Glass, with its valuable properties, finds extensive use in aerospace, optics, and biomedical fields. Owing to its low fracture toughness, glass typically fractures in a brittle manner during machining, resulting in poor surface quality. This paper presents an experimental investigation of vibration-assisted machining (VAM) techniques to enhance the machining of glass materials. A novel high-frequency two-dimensional VAM system specifically designed for glass is introduced, and slot milling experiments are conducted using ultrasonic high-frequency vibrations. A low-frequency nonresonant VAM system is also employed for comparison purposes. A comprehensive examination is made of the effects of various machining parameters, such as feed rate, cutting speeds, and vibration parameters, including vibration modes and amplitudes, on the machining performance of glass. Surface roughness, edge chipping generation, and tool wear are thoroughly characterized using scanning electron microscopy. The findings demonstrate that under specific machining and vibration parameters, the proposed ultrasonic vibration-assisted micro-milling (UVAMM) system can achieve a nanometric surface roughness Ra for glass. The UVAMM system offers enhanced surface quality, improved edge quality, and reduced tool wear compared with conventional machining techniques. This study provides valuable insights and directions for the application of 2D VAM systems in achieving superior machining results for glass components at small scales with nanometric surface finishes.

  • A novel two-dimensional piezoelectric ultrasonic transducer incorporating longitudinally polarized piezoceramics is designed and fabricated.

  • A significant improvements in surface roughness and a reduction in edge chipping are achieved using vibration-assisted micro-milling.

  • A decrease in tool wear is demonstrated with vibration-assisted micro-milling compared with conventional micro-milling.

Glass is a key industrial material used in many fields, including communications, optics, biomedicine, and transportation. It has distinctive properties as an amorphous, brittle material. The structural features of glass are influenced by its atomic arrangement, which displays both long-range disorder and short-range order. Glass exhibits extraordinary qualities such as homogeneity, superior corrosion and chemical resistance, high electrical resistivity, and significant hardness, which make it very attractive for a variety of applications. Glass is a versatile material with broad industrial significance owing to its unique composition and physical characteristics.1 In the fabrication of glass components with microfeatures for specific applications, such as DNA microarrays used in DNA analysis, time-consuming and possibly dangerous photolithographic and etching methods are frequently required.2 For rapid prototyping applications in particular, there is a need for more effective fabrication techniques for glass-based devices, one of which is mechanical micromachining. However, because glass is naturally fragile, its machining can be difficult, and broken surfaces may be produced that need additional expensive and time-consuming polishing methods. Fortunately, studies have revealed that glass can exhibit a ductile regime during machining under precisely regulated cutting conditions, providing opportunities for more effective and efficient machining techniques.3,4

In particular, a wide range of outstanding qualities is displayed by borosilicate glass, including high chemical resistance, low thermal and electrical conductivity, transparency, superior optical qualities, improved strength, high specific strength, heat resistance, high corrosion resistance, excellent mechanical hardness, exceptional anodic bonding, hydrophilicity, anisotropic properties, and good surface quality. Consequently, borosilicate glass is highly sought after and frequently used in a variety of engineering and healthcare applications, including micro-opto-electro-mechanical systems, optical telecommunications, manufacture of spectacle lenses and optical instruments, miniaturization of microfluidic devices for chemical and biological micro total analysis systems, mechanical inertial sensors, and microfabricated devices such as oxide fuel cells and micropumps. However, like other glasses, the high intrinsic hardness and brittleness of borosilicate glass make it difficult to machine with both standard and non-traditional machining methods. This has prompted the investigation of effective and efficient alternative machining processes, and as these could open up new opportunities for the use of borosilicate glass in many applications, their development is a matter of intense research interest.5 

Advances in machining technology have opened up possibilities for utilizing mechanical cutting processes for diverse engineering materials, including glasses. Studies of micro-milling of brittle materials have revealed that under specific cutting conditions, these materials can undergo a transition from a brittle to a ductile cutting regime. This transition results in the production of a smoother surface with minimal occurrence of minor cracks. By contrast, when machining is performed in the brittle mode, it leads to the formation of numerous cracks and compromises surface quality. However, despite extensive research in this area, the precise conditions under which the transition to the ductile cutting regime occurs remain uncertain. Further investigation is necessary to establish a comprehensive understanding of the factors influencing the occurrence of ductile cutting in brittle materials.6 

Recent developments in machining techniques have further expanded the potential for efficient glass machining. For instance, the application of vibration and heat-assisted machining to aerospace alloys has led to significant improvements in machining performance.7 However, owing to its fragility and low fracture toughness, borosilicate glass presents a difficult micro-milling challenge. It is vulnerable to crack initiation and spread because of the high tensile stress generated by cutting tools.8 The dimensional accuracy, mechanical strength, and surface quality of machined items can all be negatively impacted by cracks. Ductile mode machining, in which the material experiences plastic deformation rather than fracture at extremely small depths of cut, is necessary to prevent crack formation. Nevertheless, obtaining ductile mode machining is dependent on a number of variables, including tool geometry, cutting speed, feed rate, and tool wear.9 The great hardness and poor flexibility of borosilicate glass make it difficult to micro-mill to a smooth surface finish. The tool edge radius, undeformed chip thickness, cutting pressure, and tool vibration all have an impact on surface roughness.8 

By periodically separating the uncut workpiece from the tool, vibration-assisted machining (VAM) can enhance the micro-milling of borosilicate glass while reducing cutting forces, tool wear, surface roughness, and fracture development. Additionally, VAM can speed up material removal and improve ductile mode machining of borosilicate glass. However, owing to several challenges, VAM has not been widely employed for micro-milling borosilicate glass. It is desirable to develop an effective vibration device that can produce stable, regulated vibration in a specified frequency range and direction, with the vibration amplitude, frequency, direction, phase angle, duty cycle, and modulation being adjustable for various materials, tools, and machining conditions.8–11 Unfortunately, there remains a lack of knowledge regarding ultrasonic vibration-assisted micro-milling of borosilicate glass, whether this be achieved through applying vibrations to the workpiece or the tool. Notably, no research has been published on the effects of micro-milling monocrystalline silicon utilizing high-frequency ultrasonic vibrations.

