-tool motion is controlled with a minimum incremental motion of 0.03 μm in z-axis and 10 μm in both x-axis and y-axis. An integrated piezo-motor controller/driver with computer interface is used in z-axis, enabling a highly reliable tool positioning with 30 nm sensitivity. This configuration provides an accurate control over the infeed of the tool in the direction of the workpiece oscillation in order to maintain a preset machining gap.

Workpiece Clamping System

The task of holding the workpiece against the ultrasonic horn hampers the efficient use of the micro-USM with workpiece vibration method. The vibration frequency of the horn is 50 kHz in the developed system. As such, attachment of the workpiece to the horn is a critical stage in the process, which may affect the stability and accuracy of the machining process. The micro-tool used has a diameter up to 300 μm, which is susceptible to breakage in the event of tool collision.
An aluminum booster together with a full-wave titanium horn, which has a recess to accommodate the workpiece, is used to transmit the vibration from ultrasonic transducer to the workpiece. Nominal vibration amplitude of the system ranges between 0.8 and 5.5 μm using a combination of the reverse booster and horn with the same gain ratio of 0.5:1 and adjustment of the output power of the ultrasonic generator. As shown Fig. 19, a vacuum chuck is incorporated to the ultrasonic horn for workpiece clamping. It is important to introduce the vacuum tube to the ultrasonic horn at the position of the nodal point where the amplitude of vibration is close to zero. The vacuum tube is connected to the vacuum pump through a liquid separator with filter. The proposed method enables a rapid clamping and unclamping of the workpiece without introducing crack to thin and fragile workpieces.

Measurement and Control of the Static Load

In the developed system, the static load between micro-tool and vibrated workpiece is measured and controlled using a precision force sensor integrated with the tooling system as illustrated in Fig. 20. The proposed design has a merit in that the sensor can be mounted on tooling system without hampering the rotation of the micro-tool. Furthermore, utilizing the sensor in the tool side eliminates the measurement errors arising from dead weight of the ultrasonic stack and fixture as well as the noise and vibration from ultrasonic horn. The sampling frequency of the sensor is up to 1 kHz which can be utilized in both tension and compression; hence, it is capable of highspeed measurement of the variation in the machining force. The static load can be constantly monitored by the force sensor in order to maintain the preset machining gap through controlling the infeed tool motion. The system also has a built-in overload protection feature.

Material Removal Mechanisms and Modes in USM and Micro-USM

The mechanisms of material removal in USM process can be categorized into four types, namely, direct hammering, free impact, cavitation erosion, and chemical action. Detailed descriptions of these mechanisms are presented in Table 2. One material removal mechanism or a combination of mechanisms will dominate the material removal process depending on the process parameters (Thoe et al. 1998). While direct hammering and free impact are primarily responsible for material removal in USM, cavitation erosion and chemical action are of secondary significance with the majority of the workpiece materials such as glass, ceramics, and hardened steel (Thoe et al. 1998). Cavitation erosion has been reported to play an important part in machining of the porous materials (Weller 1984).

The effect of these mechanisms results in material removal from the workpiece either by crack formation and fracture in brittle mode or by cutting and shearing in ductile mode. In addition, during the USM process, there might be a situation where the material at transient surface of the workpiece is not completely removed but only displaced upon impingement of the abrasive particles.
When material is impacted by a hard angular particle at high speeds or under large contact forces, plastic deformation occurs in the contact zone due to high compressive and shear stresses, and a radial crack is formed. After impact, the plastic deformation results in large tensile stresses leading to initiation of lateral cracks and eventually material removal (Wensink and Elwenspoek 2002). When the speed of particles or the amount of force acting on them during the process is below the threshold values required for crack initiation in the brittle material, particles may cause only plastic deformation to the surface, which brings about material removal by ploughing and cutting in a ductile mode instead of a brittle one. This change in erosion mode from brittle fracture-dominated behavior to plastically dominated behavior is called brittle-ductile transition (Wensink and Elwenspoek 2002). The material removal mode, whether it occurs in brittle or ductile mode, and transition between these modes have a direct effect on both surface integrity and material removal rate associated with process quality and productivity, respectively.
Although the underlying machining principle of micro-USM is similar to that of conventional USM (Zhang et al. 2005), downscaling of the process parameters such as tool size, vibration amplitude, machining load, and particles size so as to minimize the contact zone of abrasive particles with tool and workpiece to the microscale range may affect the machining condition. Also, the nature of material removal in micro-USM, whether the material is removed through ductile deformation or crack generation followed by brittle fracture, may differ from that in USM due to the influence of miniaturized tool-abrasive-workpiece interaction intensity (Zhang et al. 2006). For instance, the accumulation of workpiece debris and crushed abrasives in a small machining gap and difficulty in removing them from the machining zone due to susceptibility of the micro-tool to deformation and breakage will affect the manner by which material is removed from the workpiece surface (Yu et al. 2006). This also could lead to further complexity in tool-abrasive-workpiece interactions and consequently more intricate material removal mechanisms in micro-USM. Research works concerning the material removal modes in micro-USM are very limited in the literature, while no studies have been reported regarding the investigation on dominant material removal mechanisms that contribute to the machining process. 

