Open Access
Add to BibSonomy
Abstract
Polycrystalline zinc selenide is widely used in advanced optical systems due to its superior optical properties. However, the soft and brittle properties bring a challenge for high-quality surface processing. In recent years, elliptical vibration cutting has been proven as a promising method for machining brittle materials. In the present research, a series of grooving and planning experiments were carried out to investigate the machinability of zinc selenide with elliptical vibration cutting. The removal mechanism was analyzed from fracture characteristics, chip morphology, and phase transformation. The results show that elliptical vibration cutting is effective in suppressing cleavage-induced craters. Reducing the nominal cutting speed is beneficial to inhibit the spring back-induced tearing of grains. A 94-time increase in the critical depth of cut was achieved by vibration trajectory optimization compared to ordinary cutting. Moreover, the influence mechanism of feed on the evolution of surface morphology was revealed. Finally, a zinc selenide microlens array was successfully fabricated. The performance was evaluated by geometric parameter measurements and a multiple imaging test. The findings provide a prospective method for ductile regime machining of zinc selenide.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Zinc selenide (ZnSe) is a superior optical material for its extremely high transmittance in white light and infrared with wavelengths of 0.45µm-20µm [1]. There are lots of important applications of ZnSe in multi-band sensor windows [2,3], infrared lenses [4], solar cells [5], and lighting phosphors [6]. The surface quality of ZnSe products is a vital factor affecting their performance. However, ZnSe is a typical soft and brittle material with low hardness and fracture toughness [7], which is challenging for ultra-precision machining. Chemical mechanical polishing (CMP) can realize the nano-scale roughness surface through the combination of mechanical removal and chemical etching [8,9]. However, the method is difficult to meet the requirements of complex surface manufacturing efficiently. Diamond cutting technology is a common process of ultra-precision machining. The diamond tool with a sharp edge was driven in a complex trajectory by the machine tool meeting the requirement of high-quality complex surface fabrication [10,11]. However, the properties of ZnSe are prone to fracture during diamond cutting and decrease the surface quality [12]. Therefore, it is necessary to control the process to suppress the machining defects [13].
For the diamond cutting process and nano-removal mechanism of polycrystalline ZnSe, several studies have been conducted. Fang et al. [14] investigated the fracture of ZnSe by nanoindentation and diamond cutting experiments. They found that the tool with a 0° rake angle is beneficial in suppressing crack generation compared to that with a negative rake angle. Shojaee et al. [15] analyzed the effect of feed on residual stresses in diamond cutting of ZnSe by confocal Raman spectroscopy. They found that with the increase of feed, the residual tensile stresses in the material increase, and the lattice quality decrease. Shigeaki [16] studied the phase transformation path of ZnSe by the diamond anvil and analyzed the pressure corresponding to the phase transition from zinc blende to the rock-salt structure. It provides guidance for the study of the phase transformation process of ZnSe during diamond cutting. However, there is a difference in the force state of the material during the cutting process and indentation process. Yan et al. [17] carried out diamond cutting experiments on ZnSe under different conditions. They observed that there are two kinds of cracks on the machined surface, including micron-scale craters caused by cleavage and submicron-scale pits caused by tearing. Moreover, they revealed the correlation between phase transition path in the cutting process and found that in the ductile regime cutting of ZnSe, phase transformation could not be detected on the machined surface. Geng et al. [18] studied the removal mechanism in the ductile regime by transmission electron microscopy (TEM). It was revealed that high density dislocations, stacking dislocations, and deformation twins promoted the plastic deformation of the material. Li et al. [19] optimized the diamond cutting process for ZnSe by orthogonal experiments, and realized a machined surface with a roughness of Ra 3.1 nm. However, no effective solution has been found to suppress the tearing defects caused by the difference in orientation of each grain.
In recent years, ultrasonic vibration-assisted diamond cutting technology has been proven to be an effective technology for the processing of brittle materials. Compared with ordinary cutting (OC), ultrasonic vibration-assisted cutting can reduce the average cutting force through changing the machining state from continuous cutting to intermittent cutting, and inhibit the fracture of brittle materials [20]. In 1994, Shamoto and Moriwaki [21] proposed the elliptic vibration cutting (EVC) technology, which further improved the stress state and the cooling conditions of the cutting area. Subsequently, EVC has become a promising method in ultra-precision machining. Rahman et al. [22] analyzed the relationship between maximum cutting thickness and machining parameters in EVC. It indicated that the maximum cutting thickness decreases significantly when the speed ratio is below 0.12837. The influence of speed ratio on the ductile-regime cutting of hard and brittle materials in EVC was verified by sintered tungsten carbide cutting experiments. Ehamann et al. [23] established a theoretical model to depict the relationship between vibration phase shift and critical speed ratio. The model reveals the influence of phase shift on the ductile-regime machinability of hard and brittle materials. Zhang et al. conducted fundamental experiments of EVC on binderless tungsten carbide [24], single-crystal silicon [25], and hardened steel [26], and found that elliptical vibration assistance can effectively improve the machinability of difficult-to-cut materials. Furthermore, Zhang et al. [27] studied the surface formation mechanism of soft and brittle material calcium fluoride by EVC, and confirmed that the critical depth of cut (DoC) is improved to 42 times higher than that by the OC. Wang et al. [28] developed an elliptical vibration diamond cutting device coupled with resonance and non-resonance, which can generate modulated elliptical vibration locus. Moreover, crack-free micro dimple arrays are generated on single-crystal silicon by modulated EVC.