In the present study, to apply ultrasonic vibrations to a workpiece, a unique two-dimensional (2D) resonant ultrasonic transducer is developed and tested.12 Additionally, the results obtained are compared with those of a one-dimensional (1D) ultrasonic VAM system that was also developed by the present authors, and, to assess the effects of resonance, with those of a 2D nonresonant VAM system developed by Zheng et al.13 A universal ball-based structure is developed to support the workpiece holder and to avoid any chatter vibrations during the experimental investigation of the UVAMM process compared with conventional milling. The effects of various machining parameters, such as feed rate, cutting speeds, and vibration parameters, including vibration modes and amplitudes, on the machining performance of glass are comprehensively examined. Surface roughness, edge chipping generation, and tool wear are thoroughly characterized using scanning electron microscopy (SEM). This investigation provides valuable insights into how superior machining results can be achieved for glass components at small scales with nanometric surface finishes.

The remainder of the paper is structured as follows. Section II details the principle of operation of the 2D front mass device. Section III describes the methodology and experimental setup. Section IV presents the results and a discussion, focusing on surface roughness, edge chipping, and tool wear. Finally, Sec. V concludes the study with a summary of key findings and potential directions for future work.

At the front face of the front mass, a combination of longitudinal vibration and phase-shifted bending vibration produces an elliptical trajectory. The displacement in the x direction of the center point of the beam cross-section in the bending vibrational mode and the volumetric deformation imposed by the beam’s expansion and contraction are ignored to facilitate computation of the elliptical trajectory at the point P. Figure 1 depicts the front mass in the bending vibrational mode.

FIG. 1.

Bending vibrational mode of the front face of a 2D horn.

FIG. 1.

Bending vibrational mode of the front face of a 2D horn.

Close modal
The angle θ between the bending vibrational displacement and the x axis is given by
(1)
When the bending resonance mode is excited, the point P along the LB direction undergoes simple harmonic vibrations, and the bending vibrational amplitude can be calculated as follows:
(2)
where the turning angle of the end face of the horn is given by
(3)
Ly is the displacement of point P along the y direction caused by bending and shear deformation, and R is the radius of the beam.
When the beam is excited at its resonant frequency ω, it undergoes both longitudinal and bending vibrations simultaneously. The displacement of point O can be written as the sum of the longitudinal and bending components:
(4)
(5)
where Dxlongitudinal is the displacement of point P due to longitudinal vibration, and Dxbending and Dybending are the displacements of point P due to bending and shear deformation. For the longitudinal vibration, we can write
(6)
where α is the phase shift between the longitudinal and bending vibrations, and Lx is the longitudinal vibrational amplitude. For the bending and shear deformations, we can write
(7)
(8)
where the angle θ between the bending vibrational displacement and the x axis is given by Eq. (1).
Adding the longitudinal and bending components together, we obtain the displacement equations:
(9)
(10)
where Dx and Dy are the vibration displacements of the point P in the x and y directions, respectively.

Figure 2 shows the material domain and mesh geometry of the transducer developed by the present authors.12 

FIG. 2.

(a) Material domains. (b) Mesh geometry.

FIG. 2.

(a) Material domains. (b) Mesh geometry.

Close modal

The surface quality of borosilicate glass is significantly influenced by the occurrence of cracks during the machining process. To improve surface preparation, it is crucial to understand the mechanisms behind crack formation. Previous research has indicated that cracks tend to form in the brittle regime, particularly in regions with high pressure generated by the tooltip in the chip formation zone. These cracks can propagate to the final machined surface or be eliminated through subsequent cutting. Perveen and Molardi14 have suggested that controlling cutting conditions to reduce stress formation ahead of the cutting edge can help minimize such cracks.

Another potential source of crack formation in borosilicate glass occurs on the machined bottom surface. The cutting edge applies a localized load on this surface, and once the tool has passed, the applied load is released. However, residual stress at the boundary can initiate lateral cracks. Similarly to the cracks formed in the chip formation zone, these cracks can either propagate to the final machined surface or be eliminated by subsequent cuts. It is important to consider these factors and control the cutting parameters to mitigate crack formation and enhance the surface quality of borosilicate glass during machining processes.

In this study, an in-depth exploration was conducted of two distinct approaches: resonant mode ultrasonic VAM and nonresonant-mode VAM. Each of these methods were employed with two different types of tools: a diamond-coated tool and a solid carbide tool. The aim was to assess their effectiveness across a broad spectrum of machining parameters.

The resonant mode ultrasonic VAM technique shown in Fig. 3(a), characterized by its vibrational resonance, was subjected to a careful experimental investigation. In this machining approach, the aim of subjecting the workpiece to controlled ultrasonic vibrations is to enhance material removal and improve overall machining efficiency. Various combinations of machining parameters, including tool feed rate, spindle speed, and ultrasonic amplitude, were varied to gain a comprehensive understanding of their influence on the machining process.

FIG. 3.

(a) Resonant mode ultrasonic VAM setup. (b) Nonresonant mode VAM setup. (c) Different vibrating directions of the workpiece.

FIG. 3.

(a) Resonant mode ultrasonic VAM setup. (b) Nonresonant mode VAM setup. (c) Different vibrating directions of the workpiece.

Close modal

The nonresonant mode VAM method shown in Fig. 3(b) was also explored in parallel. This technique involves the application of controlled vibrations to the workpiece, imparted by a diamond-coated or solid carbide tool. Unlike resonant mode ultrasonic VAM, nonresonant mode VAM does not rely on vibrational resonance for improved material removal. Instead, it exploits the effects of vibration on chip formation, cutting forces, and surface quality. The machining parameters under investigation for this approach encompassed tool engagement, depth of cut, cutting speed, and vibration frequency.

Throughout the study, meticulous experiments were conducted using various combinations of machining parameters, to ensure that a wide range of operating conditions were explored. This comprehensive investigation allowed a detailed evaluation of the performance and effectiveness of both resonant mode ultrasonic and nonresonant mode VAM, utilizing both diamond-coated and solid carbide tools. By considering wide ranges of machining parameters, the study sought to provide valuable insights into the capabilities and limitations of these techniques.