Intricacies Involved in Study of Material Removal Mechanisms in USM and Micro-USM

Characteristics of the machined surface are dependent on the manner by which the material is removed from the workpiece. The predominant mechanism of material removal has a direct influence on material removal rate and machined surface quality. The study of different conditions in tool-abrasive-workpiece interaction within the slurry fluid is complex due to the following issues:

• Ultrasonic machining is a free abrasive machining process. That is, abrasive particles move randomly and freely within the machining zone during the material removal process. Thus, the location of each particle is changing by instant displacement and rotation within the slurry.
• Generally, the shape of the particles is irregular and with a distributed size over the machining zone. That is, the real particles engaged in machining operation do not have exactly the same size, but they follow a nominal average size. Besides, abrasive particles are not ideally rigid and incompressible; thus, they break into smaller pieces due to collision with the workpiece and tool or between themselves resulting in variation of the abrasive size and distribution across the machining gap.
• The distribution of the abrasive particles in the affinity of the machined surface and tool may become uneven as a result of slurry flow and agitation due to micro-explosion of the bubbles inside the slurry fluid. Also, there is a likelihood that the conditions of the machining gap become abnormal due to embedding of the particles into the workpiece or tool surface (Yu et al. 2006). These factors may lead to a situation whereby the mechanical vibration would not be transmitted to the slurry fluid and abrasive particles effectively.

An eclectic of these factors introduces micro-USM as a stochastic process. Further, as discussed earlier in this chapter, the effective material removal mechanisms may be different in the case of micro-USM as compared to that of conventional USM due to more complexity of the tool-abrasive-workpiece interaction in the machining zone.

Investigation on Material Removal Mechanisms in Micro-USM

Experimental observations by authors (the results to be published in a separate paper) demonstrated that the material removal in micro-USM is primarily caused by the erosive wear associated with ultrasonic vibration rather than by abrasive wear resulting from micro-tool rotation during the process. The erosive wear in micro-USM mainly takes place via three material removal mechanisms similar to those of conventional USM as described earlier in this chapter. These mechanisms include pure cavitation due to bubble implosion inside the slurry liquid, direct hammering of abrasive particles on the workpiece surface by micro-tool, and impact of the free moving abrasive particles. While the role of pure cavitation has been reported to be insignificant in USM, direct hammering and free particle impact are regarded as main removal mechanisms. Also, free particle impact has been confirmed to account for only about 22 % of total material removal in USM (Khairy 1990). Thus, direct hammering has been identified as the primary cause of the material removal in conventional USM. The relative contribution of the removal mechanisms might be different in micro-USM as compared to that of conventional USM due to the diverse removal conditions inside the machining gap for two processes. As such, a basic study is conducted to identify the contribution of the different mechanisms to material removal in micro-USM. These mechanisms are depicted in Fig. 21.

Material Removal by Pure Cavitation

To identify the role of pure cavitation in material removal, a set of machining experiments were conducted by using deionized water and without any abrasive particles. Tungsten rods with a diameter of 300 μm and monocrystalline silicon with thickness of 525 μm were used as micro-tool and workpiece, respectively. The vibration frequency of 50 kHz and vibration amplitude of 2 μm were used for all experiments. A machining duration of 5 min was maintained for all experimental runs, and each run was repeated three times. The force introduced by ultrasonic vibration was recorded using the precision force sensor with interface software. This force value corresponds to the adjusted gap distance, and it is used as the feedback to maintain the gap distance during the process. Figure 22 shows the profile of a typical microhole with maximum depth of 6.8 μm obtained at gap distance of 2 μm.