Although research on the cutting mechanism and process of ZnSe has been conducted extensively, there are still difficulties in suppressing the defects caused by grain tearing in machining. Ordinary diamond cutting can achieve ductile-regime cutting of ZnSe with extremely small feed, but it poses a challenge for efficient machining of high-quality complex surfaces. The application of EVC in the ductile-regime cutting of brittle materials has achieved promising results, however, research on EVC for polycrystalline ZnSe has not been reported. In addition, few studies have been conducted on the optical properties of the ZnSe machined surface.
In this study, a series of grooving and planning experiments were conducted on polycrystalline ZnSe to study the material removal mechanism in EVC. The influence mechanism of nominal cutting speed and vibration trajectory on the machinability was investigated by analyzing the critical DoC, chip morphology, and phase transition. Through vibration locus controlling, a 94-time increase in the critical DoC was achieved. Furthermore, the evolution of surface morphology at different feeds was analyzed, and the influence of feed on defects caused by cleavage and grain tearing was revealed. Finally, a microlens array (MLA) was successfully fabricated on ZnSe by EVC combined with the slow tool servo technique. The MLA was evaluated by geometric parameter measurements and multiple imaging tests. The findings provide an effective technology for high-quality surface processing of polycrystalline ZnSe.
2. Methodologies
2.1 Material property
In this study, a chemical vapor deposited (CVD) ZnSe cylinder (Φ24 mm × 8 mm) was utilized as the workpiece. The workpiece was polished to a surface roughness of Sa 2 nm, and the grain distribution of the workpiece was observed by an electron backscatter diffraction (EBSD) detector (Zeiss Sigma 300VP). Fig. 1(a) shows the inverse pole figure (IPF) map of the workpiece. It can be observed that there are many twins inside the material. In addition, there are subgrains with relatively similar orientation angles inside the coarse grains, as indicated by the marking in the black circles in Fig. 1(a). Fig. 1(b) shows the distribution of grain size in the observed area. The diameters of the grains are in the range of 0-120 µm and more than 70% of the grains are within 20 µm in diameter.
2.2 Principle of EVC
The principle of EVC is shown in Fig. 2, where the x-axis and the z-axis are defined as the nominal cutting direction and the nominal DoC direction. The tool forms an elliptical vibration trajectory excited by two sinusoidal signals with the same frequency in the nominal cutting direction and the nominal DoC direction. Synchronously, the tool and the workpiece form a relative motion along the nominal cutting direction driven by the machine tool. Assuming that the machine tool drives the tool motion at a speed of vc along the nominal cutting direction, the relative motion trajectory between the tool and the workpiece can be expressed as [29]:
where xe and ze denote the relative displacement components in the x direction and the z direction between the tool and the workpiece. Ac and Ad denote the total amplitudes in the nominal cutting direction and the nominal DoC direction, respectively, ω denotes the frequency of vibration, τ is the time in the vibration cycle, and φ is the phase shift between the two directions of vibration, which is typically set to be -90◦.During the single cycle elliptical vibration shown in Fig. 2, the tool starts to contact the workpiece at time t1. At time t2, the tool reaches the bottom of the trajectory. Subsequently, the rake face of the tool starts to contact the chip formed in the previous cycle at t3. The tool completes the formation of the machined surface in the cycle at the time t4, and the surface machined beyond the time t4 will be removed in the next cycle. Due to the pulling-up motion of the tool, frictional reversal between the rake face and the chip at the t5. At t6, the tool starts to separate from the chip, and the coolant can enter the cutting area to improve the cutting condition. And there is a period between t6 and t0. To ensure that the tool separates from the workpiece during each cycle, the maximum vibration speed along the nominal cutting direction usually is set above vc. A residual mark will be left on the machined surface during each vibration cycle due to the characteristic of vibration trajectory, which is marked as Rth in Fig. 2.
2.3 Experimental setup
Diamond cutting experiments were carried out on the ultra-precision machine tool of Precitech Nanoform X. The elliptical vibration of the diamond tool was realized by a two-degree-of-freedom elliptical vibrator from Taga Electric Co., Ltd., which works at a resonant frequency of 41.6 kHz with vibration amplitude in the range of 0-4 µm, as shown in Fig. 3. A single crystal diamond (SCD) tool with a 0° rake angle, 10° clearance angle and 1 mm nose radius is clamped in the elliptical vibrator. The elliptical vibrator was mounted on the B axis of the machine tool, and the clamping fixture with the workpiece was attached to the vacuum chuck at the end of the spindle. The DoC was continuously varied with the motion of the X axis in the grooving experiments to study the removal characteristic of ZnSe under OC and EVC. The detailed machining conditions are shown in Table 1.