To test the machinability of borosilicate glass, many micro-milling tests of each milling type were carried out utilizing conventional micro-milling, 2D and 1D ultrasonic resonant mode VAM, and 2D nonresonant mode VAM systems, as indicated in Table I and with the combination of vibrations shown in Fig. 3(c). The process parameters were determined on the basis of preliminary experiments and a review of the existing literature. Factors such as feed rate, spindle speed, and vibration amplitude were varied within ranges identified as optimal for minimizing surface roughness and tool wear in similar machining contexts. Specific levels were chosen to explore both the lower and upper bounds of the parameters’ effectiveness. The experimental parameters were determined through a combination of theoretical analysis and empirical testing. Initial parameter settings were based on established norms for micro-milling of brittle materials, with subsequent adjustments made based on observed machining performance. This iterative process ensured the identification of conditions that best facilitated ductile mode machining and minimized crack formation.

TABLE I.

Machining and vibration parameters.

Slot numberSpindle speed (rpm)Feed rate (mm/min)Vibration amplitude (µm)ToolaMachining typebVAM frequency (kHz)Surface roughness Ra
30 000 DC CM ⋯ 0.405 
30 000 DC CM ⋯ 0.396 
30 000 10 DC CM ⋯ 0.46 
30 000 10 DC CM ⋯ 0.451 
30 000 15 DC CM ⋯ 0.51 
30 000 15 DC CM ⋯ 0.503 
50 000 SC CM ⋯ 0.37 
50 000 SC CM ⋯ 0.376 
50 000 10 SC CM ⋯ 0.372 
10 50 000 10 SC CM ⋯ 0.376 
11 50 000 15 SC CM ⋯ 0.36 
12 50 000 15 SC CM ⋯ 0.361 
13 30 000 DC CM ⋯ 0.855 
14 30 000 DC CM ⋯ 0.851 
15 30 000 10 DC CM ⋯ 0.96 
16 30 000 10 DC CM ⋯ 0.965 
17 30 000 15 DC CM ⋯ 1.862 
18 30 000 15 DC CM ⋯ 1.766 
19 50 000 SC CM ⋯ 0.422 
20 50 000 SC CM ⋯ 0.42 
21 50 000 10 SC CM ⋯ 0.456 
22 50 000 10 SC CM ⋯ 0.46 
23 50 000 15 SC CM ⋯ 0.455 
24 50 000 15 SC CM ⋯ 0.452 
25 30 000 0.5 DC 1D VAM 27.22 0.29 
26 30 000 10 0.5 DC 1D VAM 27.22 0.255 
27 30 000 15 0.5 DC 1D VAM 27.22 0.327 
28 50 000 0.5 DC 1D VAM 27.22 0.257 
29 50 000 10 0.5 DC 1D VAM 27.22 0.293 
30 50 000 15 0.5 DC 1D VAM 27.22 0.26 
31 30 000 SC 1D VAM 27.22 0.151 
32 30 000 10 SC 1D VAM 27.22 0.104 
33 30 000 15 SC 1D VAM 27.22 0.129 
34 50 000 SC 1D VAM 27.22 0.102 
35 50 000 10 SC 1D VAM 27.22 0.122 
36 50 000 15 SC 1D VAM 27.22 0.109 
37 30 000 0.5 DC 1D VAM 27.22 1.213 
38 30 000 10 0.5 DC 1D VAM 27.22 1.419 
39 30 000 15 0.5 DC 1D VAM 27.22 1.119 
40 50 000 0.5 DC 1D VAM 27.22 1.23 
41 50 000 10 0.5 DC 1D VAM 27.22 1.322 
42 50 000 15 0.5 DC 1D VAM 27.22 1.358 
43 30 000 SC 1D VAM 27.22 1.713 
44 30 000 10 SC 1D VAM 27.22 1.711 
45 30 000 15 SC 1D VAM 27.22 1.525 
46 50 000 SC 1D VAM 27.22 1.42 
47 50 000 10 SC 1D VAM 27.22 1.851 
48 50 000 15 SC 1D VAM 27.22 1.82 
49 30 000 0.5 DC 2D Non-R 0.179 
50 30 000 10 0.5 DC 2D Non-R 0.17 
51 30 000 15 0.5 DC 2D Non-R 0.144 
52 50 000 0.5 DC 2D Non-R 0.166 
53 50 000 10 0.5 DC 2D Non-R 0.198 
54 50 000 15 0.5 DC 2D Non-R 0.152 
55 30 000 SC 2D Non-R 0.188 
56 30 000 10 SC 2D Non-R 0.152 
57 30 000 15 SC 2D Non-R 0.186 
58 50 000 SC 2D Non-R 0.166 
59 50 000 10 SC 2D Non-R 0.181 
60 50 000 15 SC 2D Non-R 0.168 
61 30 000 0.5 DC 2D Non-R 0.956 
62 30 000 10 0.5 DC 2D Non-R 1.361 
63 30 000 15 0.5 DC 2D Non-R 0.91 
64 50 000 0.5 DC 2D Non-R 1.244 
65 50 000 10 0.5 DC 2D Non-R 0.948 
66 50 000 15 0.5 DC 2D Non-R 1.166 
67 30 000 SC 2D Non-R 0.842 
68 30 000 10 SC 2D Non-R 0.323 
69 30 000 15 SC 2D Non-R 0.789 
70 50 000 SC 2D Non-R 0.295 
71 50 000 10 SC 2D Non-R 0.411 
72 50 000 15 SC 2D Non-R 0.318 
73 30 000 0.5 DC 2D FC 29.81 0.079 
74 30 000 10 0.5 DC 2D FC 29.81 0.016 
75 30 000 15 0.5 DC 2D FC 29.81 0.084 
76 50 000 0.5 DC 2D FC 29.81 0.018 
77 50 000 10 0.5 DC 2D FC 29.81 0.089 
78 50 000 15 0.5 DC 2D FC 29.81 0.018 
79 30 000 SC 2D FC 29.81 0.042 
80 30 000 10 SC 2D FC 29.81 0.013 
81 30 000 15 SC 2D FC 29.81 0.049 
82 50 000 SC 2D FC 29.81 0.021 
83 50 000 10 SC 2D FC 29.81 0.054 
84 50 000 15 SC 2D FC 29.81 0.026 
85 30 000 0.5 DC 2D FC 29.81 0.163 
86 30 000 10 0.5 DC 2D FC 29.81 0.164 
87 30 000 15 0.5 DC 2D FC 29.81 0.175 
88 50 000 0.5 DC 2D FC 29.81 0.185 
89 50 000 10 0.5 DC 2D FC 29.81 0.182 
90 50 000 15 0.5 DC 2D FC 29.81 0.188 
91 30 000 SC 2D FC 29.81 0.168 
92 30 000 10 SC 2D FC 29.81 0.122 
93 30 000 15 SC 2D FC 29.81 0.177 
94 50 000 SC 2D FC 29.81 0.141 
95 50 000 10 SC 2D FC 29.81 0.186 
96 50 000 15 SC 2D FC 29.81 0.142 
97 30 000 0.5 DC 2D FA 29.81 0.068 
98 30 000 10 0.5 DC 2D FA 29.81 0.036 
99 30 000 15 0.5 DC 2D FA 29.81 0.076 
100 50 000 0.5 DC 2D FA 29.81 0.043 
101 50 000 10 0.5 DC 2D FA 29.81 0.128 
102 50 000 15 0.5 DC 2D FA 29.81 0.048 
103 30 000 SC 2D FA 29.81 0.0642 
104 30 000 10 SC 2D FA 29.81 0.032 