Material Removal by Particle Direct Hammering

Another set of experiments was planned and conducted to investigate the role of direct hammering mechanism in material removal in the micro-USM process. PCD particles with nominal size of 3 μm mixed with DI water were utilized as abrasive slurry. Vibration amplitude was set at 2 μm, and small machining gaps in the range of 0.2–2 μm were applied. This is to ensure that abrasive particles can be pushed and impacted directly by end face of the micro-tool onto the workpiece surface, that is, the condition required for direct hammering mechanism to take place in micro-USM process. Other machining conditions were similar to that described for the experiments in the previous experiment.
The depth of microholes was measured, and maximum machined depth of 28.4 μm was obtained when machining gap was adjusted at 0.2 μm. The profile of a typical microhole machined under these process conditions is depicted in Fig. 23.

Material Removal by Free Particle Impact

Machining experiments were conducted to investigate the effect of free particle impact in the material removal process. PCD particles with nominal size of 3 μm mixed with DI water were utilized as abrasive slurry. Vibration amplitude was set at 2 μm, and machining gap was adjusted in the range of 5–200 μm. Since the machining gap was set to be larger than the particles size, direct hammering of the particles by tool face was unlikely to occur in the process. Therefore, material removal was brought about primarily by the impact of the free moving particles striking the workpiece surface.
The depth of microholes machined at various gap distances was measured and analyzed. As shown in Fig. 24, a microhole with a depth of approximately 218.2 μm was obtained, which corresponds to the microhole machined at 30 μm gap distance.

Contribution of Different Material Removal Mechanisms in Micro-USM

A comparison of the three removal mechanisms in micro-USM with respect to their contributions to material removal is presented in Fig. 25. Average machining depth was calculated for three machining runs at which the maximum depth was achieved under different mechanisms in the process. While pure cavitation produced a shallow microhole with an average depth of only 6.8 μm during 5 min of the machining process, direct hammering and free particle impact produced microholes with an average depth of 28.4 μm and 218.2 μm, respectively, over the same machining time. Accordingly, the depth of the microholes produced under the free particle impact mechanism is approximately 32 times and 8 times higher than that of pure cavitation and direct hammering, respectively. This could be attributed to high velocity of the free moving particles inside the machining gap which are subjected to ultrasound in micro-USM. Results of this study evidence that unlike conventional USM, free particle impact is the main contributor to material removal in micro-USM.

Predictive Modeling of Ductile and Brittle Removal Modes in Micro-USM

The results of the investigation on abrasive particle erosion of some brittle materials have demonstrated that the ductile-brittle transition may occur as the size of particles is reduced (Hutchings 1992). Since the size of abrasive particles used in micro-USM is smaller than that of conventional USM by a factor of 10 or more, the likelihood of such a ductile-brittle transition is higher in micro-USM as compared to that of conventional USM. Being a loose abrasive process, the material removal mode in micro-USM can be influenced by setting the process parameters including particle size at desired levels in order to enhance the process performance.
Investigations on material removal characteristics have been carried out based on the morphology of surfaces, processed by multiple-particle impingements. Unlike the multiple-particle impact, the method of single-particle impact has a potential to provide more basic and explicit information about material removal modes and mechanisms. Investigation on craters produced by single-particle impact is an attractive approach with the capability of providing practical solutions to overcome the ambiguity around the issue of determining the material removal modes in abrasive-based micromachining processes. Therefore, the study of single abrasive particle impingement may open new avenues toward a more fundamental and objective approach to investigate the brittle and ductile machining modes and transition between them in micro-USM process.

Approach to Development of the Predictive Model

Figure 26 outlines a model for prediction of material removal mode in micro-USM process. The model consists of three parts: estimation of the apparent threshold kinetic energy for radial and lateral fracture in workpiece material, estimation of the kinetic energy of an impinging particle inside the machining zone, and criteria for determining the material removal mode. The analysis to estimate the threshold kinetic energy is performed based on the well-established indentation fracture theory for microcracks initiation by hard angular particles in work materials. The application of this theory to predict the thresholds for ductile and brittle transitions in other processes such as particle erosion have been described by Hutchings (Hutchings 1992).
The model estimates the kinetic energy of a single impinging abrasive particle driven by the ultrasound in the process. The primacy on the type of material removal mode lies at the foundation of comparative principle between the apparent threshold kinetic energy of the workpiece material and estimated kinetic energy of the impacting particle based on the input parameters of micro-USM process.