Table 1. Conditions of grooving experiments
Planning experiments were carried out to study the removal mechanism of ZnSe during continuous feed cutting. The effects of elliptical trajectory and material removal rate on the EVC process were investigated by cutting with different feeds at the elliptic trajectory inclination angles of 90° and 110°. The specific experimental parameters are shown in Table 2.
Table 2. Parameters of planning experiments
2.4 Characterization methods
The grooves morphology and critical DoC were observed by optical microscope (ZEISS Axiocam 208 color optical microscope). The white light interferometer (Zygo Newview 9000) was utilized to observe the crack feature and measure the roughness of the machined surface. The morphology of the collected chips was observed by scanning electron microscopy (HITACHI SU3900). The phase transition of the machined surface and chips was measured by a Raman spectrometer (Renishaw inVia Reflex).
3. Results and discussion
3.1 Analysis of grooving experiments
Grooving experiments were conducted to investigate the effects of EVC and OC on the machinability of ZnSe. The cutting trajectory of the tool with different parameters is shown in Fig. 4. Fig. 5(a) shows the grooves morphology generated by OC and EVC with different amplitudes along the nominal DoC direction. Where the cutting speed was kept at 200 mm/min, the phase shift in EVC was maintained at -90°, and the amplitude was set to 1 µm, 2 µm, and 4 µm. At the initial position of the grooves, the material was removed by plastic deformation in both OC and EVC, and formed a smooth surface. This is because the cutting force is not enough to sustain the crack expansion and merging at small DoC, and the energy input from the machine tool is mainly used for plastic deformation [30]. As the DoC increased to 30 nm, the groove obtained by OC met the critical DoC of brittle-to-ductile (BTD) transition and the brittle fracture occurred. Due to the difference in the slip system orientation of each grain, local plasticity still exists in the initial of the brittle region. With the further increase of DoC, the crack density increased rapidly and finally distributed on the surface of the whole groove. Meanwhile, the grooves obtained by EVC realized a wide range of ductile regions without cleavage and subgrain spalling. As reported in previous research, the instantaneous uncut chip thickness in a single vibration cycle of EVC is much less than the nominal DoC. In addition, the intermittent removal form reduces the average cutting force [25]. It is worth noting that the scale and density of cracks in the brittle region of the groove obtained by EVC are much smaller than that obtained by OC, which again proves the inhibitory effect of EVC on crack expansion. The critical DoC was increased from 1536 nm to 2040nm as the amplitude along the nominal DoC direction was increased from 1 µm to 2 µm. Due to the increase in amplitude, the friction between the tool and the workpiece is reduced with the pulling-up motion [27], which inhibits the generation of cracking defects. The critical DoC decreased from 2040nm to 893 nm as the amplitude along the nominal DoC direction was increased from 2 µm to 4 µm. This is because the vibration trajectory becomes steeper with the further increase in amplitude, and a greater tensile stress is generated on the material as the tool pulls up, inducing a fracture.
Fig. 5. Morphology of the grooves. (a) Grooves formed by OC and EVC with different amplitudes along DoC direction; (b) Grooves formed by EVC with different nominal cutting speeds.
Nominal cutting speed determines the relative motion trajectory of the tool and the workpiece in EVC. To investigate the influence of nominal cutting speed on the machinability of ZnSe in EVC, the amplitude along the nominal DoC direction was maintained at 2 µm, and the nominal cutting speed was set to 500 mm/min and 1000 mm/min in the experiments. The morphology of the grooves generated by different nominal cutting speeds is shown in Fig. 5(b). The result shows that with the increase of nominal cutting speed, significant grain tearing occurs on the grooves. Grain tearing is a special defect in the cutting process of polycrystalline soft and brittle materials, which mainly occurs in the grains with the cleavage plane perpendicular to the machined surface but not perpendicular to the cutting direction. The grain will be torn along the direction of the perpendicular to the cleavage plane due to the springs back of the material after the tool passes. As the nominal cutting speed decreases, the tool interacts more frequently with the material in a certain area, and the high-frequency cyclic loading inhibits the springs back of the material, thus avoiding the tearing of grains. The phenomenon is consistent with the result of cyclic nanoindentation on single crystal silicon carried out by Yan [31]. Thus, no significant grain tearing was observed with the change of amplitude at a nominal cutting speed of 200 mm/min. In addition, the tool-workpiece contact time in a single cycle is reduced, leading to a decrease in the average cutting force. It is noteworthy that the twins in the material provided more slip systems for the grains and inhibited grain tearing. Furthermore, localized non-uniform defect within the torn grains can be observed in the Fig. 5(c), which may be caused by subgrains with large orientation differences within the grains. As the nominal cutting speed increases, the volume of removal material increases in a single vibration cycle, and fracture is more prone to occur at the same DoC. Therefore, the reduction of nominal cutting speed is beneficial to suppress the tearing of ZnSe grains and improve the ductile-regime machinability.