105 30 000 15 SC 2D FA 29.81 0.0864 
106 50 000 SC 2D FA 29.81 0.041 
107 50 000 10 SC 2D FA 29.81 0.089 
108 50 000 15 SC 2D FA 29.81 0.038 
109 30 000 0.5 DC 2D FA 29.81 0.163 
110 30 000 10 0.5 DC 2D FA 29.81 0.164 
111 30 000 15 0.5 DC 2D FA 29.81 0.175 
112 50 000 0.5 DC 2D FA 29.81 0.185 
113 50 000 10 0.5 DC 2D FA 29.81 0.182 
114 50 000 15 0.5 DC 2D FA 29.81 0.188 
115 30 000 SC 2D FA 29.81 0.168 
116 30 000 10 SC 2D FA 29.81 0.122 
117 30 000 15 SC 2D FA 29.81 0.177 
118 50 000 SC 2D FA 29.81 0.141 
119 50 000 10 SC 2D FA 29.81 0.186 
120 50 000 15 SC 2D FA 29.81 0.142 
Slot numberSpindle speed (rpm)Feed rate (mm/min)Vibration amplitude (µm)ToolaMachining typebVAM frequency (kHz)Surface roughness Ra
30 000 DC CM ⋯ 0.405 
30 000 DC CM ⋯ 0.396 
30 000 10 DC CM ⋯ 0.46 
30 000 10 DC CM ⋯ 0.451 
30 000 15 DC CM ⋯ 0.51 
30 000 15 DC CM ⋯ 0.503 
50 000 SC CM ⋯ 0.37 
50 000 SC CM ⋯ 0.376 
50 000 10 SC CM ⋯ 0.372 
10 50 000 10 SC CM ⋯ 0.376 
11 50 000 15 SC CM ⋯ 0.36 
12 50 000 15 SC CM ⋯ 0.361 
13 30 000 DC CM ⋯ 0.855 
14 30 000 DC CM ⋯ 0.851 
15 30 000 10 DC CM ⋯ 0.96 
16 30 000 10 DC CM ⋯ 0.965 
17 30 000 15 DC CM ⋯ 1.862 
18 30 000 15 DC CM ⋯ 1.766 
19 50 000 SC CM ⋯ 0.422 
20 50 000 SC CM ⋯ 0.42 
21 50 000 10 SC CM ⋯ 0.456 
22 50 000 10 SC CM ⋯ 0.46 
23 50 000 15 SC CM ⋯ 0.455 
24 50 000 15 SC CM ⋯ 0.452 
25 30 000 0.5 DC 1D VAM 27.22 0.29 
26 30 000 10 0.5 DC 1D VAM 27.22 0.255 
27 30 000 15 0.5 DC 1D VAM 27.22 0.327 
28 50 000 0.5 DC 1D VAM 27.22 0.257 
29 50 000 10 0.5 DC 1D VAM 27.22 0.293 
30 50 000 15 0.5 DC 1D VAM 27.22 0.26 
31 30 000 SC 1D VAM 27.22 0.151 
32 30 000 10 SC 1D VAM 27.22 0.104 
33 30 000 15 SC 1D VAM 27.22 0.129 
34 50 000 SC 1D VAM 27.22 0.102 
35 50 000 10 SC 1D VAM 27.22 0.122 
36 50 000 15 SC 1D VAM 27.22 0.109 
37 30 000 0.5 DC 1D VAM 27.22 1.213 
38 30 000 10 0.5 DC 1D VAM 27.22 1.419 
39 30 000 15 0.5 DC 1D VAM 27.22 1.119 
40 50 000 0.5 DC 1D VAM 27.22 1.23 
41 50 000 10 0.5 DC 1D VAM 27.22 1.322 
42 50 000 15 0.5 DC 1D VAM 27.22 1.358 
43 30 000 SC 1D VAM 27.22 1.713 
44 30 000 10 SC 1D VAM 27.22 1.711 
45 30 000 15 SC 1D VAM 27.22 1.525 
46 50 000 SC 1D VAM 27.22 1.42 
47 50 000 10 SC 1D VAM 27.22 1.851 
48 50 000 15 SC 1D VAM 27.22 1.82 
49 30 000 0.5 DC 2D Non-R 0.179 
50 30 000 10 0.5 DC 2D Non-R 0.17 
51 30 000 15 0.5 DC 2D Non-R 0.144 
52 50 000 0.5 DC 2D Non-R 0.166 
53 50 000 10 0.5 DC 2D Non-R 0.198 
54 50 000 15 0.5 DC 2D Non-R 0.152 
55 30 000 SC 2D Non-R 0.188 
56 30 000 10 SC 2D Non-R 0.152 
57 30 000 15 SC 2D Non-R 0.186 
58 50 000 SC 2D Non-R 0.166 
59 50 000 10 SC 2D Non-R 0.181 
60 50 000 15 SC 2D Non-R 0.168 
61 30 000 0.5 DC 2D Non-R 0.956 
62 30 000 10 0.5 DC 2D Non-R 1.361 
63 30 000 15 0.5 DC 2D Non-R 0.91 
64 50 000 0.5 DC 2D Non-R 1.244 
65 50 000 10 0.5 DC 2D Non-R 0.948 
66 50 000 15 0.5 DC 2D Non-R 1.166 
67 30 000 SC 2D Non-R 0.842 
68 30 000 10 SC 2D Non-R 0.323 
69 30 000 15 SC 2D Non-R 0.789 
70 50 000 SC 2D Non-R 0.295 
71 50 000 10 SC 2D Non-R 0.411 
72 50 000 15 SC 2D Non-R 0.318 
73 30 000 0.5 DC 2D FC 29.81 0.079 
74 30 000 10 0.5 DC 2D FC 29.81 0.016 
75 30 000 15 0.5 DC 2D FC 29.81 0.084 
76 50 000 0.5 DC 2D FC 29.81 0.018 
77 50 000 10 0.5 DC 2D FC 29.81 0.089 
78 50 000 15 0.5 DC 2D FC 29.81 0.018 
79 30 000 SC 2D FC 29.81 0.042 
80 30 000 10 SC 2D FC 29.81 0.013 
81 30 000 15 SC 2D FC 29.81 0.049 
82 50 000 SC 2D FC 29.81 0.021 
83 50 000 10 SC 2D FC 29.81 0.054 
84 50 000 15 SC 2D FC 29.81 0.026 
85 30 000 0.5 DC 2D FC 29.81 0.163 
86 30 000 10 0.5 DC 2D FC 29.81 0.164 
87 30 000 15 0.5 DC 2D FC 29.81 0.175 
88 50 000 0.5 DC 2D FC 29.81 0.185 
89 50 000 10 0.5 DC 2D FC 29.81 0.182 
90 50 000 15 0.5 DC 2D FC 29.81 0.188 
91 30 000 SC 2D FC 29.81 0.168 
92 30 000 10 SC 2D FC 29.81 0.122 
93 30 000 15 SC 2D FC 29.81 0.177 
94 50 000 SC 2D FC 29.81 0.141 
95 50 000 10 SC 2D FC 29.81 0.186 
96 50 000 15 SC 2D FC 29.81 0.142 
97 30 000 0.5 DC 2D FA 29.81 0.068 
98 30 000 10 0.5 DC 2D FA 29.81 0.036 
99 30 000 15 0.5 DC 2D FA 29.81 0.076 
100 50 000 0.5 DC 2D FA 29.81 0.043 
101 50 000 10 0.5 DC 2D FA 29.81 0.128 
102 50 000 15 0.5 DC 2D FA 29.81 0.048 
103 30 000 SC 2D FA 29.81 0.0642 
104 30 000 10 SC 2D FA 29.81 0.032 
105 30 000 15 SC 2D FA 29.81 0.0864 
106 50 000 SC 2D FA 29.81 0.041 
107 50 000 10 SC 2D FA 29.81 0.089 
108 50 000 15 SC 2D FA 29.81 0.038 
109 30 000 0.5 DC 2D FA 29.81 0.163 
110 30 000 10 0.5 DC 2D FA 29.81 0.164 
111 30 000 15 0.5 DC 2D FA 29.81 0.175 
112 50 000 0.5 DC 2D FA 29.81 0.185 
113 50 000 10 0.5 DC 2D FA 29.81 0.182 
114 50 000 15 0.5 DC 2D FA 29.81 0.188 
115 30 000 SC 2D FA 29.81 0.168 
116 30 000 10 SC 2D FA 29.81 0.122 
117 30 000 15 SC 2D FA 29.81 0.177 
118 50 000 SC 2D FA 29.81 0.141 
119 50 000 10 SC 2D FA 29.81 0.186 
120 50 000 15 SC 2D FA 29.81 0.142 
a