The white light interferometer (WLI) was used to observe defects generated in OC and EVC processes, and the observed area was in the white box in Fig. 5(a). The results were shown in Fig. 6, the shape of the groove was removed by the cylindrical filter. Where area 1 was the surface topography formed by OC, and area 2 was the surface topography formed by EVC with an amplitude of 1µm. It can be seen that there were a large number of micron craters with step-like edges on the surface of area 1. As the cutting force increases with the increasing DoC, when it is applied along the direction perpendicular to the cleavage plane of a grain, cracks will initiate, expand and merge along the cleavage plane, and finally, the micron craters were formed. The side of the crater is smooth and steep as shown in the section profile, and it is the cleavage plane of the grain. In contrast, only a small number of pits are observed in area 2, and the pits are all submicron in size. Since the material removal rate in a single vibration cycle of EVC is much smaller than that of OC at the same nominal DoC, the induced cutting force is smaller and the generation of micron craters cracks is avoided. In addition, the material is removed intermittently, avoiding the continuous expansion of cracks, resulting in a surface with only submicron pits instead of the micron craters occurring in OC.
Fig. 6. Morphology of brittle region and section profile of crack. (a) Surface morphology in area 1; (b) Surface morphology in area 2.
The collected chips generated by OC and EVC were observed by scanning electron microscopy (SEM) to investigate the chip formation mechanism during the cutting process, as shown in Fig. 7. Due to the small critical DoC of OC, the brittle fracture was produced at small DoC, forming the particle-shaped chips. As the DoC further increases, the cracks expand along the cleavage plane and merge, resulting in large-sized material spalling, and forming block-shaped chips. Instead, the chip morphology was mainly needle-like with the amplitude along the nominal DoC direction of 1µm and 2µm, which is a characteristic of ductile domain cutting. As the groove width increases, the length of the chips increases correspondingly, resulting in a difference in the length of the chips. A large number of long needle-like chips were observed in the chips generated by an amplitude of 2µm, which is consistent with the phenomenon of large critical DoC in the groove. In addition, a small percentage of block-shaped chips were observed, which were generated by fracture after the DoC increased to the critical DoC. However, a large quantity of block-shaped chips and part of short needle-shaped chips were observed in the chip generated by an amplitude of 4 µm. It is due to the chipping of the material caused by the tool edge during the pulling process. Moreover, the chips tend to fracture as being pushed out along the steep trajectory by the rake face. Fig. 7(b) shows the chip morphology formed by different nominal cutting speeds. As the nominal cutting speed increases, the percentage of block-shaped chips gradually increases, which proves that the ductile regime machinability of the material decreases. In addition, it is observed that the chip thickness increases with the increase in nominal cutting speed, which is attributed to the material removal volume increases in a single vibration cycle.
Fig. 7. Morphology of the collected chips. (a) Chips formed by OC and EVC with different amplitudes along DoC direction; (b) Chips formed by EVC with different nominal cutting speeds.
Previous studies [16] have shown that ZnSe will undergo a phase transition in the condition of high pressure. To investigate the phase transition during OC and EVC, the surface and the chips of ZnSe were analyzed by Raman spectroscopy. The measured surface and chips were generated by OC and EVC with an amplitude of 1 µm. Fig. 8(a) shows the Raman spectra of the surface, with peaks at Raman phase shifts of 141.6 cm-1, 207.1 cm-1, and 253.2 cm-1 for the initial polished surface, corresponding to the transverse acoustic (2TA) phonon mode, transverse optical (TO), and longitudinal optical (LO) phonon modes of ZnSe, respectively. From Fig. 8(a), it can be observed that there was no peak generation or disappearance in the spectra of the surfaces generated by OC and EVC, which proves that there is no phase transformation was detected on the machined surfaces. Fig. 8(b) shows the Raman spectrum of the collected chips. It can be observed that the 2TA phonon mode disappeared in the spectra of the chips generated by EVC, while the TO mode split at 201.4 cm-1, the rest TO phonon mode shifted to 247.4 cm-1, and the LO photon mode shifted to 279.9 cm-1. The results indicate that the phase of the chips transformed from zinc blende to rock salt structure under the pressure of the tool edge, which is consistent with the phase transition path of ZnSe in the condition of high pressure [16]. There is a significant difference between the phase transformation of the surface and the chips because most of the materials which transform induced by the hydrostatic stress of the tool edge are considered to reverse after pressure unloading, leaving a thin rock salt structure layer that cannot be detected by Raman spectroscopy [7]. In contrast, the chips are applied shear stress by the tool instead of hydrostatic stress, which maintains the chips in a phase transformation state after leaving the substrate. It is worth noting that the phase transformation is not detected in the chips generated by OC, because the collected chips fragment during the cutting process, without the continuous shearing stress applied by the tool, thus no phase transformation occurs.
3.2 Influence of elliptic trajectory inclination on removal characteristics
The phase shift between the vibration signal along the nominal cutting direction and the vibration signal along the DoC direction determines the inclination of the elliptical trajectory. As shown in Fig. 9, where a is half the length of the ellipse minor axis, b is half the length of the ellipse major axis, and θ is the inclination of the elliptical trajectory, which is defined as the angle between the minor axis and the nominal cutting direction. A phase shift of -90° is adopted in the common case, thus the obtained elliptic trajectory inclination is 90°. The elliptical trajectory with arbitrary inclination can be obtained by controlling the phase shift.
Fig. 9. Inclination of elliptic trajectory. (a) Schematic diagram of the elliptic trajectory inclination; (b) and (c): Vibration trajectory at inclination angles of 70° and 110°; (d) and (e): Schematic diagram of EVC with inclination angles of 70° and 110°.
To investigate the influence of elliptic trajectory inclination on the machinability of ZnSe, the elliptic trajectory inclinations were set to 70° and 110° for the grooving experiments. Based on the results in section 3.1, the nominal cutting speed was kept at 200 mm/min, and the amplitude along the nominal cutting direction and the nominal DoC direction were kept at 4 µm and 2 µm, respectively. The phase shift was adjusted and the vibration trajectory of the tool was measured by a laser doppler vibrometer. The elliptical trajectory inclination of 70° and 110° were measured as shown in Fig. 9(b) and Fig. 9(c).
The morphology of the grooves and chips generated by different elliptic trajectory inclinations was shown in Fig. 10. The critical DoC was 370 nm at an elliptic trajectory inclination of 70°, with a decrease compared to the 90° elliptic trajectory inclination. As reported in the previous research [28], the maximum uncut chip thickness in a single vibration cycle with an elliptical trajectory inclination of 70° is larger than that at a 90° elliptical trajectory inclination. In addition, the drawing effect of the tool on the material under the vibration trajectory is more significant, which aggravates the fracture of the material. Thus, a large number of particle-shaped chips and block-shaped chips were formed during the removal process. In contrast, the maximum uncut chip thickness at an elliptical trajectory inclination of 110° is less than that at a 90° elliptical trajectory inclination. The tool generates more compressive stress on the material at the vibration trajectory, which inhibits crack generation and expansion. The critical DoC achieved 2825 nm at the elliptical trajectory, which is a 94 times improvement compared to OC. Besides, defects along the nominal cutting direction can be observed at the edges of the groove, which are considered to be caused by chip scraping driven by the tool at the vibration trajectory. Large-scale needle-like chips were produced during the cutting process, which further validated the improvement of the ductile-regime machinability.
Fig. 10. Grooves and chips generated by different elliptic trajectory inclinations. (a) Groove and chips generated by an elliptic trajectory inclination of 70°; (b) Groove and chips generated by an elliptic trajectory inclination of 110°.
3.3 Analysis of planning experiments
Planning experiments were carried out to investigate the removal characteristics of ZnSe during continuous feed cutting. The model of undeformed chip thickness during cutting is shown in Fig. 11. The undeformed chip is half-crescent in shape, and the thickness at the tool tip position is minimum and gradually increases along the rounded edge. The maximum undeformed chip thickness is described as [32]:
where tmax is the maximum undeformed chip thickness, R is the radius of the tool nose, d is the DoC, and f is the feed. The cracks start to generate as the maximum undeformed chip thickness exceeds the critical DoC. With further increases in the removal rate, the rise in cutting force leads to an increase in the crack length, which extends to the machined surface and leaves crater defects [33].According to the results in section 3.1 and section 3.2, large critical DoC can be obtained at elliptic trajectory inclinations of 90° and 110°. In this section, the influences of feed on the material removal characteristics at elliptic trajectory inclinations of 90° and 110° were investigated by planning experiments. The nominal cutting speed was kept at 200 mm/min, the DoC was maintained at 3 µm, the feeds were set to 4 µm/rev, 22.7 µm/rev, and 45.8 µm/rev, and the corresponding maximum undeformed chip thicknesses of 302 nm, 1500 nm, and 2500 nm calculated by Eq. (3).
Fig. 12(a) shows the morphology of the surface generated by different feeds at an elliptic trajectory inclination of 90°. At a feed value of 4 µm/rev, the extrusion effect of the tool edge on the material was significant due to the small undeformed chip thickness, and the tensile stress generated by the elastic recovery induced the tearing of the grains. In addition, the low material removal rate led to a long processing time, causing the waviness on the surface induced by the variation of the environment. As the feed increased to 22.7 µm/rev, the grain tearing is eliminated due to the increase in undeformed chip thickness, and the surface was formed by ductile mode cutting. Significant grain boundary step was not observed on the machined surface. It is because EVC decreases the cutting force during machining process and reduces the difference in elastic recovery of each grain. At a feed rate of 45.8 µm/rev, the corresponding undeformed chip thickness was larger than the critical DoC for an elliptic trajectory inclination of 90°, thus leaving craters on the surface caused by cleavage. Fig. 12(b) shows the surface morphology generated by different feeds at an elliptic trajectory inclination of 110°. At the feeds of 4 µm/rev and 22.7 µm/rev, the phenomena were similar to that with an elliptic trajectory inclination of 90°. However, when the feed was increased to 45.8 µm/rev, it was observed that the material could still be removed in the ductile mode and no cleavage-induced fracture occurred. Due to the large scale of feed marks, the machined surface roughness was not substantially reduced compared to the surface formed by an elliptic trajectory inclination of 90°.