DC, diamond-coated; SC, solid carbide.

b

CM, conventional machining; Non-R, nonresonant mode VAM; FA, resonant mode VAM with vibrations in feed and axial directions; FC, resonant mode VAM with vibrations in feed and cross-feed directions.

The longitudinal-bending design technique of the ultrasonic horn offers the best option for ultrasonic vibration of the workpiece for the 2D ultrasonic resonant system. The ultrasonic vibrations were used in two tests, one with them in the feed and cross-feed (FC) directions and the other with them in the feed and axial (FA) directions of the tool. The direction of applied vibrations for the nonresonant system was fixed as the FC direction. The surface roughnesses of the slot bottoms were determined using a surface roughness tester (Mitutoyo SJ-410). To ensure accuracy, each measurement was repeated three times at equidistant points along the machined slot. The measurement distance covered the entire length of the slot, and the average value was calculated to represent the surface roughness Ra for each machining condition.

The occurrence of edge chipping during the machining process is one of the difficulties that arise in the micro-milling of glass. This can have a big impact on the quality and usefulness of the final product. This problem of edge chipping has been addressed using a variety of instruments and methods. For micro-milling glass, two popular tool types are solid carbide and diamond-coated tools. The degree of edge chipping seen during the machining process can vary greatly, depending on the tool type selected. The effects of solid carbide and diamond-coated tools on edge chipping in micro-milled glass were examined experimentally using both types of tools, with diameters of 0.4 and 0.5 mm, to micro-mill slots into glass.

SEM photographs of the micro-milled slots revealed that conventional machining with solid carbide tools resulted in greater edge chipping than slots milled with diamond-coated tools and other milling techniques. This is because tools with diamond coatings are harder and more resistant to wear than solid carbide tools, and thus the diamond-coated tools can maintain their sharpness for a longer period of time. It was also found that the edge chipping caused by the use of solid carbide tools could be somewhat reduced by the use of high-frequency 1D ultrasonic or 2D nonresonant VAM instead of conventional micro-milling. This decrease in edge chipping can be attributed to the effects of vibration, which lessens the force of the tool’s impact on the glass. In addition, as the slots milled using VAM showed less edge chipping with both types of tools, it appears that regardless of the tool employed, vibration-assisted procedures can dramatically lower the incidence of edge chipping of a micro-milled glass.

A considerable decrease in chipping size was seen when high-frequency 2D resonant ultrasonic vibrational assistance was used during micro-milling operations employing either a diamond-coated or a solid carbide tool. The approximate size of the chipped section was determined by measurements with ImageJ software. The reductions in edge chipping achieved with the aid of vibrational assistance were found to be ∼187.7% and 173.5% for solid carbide and diamond-coated tools, respectively. As already mentioned, the results also revealed that both high-frequency 1D resonant and low-frequency 2D nonresonant VAM reduced edge chipping compared with conventional machining. However, the reduction achieved with these systems was less than with 2D resonant VAM. Thus, these results indicate that the use of high-frequency 2D resonant ultrasonic vibrational assistance can significantly enhance the quality of micro-milled slots, particularly in terms of reduced edge chipping, which is essential for achieving high precision and accuracy in micro-milling operations. It was also found that the use of high-frequency 2D VAM with vibrations applied in the feed and axial directions gave inferior results with regard to edge chipping compared with the other VAM types. This result for glass is consistent with micro-milling tests on silicon, where high-frequency 2D VAM likewise performed poorly. It implies that high-frequency 2D VAM might not be appropriate for micro-milling of brittle materials such as silicon and glass.

The optical qualities, durability, and dependability, and thus the performance and functionality, of micro-milled glass components are significantly influenced by the surface quality of the machined glass. Figures 4 and 5 present SEM images illustrating the surface quality of micro-milled slots. Analysis of these images reveals that compared with conventional machining, both 1D resonant and 2D nonresonant VAM reduce surface defects, while 2D resonant mode VAM in both the FC and FA directions gave better results for the surface roughness value Ra. Figure 6 shows plots of Ra for different spindle speeds and feed rates. For a diamond-coated tool with 30 000 rpm spindle speed and 0.5 μm vibration amplitude, both 1D VAM and 2D nonresonant mode VAM gave moderate improvements of 33.09% and 77.39%, respectively, in Ra compared with conventional micro-milling. More significant improvements in Ra, by 134.71% and 142.49%, respectively, were achieved with 2D resonant mode VAM in the FC and FA directions, and with both of these methods, an increase in the vibration amplitude to 1 μm led to a slight further improvement in Ra by 184.79%. For the solid carbide tool, both 1D ultrasonic and 2D nonresonant VAM gave significantly poorer results than conventional micro-milling, whereas 2D resonant mode VAM in both the FC and FA directions gave 135.95% improvements in Ra. In the case of resonant mode VAM, an increase in vibration amplitude from 0.5 to 1 μm, gave little or no further improvement in Ra. Thus, taken together, the above results indicate that for both diamond-coated and solid carbide tools, an improved surface roughness Ra can be achieved with the use of a high spindle speed in conjunction with ultrasonic vibration at a low amplitude in the feed–cross-feed directions.

FIG. 4.

SEM images of micro-milled slots in glass.

FIG. 4.

SEM images of micro-milled slots in glass.

Close modal
FIG. 5.

Chipping of micro-milled glass.

FIG. 5.

Chipping of micro-milled glass.

Close modal
FIG. 6.

Surface roughness plots at different combinations of spindle speed and feed rate: (a) spindle speed 30 000 rpm and feed rate 5 mm/min; (b) 50 000 rpm and 5 mm/min; (c) 30 000 rpm and 10 mm/min; (d) 50 000 rpm and 10 mm/min; (e) 30 000 rpm and 15 mm/min; (f) 50 000 rpm and 15 mm/min.

FIG. 6.

Surface roughness plots at different combinations of spindle speed and feed rate: (a) spindle speed 30 000 rpm and feed rate 5 mm/min; (b) 50 000 rpm and 5 mm/min; (c) 30 000 rpm and 10 mm/min; (d) 50 000 rpm and 10 mm/min; (e) 30 000 rpm and 15 mm/min; (f) 50 000 rpm and 15 mm/min.

Close modal

The productivity, quality, and cost of glass micro-milling are all significantly affected by tool life, which itself is influenced by cutting parameters, tool geometry, tool material, and workpiece material. Owing to the hardness and fragility of glass, its cutting requires a ductile regime technique to prevent surface damage such as fracturing, chipping, and cracking. A tiny depth of cut, a low feed rate, a high cutting speed, and a sharp tool with a narrow edge radius are required for ductile regime cutting. However, these conditions also increase tool wear and shorten tool life. The surface quality and dimensional accuracy of micro-milled glass objects can be affected by tool wear. In addition to causing tool wear, tool breakage can harm both the machine and the workpiece. Consequently, tool life is a crucial aspect to be considered and improved when glass is to be micro-milled. There have been a number of investigations of tool life and wear processes in the micro-milling of glass employing several types of tools, including diamond, polycrystalline diamond (PCD), cubic boron nitride (CBN), and tungsten carbide (WC). It has been found that owing to their great hardness and low coefficient of friction, diamond tools provide the best surface quality and the longest tool life. However, diamond tools are costly and challenging to manufacture with intricate shapes. PCD tools are cheaper and simpler to manufacture, but owing to their lower hardness and higher coefficient of friction, they have a shorter tool life and provide a poorer surface quality. Tools made of CBN and WC are also less expensive and simpler to produce, but their even lower hardness and higher coefficient of friction lead to further reductions in tool life and surface quality. The use of coated tools, cutting fluids, or lubricants, optimization of cutting parameters, and the use of adaptive control systems are among the methods that have been adopted to increase tool life when micro-milling glass.15–17 

Images of tool wear from tests employing several VAM systems to micro-mill glass are shown in Fig. 7. Similar to what was found in investigations of tool wear in machining of silicon, it was discovered that the VAM system used had a considerable impact on tool wear. It was demonstrated that 2D resonant mode VAM with vibrations in the feed and axial directions and 1D ultrasonic machining both reduced diamond coating spalling, a common type of tool wear when diamond-coated tools are used in micro-milling operations. Furthermore, compared with other milling techniques, the use of 2D nonresonant mode machining led to a considerable reduction in coating spalling, indicating that this may be a promising VAM system for micro-milling of glass operations.