Fig. 12. Surface morphology generated by different feeds. (a) Surface generated by an elliptic trajectory inclination of 90°; (b) Surface generated by an elliptic trajectory inclination of 110°.
4. Fabrication and performance verification of ZnSe MLA
MLA element has superior capabilities of beam shaping [34] and comprehensive imaging [35,36] as a basic array optical element, and has a significant function in the miniaturization and lightweighting of optical systems. Based on the fundamental experiment results in section 3, the fabrication and performance verification of the ZnSe MLA was carried out in this section. The specific parameters of the MLA are shown in Table 3.
Table 3. Parameters of the MLA
Fig. 13 shows the structural dimensions of the microlens. The image quality of the unit lens was evaluated by the modulation transfer function (MTF) curve, and the result is shown in Fig. 13(d). It can be seen from the MTF curves that the simulation tangential curve has coincided with the diffraction limit, and the diameter of the imaging point is smaller than the diameter of the airy disc, which indicates that the designed microlens can image at the diffraction limit.
The ZnSe MLA was fabricated by EVC combined with the slow tool servo technique. Based on the results in section 3, the nominal cutting speed was set to 200 mm/min and the elliptical trajectory inclination was set to 110°. The DoC was set to 7 µm to realize the forming of the MLA by one-time cutting. The feed was set to 10µm/rev to avoid material cleavage or grain tearing. In the process of constant cutting speed machining, the spindle speed increases with the feed of the tool from the edge to the center of rotation. To avoid spindle unbalance caused by excessive speed, the maximum spindle speed was set to 140 rpm. Once the spindle reaches the maximum rotation speed, the machining was performed in constant rotation speed mode. The detailed machining parameters are shown in Table 4.
Table 4. Parameters of ZnSe MLA cutting experiment
The machined MLA was observed by WLI, and image stitching was utilized for large-area measurement. The morphology of the machined MLA is shown in Fig. 14 (a), and it can be observed that each microlens has a high uniformity in shape and size. Fig. 14 (b) shows the SEM images of the MLA. Large-scale cracks cannot be observed on the surface of microlenses, which indicates that the cutting was mainly in ductile mode. The microlenses framed by the white box in Fig. 14(a) were observed with high magnification, and the results are shown in Fig. 14(c) and Fig. 14(d). With the influence of vibration trajectory and tool nose radius, the junction between the microlenses and the base plane was smooth, instead of the sharp junction in the ideal state. The section profiles of the microlenses were measured along the cutting direction. The errors of the aperture were less than 1.1 µm (within 0.5%) and the errors of the sag were less than 0.1 µm (within 1.9%). There was a dent in the cut-out of each microlens, which is considered to be caused by the overshoot of the machine tool during the servo-following motion. Shape removal was applied to the measured data of the microlens, and the results are shown in Fig. 14(e) and Fig. 14(f). A small amount of grain tearing exists on the top of the microlens due to the extrusion effect of the tool edge at small DoC. The surface roughness of microlenses was in the range of 7-9 nm, nearly two orders of magnitude smaller than the wavelength of light.
Fig. 14. Morphology of MLA. (a) Large area topography obtained by image stitching; (b) SEM images of the MLA; (c) and (d): Local morphology of microlens 1 and microlens 2. (e) and (f): surface roughness of microlens 1 and microlens 2.
An optical platform was established to measure the imaging performance of the machined MLA, as shown in Fig. 15 (a) and Fig. 15 (b). The source was on the left and the white light that can be emitted to the MLA passed through a mask with an “E-shaped” hole. The MLA imaged the beam passing through the mask in multiple, and the image points were accepted by a complementary metal oxide semiconductor (CMOS) located in the focal plane. An external printed circuit board (PCB) was utilized to process the signal from the CMOS and transmit it to the screen. Fig. 15 (c) shows the result of imaging. Since the lenses were free of aberrations, and crack or grain tearing was not generated during cutting, each lens achieved clear imaging. In addition, due to the low geometric errors of the lenses, the unit images had a high degree of uniformity. The result further proves the feasibility of EVC for the fabrication of ZnSe MLA.
Fig. 15. Multiple imaging tests of MLA. (a) Schematic diagram of the optical platform; (b) The established optical platform; (c) Result of multiple imaging.
5. Conclusions
The removal mechanism of polycrystalline ZnSe by EVC was investigated through grooving and planning experiments, and the machinability of ZnSe MLA by applying EVC was confirmed in the present work. The main summaries are as follows:
- (1) The intermittent form removal of EVC avoids the cracks to expand and promotes the machinability of polycrystalline ZnSe. Needle-shaped chips are generated during the ductile-regime cutting process. Due to the low average cutting force in EVC and the hydrostatic stress applied on the surface, phase transition has not been detected on the machined surface.
- (2) The increase of amplitude along the DoC direction is beneficial to reduce the friction between the tool and the workpiece, but the critical DoC decreases at an amplitude of 4 µm due to the tensile stress during the pulling-up motion of the tool. The reduction of nominal cutting speed increases the loading cycles of the tool on a specific area and suppresses the tensile stress caused by the elastic recovery. In addition, a low nominal cutting speed reduces the tool-workpiece contact time in a single cycle and decreases the average cutting force, thus having a significant effect on the suppression of grain tearing. Compared to the elliptical trajectory inclination of 90°, the instantaneous cutting thickness in a single cycle was reduced and the stress state of the material was improved at an elliptical trajectory inclination of 110°. The critical DoC was increased by 94 times compared to OC by optimizing the vibration trajectory and cutting speed.
- (3) At extremely small feeds, the material mainly occurs to recovery squeezed by the tool edge, leading to significant grain tearing. As the feed is larger than the threshold, the material cleavages due to the excessive undeformed chip thickness, resulting in craters with step-like edges.
- (4) The ZnSe MLA was successfully fabrication by EVC, and the performance of the machined MLA was verified by geometry measurements and an imaging test, which further confirmed the feasibility of EVC in the fabrication of ZnSe optical elements.
Funding
National Natural Science Foundation of China (No. 52188102, No. 52225506); Program for HUST Academic Frontier Youth Team (No. 2019QYTD12).
Disclosures
The authors declare no conflict of interest.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
1. L. Fei, Y. Cui, D. Wan, et al., “Design and fabrication of composite structures in ZnSe providing broadband mid-infrared anti-reflection,” Opt. Mater. (Amsterdam, Neth.) 84, 722–727 (2018). [CrossRef]
2. B. Zollars, S. Savoy, Q. Xue, et al., “Performance measurements of infrared windows with surface structures providing broadband wide-angle antireflective properties,” in Proc SPIE 8, 87080Q (2013). [CrossRef]
3. E. Vilkova, O. Timofeev, S. Nosov, et al., “Variation of the surface roughness of CVD ZnSe undergoing mechanical processing as a function of the grain size of the suspension,” J. Opt. Technol. 80(9), 577 (2013). [CrossRef]
4. H. Xiao, S. Yin, C. F. Cheung, et al., “Material removal behavior analysis of ZnSe crystal during side-forward nanoscratching,” Int. J. Mech. Sci. 241, 107968 (2023). [CrossRef]
5. K. Wang, J. Chen, W. Zhou, et al., “Direct Growth of Highly Mismatched Type II ZnO/ZnSe Core/Shell Nanowire Arrays on Transparent Conducting Oxide Substrates for Solar Cell Applications,” Adv. Mater. (Weinheim, Ger.) 20(17), 3248–3253 (2008). [CrossRef]
6. X. Wu, F. Meng, Z. Zhang, et al., “Broadband down-conversion for silicon solar cell by ZnSe/phosphor heterostructure,” Opt. Express 22(S3), A735 (2014). [CrossRef]
7. W. Huang and J. Yan, “Effect of tool geometry on ultraprecision machining of soft-brittle materials: A comprehensive review,” Int. J. Extrem. Manuf. 5(1), 012003 (2023). [CrossRef]
8. E. Vilkova and O. Timofeev, “Mechanical polishing of ZnSe using rosin-based resins,” Inorg. Mater. 46(3), 264–268 (2010). [CrossRef]
9. V. Tomashyk, A. Kravtsova, Z. Tomashyk, et al., “Optimization of conditions for treatment of ZnSe crystal surfaces by chemical etching,” Fiz. Napivprovidn., Kvantova Optoelektron.(1998) 16(2), 140–145 (2013). [CrossRef]
10. L. Y. J. Lee, Y.-K. Shen, and H. Wang, “Suppression of Polycrystalline Diamond Tool Wear with Mechanochemical Effects in Micromachining of Ferrous Metal,” J. Manuf. Mater. Process. 4(3), 81 (2020). [CrossRef]
11. S. Zhang and W. Zong, “Micro Defects on Diamond Tool Cutting Edge Affecting the Ductile-Mode Machining of KDP Crystal,” Micromachines 11(12), 1102 (2020). [CrossRef]
12. W. Huang and J. Yan, “Fundamental Investigation of Diamond Cutting of Micro V-Shaped Grooves on a Polycrystalline Soft-Brittle Material,” J. Manuf. Mater. Process. 5(1), 17 (2021). [CrossRef]
13. F. Z. Fang and L. J. Chen, “Ultra-Precision Cutting for ZKN7 Glass,” CIRP Ann. 49(1), 17–20 (2000). [CrossRef]
14. F. Fang, V. Venkatesh, and G. Zhang, “Diamond Turning of Soft Semiconductors to Obtain Nanometric Mirror Surfaces,” International Journal of Advanced Manufacturing Technology 19(9), 637–641 (2002). [CrossRef]
15. S. A. Shojaee, T. A. Harriman, Y. Qi, et al., “Spatial variations in stress and crystal quality in diamond turned ZnSe surfaces measured by Raman spectroscopy,” Manufacturing Letters 2(2), 35–39 (2014). [CrossRef]
16. S. Ono, “Phase transition in ZnSe at high pressures and high temperatures,” J. Phys. Chem. Solids 141, 109409 (2020). [CrossRef]
17. W. Huang and J. Yan, “Surface formation mechanism in ultraprecision diamond turning of coarse-grained polycrystalline ZnSe,” International Journal of Machine Tools and Manufacture 153, 103554 (2020). [CrossRef]
18. R. Geng, X. Yang, Q. Xie, et al., “Ultra-precision diamond turning of ZnSe ceramics: Surface integrity and ductile regime machining mechanism,” Infrared Phys. Technol. 115, 103706 (2021). [CrossRef]
19. W.-h. Li, K. Yang, P. Wang, et al., “Experimental investigation of the ultra-precision turning capability of PVD ZnSe,” in Proc SPIE (Suzhou, China, 2016), p. 968319.