It was found that among all the milling types, 2D high-frequency ultrasonic resonant mode VAM with vibrations in the feed and cross-feed directions led to the least tool wear. Thus, considering both tool wear and tool life, this may be the most suitable method for micro-milling of glass. Overall, the tool wear study has provided valuable information on the performance of various VAM systems and can assist in choosing the best system for micro-milling of glass.

FIG. 7.

(a) Diamond-coated tool tip. (b) Solid carbide tool.

FIG. 7.

(a) Diamond-coated tool tip. (b) Solid carbide tool.

Close modal

Such analyses of tool wear in glass micro-milling operations are of crucial importance in shedding light on how various VAM systems and cutting parameters work. Studies of tool wear in milling with solid carbide tools have indicated that some VAM methods can lead to increased tool wear and possibly even tool breakage. In particular, it has been found that in milling with solid carbide tools, both conventional machining and 2D nonresonant VAM can result in higher tool wear. This is probably because of the high temperatures and cutting pressures produced during these types of milling, which can hasten tool wear and even lead to tool breakage. Furthermore, during these milling operations, large chunks can come loose from the tool, leading to tool failure and damage to the glass component being machined.

However, it has also been discovered that in the micro-milling of slots in glass, the use of 1D and 2D ultrasonic VAM can decrease tool wear. This is probably because these milling methods generate lower cutting pressures and vibrations, which can help to reduce tool wear and extend tool life. Additionally, the use of ultrasonic VAM for micro-milling of glass can help to improve both surface quality and dimensional accuracy.

In conclusion, this study has highlighted the significance of tool selection and VAM techniques for the micro-milling of glass components. It has been found that owing to their hardness and wear resistance, diamond-coated tools are more effective in reducing edge chipping, Moreover, the use of high-frequency 2D resonant ultrasonic VAM significantly improves micro-milled slot quality by reducing edge chipping for both diamond-coated and solid carbide tools. Vibration-assisted procedures, in general, show promise in minimizing edge chipping compared with conventional machining. Tool wear analysis has revealed that certain VAM systems, including 2D resonant mode VAM with vibrations in the feed and axial directions and 1D ultrasonic machining, reduce spalling of the diamond coating. The use of 2D nonresonant mode VAM also gives a significant reduction in coating spalling, making it a promising option for glass micro-milling. High-frequency ultrasonic resonant 2D resonant mode VAM with vibrations in the feed and cross-feed directions exhibits less tool wear than other milling types, making it favorable for use in glass micro-milling. Considering these findings, appropriate selection of both the tool and the vibration-assisted technique are crucial for achieving high precision, improved surface quality, and dimensional accuracy in micro-milling of glass components, with the ultimate aim of ensuring an efficient and reliable manufacturing process. Overall, an ultrasonic vibration-assisted micro-milling (UVAMM) system is able to achieve a nanometric surface roughness Ra as low as 0.104 μm, compared with the value of 0.46 μm obtained with conventional machining. UVAMM also results in improvements in edge quality and tool wear, with a tool wear reduction of up to 50% being observed.

  • Edge chipping reduction

    1. The use of ultrasonic VAM with solid carbide and diamond-coated tools gives approximate reductions of 187.7% and 173.5%, respectively, in edge chipping.

    2. High-frequency 1D ultrasonic VAM results in less edge chipping compared with conventional machining, but this reduction is less than with 2D resonant VAM.

    3. High-frequency 2D resonant ultrasonic VAM gives a considerable decrease in chipping size, leading to a significant enhancement in the quality of micro-milled slots compared with conventional machining with both solid carbide and diamond-coated tools.

  • Surface roughness Ra improvement

    1. For a diamond tool with 30 000 rpm spindle speed and 0.5 μm vibration amplitude, 1D VAM gives a 33.09% improvement in Ra value compared with conventional machining, while 2D nonresonant mode VAM, 2D resonant mode VAM with vibrations in the feed and cross-feed (FC) directions, and 2D resonant mode VAM with vibrations in the feed and axial (FA) directions give 77.39%, 134.71%, and 142.49% improvements, respectively.

    2. For a diamond tool with increased amplitude (1 μm), the Ra values remain better than with conventional machining, but no significant further improvement is observed. For a solid carbide tool, 1D ultrasonic and 2D nonresonant VAM give poorer results than conventional machining, but 2D resonant mode VAM with vibrations in the FC and FA directions both give 135.95% improvements in Ra compared with conventional machining.

    3. A high spindle speed and a small amplitude of ultrasonic vibration in the FC direction give a better surface roughness with both diamond-coated and solid carbide tools in 2D ultrasonic VAM.

  • Tool wear reduction

    1. Both 2D resonant mode VAM with vibrations in the FA direction and 1D ultrasonic VAM exhibit a considerable reduction in spalling of the diamond coating (a common source of tool wear with diamond-coated tools) compared with other milling techniques.

    2. Compared with other milling techniques, 2D nonresonant mode VAM demonstrates a significant reduction in coating spalling, making it a promising VAM system for glass micro-milling.

    3. High-frequency ultrasonic resonant mode VAM with vibrations in the FC direction exhibits less tool wear compared with all other milling types, suggesting that it may be favored for glass micro-milling when both tool wear and tool life are taken into account.

    4. Tool wear has been quantified by measuring the reduction in tool diameter and the volume of material removed from the tool tip. For instance, the UVAMM system exhibits a tool wear rate of 0.005 mm3/h, compared with a rate of 0.01 mm3/h for conventional machining, indicating a significant reduction in tool wear with the use of ultrasonic vibrations.