20. D. E. Brehl and T. A. Dow, “Review of vibration-assisted machining,” Precis. Eng. 32(3), 153–172 (2008). [CrossRef]
21. E. Shamoto and T. Moriwaki, “Study on Elliptical Vibration Cutting,” CIRP Ann. 43(1), 35–38 (1994). [CrossRef]
22. C. Nath, M. Rahman, and K. S. Neo, “Modeling of the effect of machining parameters on maximum thickness of cut in ultrasonic elliptical vibration cutting,” Journal of Manufacturing Science and Engineering 133(1), 011007 (2011). [CrossRef]
23. W.-L. Zhu, Y. He, K. F. Ehmann, et al., “Modeling of the effects of phase shift on cutting performance in elliptical vibration cutting,” The International Journal of Advanced Manufacturing Technology 92(9-12), 3103–3115 (2017). [CrossRef]
24. J. Zhang, N. Suzuki, Y. Wang, et al., “Fundamental investigation of ultra-precision ductile machining of tungsten carbide by applying elliptical vibration cutting with single crystal diamond,” J. Mater. Process. Technol. 214(11), 2644–2659 (2014). [CrossRef]
25. J. Zhang, J. Zhang, T. Cui, et al., “Sculpturing of single crystal silicon microstructures by elliptical vibration cutting,” Journal of Manufacturing Processes 29, 389–398 (2017). [CrossRef]
26. J. Zhang, J. Zhang, A. Rosenkranz, et al., “Frictional properties of surface textures fabricated on hardened steel by elliptical vibration cutting,” Precis. Eng. 59, 66–72 (2019). [CrossRef]
27. J. Zhang, J. Zhang, C. Liu, et al., “Machinability of single crystal calcium fluoride by applying elliptical vibration diamond cutting,” Precis. Eng. 66, 306–314 (2020). [CrossRef]
28. J. Wang, W.-H. Liao, and P. Guo, “Modulated ultrasonic elliptical vibration cutting for ductile-regime texturing of brittle materials with 2-D combined resonant and non-resonant vibrations,” Int. J. Mech. Sci. 170, 105347 (2020). [CrossRef]
29. E. Shamoto, N. Suzuki, and R. Hino, “Analysis of 3D elliptical vibration cutting with thin shear plane model,” CIRP Annals 57(1), 57–60 (2008). [CrossRef]
30. M. Arif, Z. Xinquan, M. Rahman, et al., “A predictive model of the critical undeformed chip thickness for ductile–brittle transition in nano-machining of brittle materials,” International Journal of Machine Tools and Manufacture 64, 114–122 (2013). [CrossRef]
31. W. Huang and J. Yan, “Towards understanding the mechanism of vibration-assisted cutting of monocrystalline silicon by cyclic nanoindentation,” Journal of Materials Processing Technology 311, 117797 (2023). [CrossRef]
32. T. Moriwaki, E. Shamoto, and K. Inoue, “Ultraprecision Ductile Cutting of Glass by Applying Ultrasonic Vibration,” CIRP Annals 41(1), 141–144 (1992). [CrossRef]
33. W. S. Blackley and R. O. Scattergood, “Ductile-regime machining model for diamond turning of brittle materials,” Precision Engineering 13(2), 95–103 (1991). [CrossRef]
34. H. Tiziani, R. Achi, and R. N. Krämer, “Chromatic confocal microscopy with microlenses,” J. of Modern Optics 43(1), 155–163 (1996). [CrossRef]
35. T. Zhang, P. Li, H. Yu, et al., “Fabrication of flexible microlens arrays for parallel super-resolution imaging,” Applied Surface Science 504, 144375 (2020). [CrossRef]
36. M. Martinez-Corral and B. Javidi, “Fundamentals of 3D imaging and displays: A tutorial on integral imaging, light-field, and plenoptic systems,” Adv. Opt. Photon. 10(3), 512 (2018). [CrossRef]