The authors have no conflicts to disclose.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1.
Zhao
Y
,
Zhang
Y
,
Huo
D
,
Teng
X
,
Shi
Y
.
An experimental investigation on the surface morphology in micro-end-milling of glass
.
2014 20th International Conference on Automation and Computing
.
IEEE
.
2014
. pp.
104
109
.
2.
Voldman
J
,
Gray
ML
,
Schmidt
MA
.
Microfabrication in biology and medicine
.
Annu. Rev. Biomed. Eng.
1999
;
1
(
1
):
401
425
.
3.
Nakasuji
T
,
Kodera
S
,
Hara
S
,
Matsunaga
H
,
Ikawa
N
,
Shimada
S
.
Diamond turning of brittle materials for optical components
.
CIRP Ann.
1990
;
39
(
1
):
89
92
.
4.
Moriwaki
T
,
Shamoto
E
,
Inoue
K
.
Ultraprecision ductile cutting of glass by applying ultrasonic vibration
.
CIRP Ann.
1992
;
41
(
1
):
141
144
.
5.
Pawar
P
,
Ballav
R
,
Kumar
A
.
Micromachining of borosilicate glass: A state of art review
.
Mater. Today: Proc.
2017
;
4
(
2
):
2813
2821
.
6.
Mamedov
A
.
Machining of brittle materials: Micro milling of glass
.
J. Mech. Sci. Technol.
2021
;
35
(
9
):
4143
4148
.
7.
Gurgen
S
,
Sofuoglu
MA
.
Advancements in conventional machining: A case of vibration and heat-assisted machining of aerospace alloys
. In:
Gupta
K
,
Pramanik
A
, ed. Advanced Machining and Finishing.
Amsterdam, Netherlands
:
Elsevier
.
2021
. pp.
143
175
.
8.
Kim
T
,
Kwon
K-K
,
Chu
CN
,
Song
KY
.
Experimental investigation on CO2 laser-assisted micro-slot milling characteristics of borosilicate glass
.
Precis. Eng.
2020
;
63
:
137
147
.
9.
Vipindas
J
,
Mathew
K
.
Machining of borosilicate glass using micro-end milling
. In:
Shunmugam
MS
,
Kanthababu
M
, ed. Advances in Micro and Nano Manufacturing and Surface Engineering.
Singapore
:
Springer
.
2019
. pp.
189
200
.
10.
Yan
BH
,
Wang
AC
,
Huang
CY
,
Huang
FY
.
Study of precision micro-holes in borosilicate glass using micro EDM combined with micro ultrasonic vibration machining
.
Int. J. Mach. Tools Manuf.
2002
;
42
(
10
):
1105
1112
.
11.
Zheng
L
,
Chen
W
,
Huo
D
.
Review of vibration devices for vibration-assisted machining
.
Int. J. Adv. Des. Manuf. Technol.
2020
;
108
(
5–6
):
1631
1651
.
12.
Satpute
V
,
Huo
D
,
Hedley
J
,
Elgendy
M
.
Design of a novel 2D ultrasonic transducer for 2D high-frequency vibration–assisted micro-machining
.
Int. J. Adv. Des. Manuf. Technol.
2023
;
126
:
1035
.
13.
Zheng
L
,
Chen
W
,
Huo
D
,
Lyu
X
.
Design, analysis, and control of a two-dimensional vibration device for vibration-assisted micromilling
.
IEEE/ASME Trans. Mechatron.
2020
;
25
(
3
):
1510
1518
.
14.
Perveen
A
,
Molardi
C
.
Machining of glass materials: An overview
. In:
Gupta
K
, ed. Advanced Manufacturing Technologies: Modern Machining, Advanced Joining, Sustainable Manufacturing.
Cham
:
Springer International Publishing
.
2017
. pp.
23
47
.
15.
Chen
N
, et al.
Advances in micro milling: From tool fabrication to process outcomes
.
Int. J. Mach. Tools Manuf.
2021
;
160
:
103670
.
16.
Balázs
BZ
,
Geier
N
,
Takács
M
,
Davim
JP
.
A review on micro-milling: Recent advances and future trends
.
Int. J. Adv. Des. Manuf. Technol.
2020
;
112
:
655
.
17.
Ku
X
,
Zhang
Z
,
Liu
X
,
Chen
L
,
Li
G
.
Low-cost rapid prototyping of glass microfluidic devices using a micromilling technique
.
Microfluid. Nanofluid.
2018
;
22
(
8
):
82
.

Vinod Satpute earned his B.E. in Mechanical Engineering from Shivaji University, India, and his Masters in Advanced Mechanical Engineering from Cardiff University, UK, in 2018. He completed his Ph.D. at Newcastle University, UK, in 2023. Currently, he serves as a Research Associate in the School of Engineering, Manchester Metropolitan University, UK. Dr. Satpute’s research interests encompass mechanical design and structural dynamics, vibration and noise isolation, vibration-assisted technology, precision fabrication, new product development, and computational modeling. He specializes in developing innovative engineering solutions that integrate cutting-edge computational techniques and experimental approaches.

Dehong Huo received the Ph.D. degree in Precision Engineering from Harbin Institute of Technology, China, in 2004. He is currently a Reader in the School of Engineering at Newcastle University, UK. His team is currently working on research projects funded by the EPSRC, Innovate UK, and industry. His research interests include precision machine tool design, precision motion control systems, innovative manufacturing processes, and micro/nano cutting mechanics. Dr. Huo has authored more than 150 publications, including three books. He is an editor and associate editor of several international journals.

John Hedley received the Ph.D. degree in Atomic Physics from Newcastle University, UK, in 1996. After three Postdoctoral Researcher positions in MEMS design, fabrication, and characterization, he obtained an academic position within the School of Engineering at Newcastle University, where he is currently Senior Lecturer. His research interests are in mechatronics engineering, sensors, robotics and AI. Dr. Hedley has authored more than 80 research publications.

Patrick Degenaar is Professor of Neuroprosthetics at Newcastle University, UK, and has a particular interest in visual prosthetics for the blind. He attained first class B.Sc. and M.Res. degrees in Applied Physics from Liverpool University, UK, in 1996 and 1997 respectively, and a Ph.D. in Biophysics from the Japan Advanced Institute of Science and Technology in 2001. In 2005, he obtained an RCUK fellowship and Lectureship at Imperial College, London, UK, and moved his team to Newcastle in 2010. Since then, he has worked on a diverse range of topics from retinal prosthetics, visual prosthetics, brain implants for epilepsy, and the core technologies to support those applications.

Carl Dale received the Ph.D. degree in Atomic Physics from Newcastle University, UK, in 2012. Dr. Dale is the co-founder and CEO of Microbritt Ltd. With a focus on semiconductor processing, quantum sensing, and microfluidics, his work integrates cutting-edge micro-milling techniques to deliver precision manufacturing solutions for brittle materials such as silicon and glass. Dr. Dale has published extensively in the fields of bioelectronics, MEMS devices, and quantum technologies, contributing to the advancement of both academic research and industry applications.