- Email: [email protected]
Vibration assisted wire electrochemical micro machining of array micro tools
Accepted Manuscript Title: Vibration assisted wire electrochemical micro machining of array micro tools Author: Kun Xu Yongbin Zeng Peng Li Di Zhu PII: DOI: Reference:
S0141-6359(16)30264-1 http://dx.doi.org/doi:10.1016/j.precisioneng.2016.10.004 PRE 6472
To appear in:
Precision Engineering
Received date: Revised date: Accepted date:
20-6-2016 27-9-2016 7-10-2016
Please cite this article as: Xu Kun, Zeng Yongbin, Li Peng, Zhu Di.Vibration assisted wire electrochemical micro machining of array micro tools.Precision Engineering http://dx.doi.org/10.1016/j.precisioneng.2016.10.004 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Vibration assisted wire electrochemical micro machining of array micro tools
Kun Xu, Yongbin Zeng*, Peng Li, Di Zhu
(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics & Astronautics, Nanjing 210016) TEL:86-25-84896601 Fax:86-25-84895912 E-mail:
[email protected],
[email protected](corresponding
[email protected], [email protected]
1
author),
Highlights
The influence of bubble behaviour on electric field was studied through simulations.
By using electrodes vibration, bubbles were driven out of the machining gap quickly.
Optimal machining parameters were obtained though experimental study.
Micro square columnar tool arrays with good shapes were fabricated.
The micro dimple array was fabricated by employing tool array as the cathode.
Abstract: Micro structures and components are widely used in modern industries, and micro machining has therefore become a popular research topic. As micro tools are essential in micro machining, wire electrochemical micro machining is introduced in the fabrication of micro tools in this paper, and micro square column tool arrays are fabricated using wire cathodes by two steps. In order to improve the machining efficiency and quality, an electrode vibration technique is used, and the effects of bubble behaviour on slit width homogeneity and edge radius are studied through simulations of the electric field. The influences of various machining parameters such as vibration conditions, electrical properties, electrolyte concentration and feedrate on the standard deviation of the slit width and on the value of the edge radius are investigated. In addition, the micro dimple array is fabricated using electrochemical
2
micro machining by employing the micro square column tool array as the cathode. Keywords: Micro tool array; Electrodes vibration; Bubble behavior; Wire electrochemical micro machining 1. Introduction Micro machining technology has been widely used in information technology [1], aerospace industries, automotive engines, biomedical industries, micro-fluidic systems and micro electromechanical systems [2]. To meet different continuously improving requirements, including higher production efficiency, better machining accuracy, smooth and burr free surface and so on, many micro machining approaches have been developed, such as laser beam machining [3], micro milling [4], micro electrochemical machining [5], micro electro discharge machining [6] and micro electrochemical discharge machining [7]. Among them, micro electrochemical machining is receiving considerable attention as it can be thought of a controlled anodic dissolution at atomic level, and can produce smooth and burr-free surfaces without tool wear. Different process methods have been studied, such as micro electrochemical milling, micro electrochemical drilling [8] and wire electrochemical micro machining (WEMM) [9]. In order to obtain high machining accuracy, micro tools with small sizes and high precisions are necessary. As the fabrication of micro tools is essential, it has been a popular research topic in recent years. Efforts have been made to fabricate micro tools using wire electro discharge grinding (WEDG), micro electro discharge machining and micro 3
electrochemical machining. Sheu developed a twin-wire electro discharge machining system, and fabricated Ø20 μm micro tools with rough and finish machining on the same machine [10]. Morgan produced polycrystalline diamond tools made using micro EDM [11]. Fan and Hourng discussed the effects of rotation of an anodic tool on the diffusion layer and rate of dissolution with increasing rotational speed, and cylindrical microelectrodes of diameter 100 μm were prepared [12]. Cabrera fabricated ultra-thin microtools with a final diameter typically ranging from 3 to 30 μm using electrochemical etching [13]. Ghoshal and Bhattacharyya studied the influence of vibration on micro tool fabrication by electrochemical machining, and a micro tool of tip diameter 1 μm and tip angle of 11.59° was fabricated [14]. Wang prepared
cylindrical
nanoprobes
using
a
vibrating
liquid
membrane
by
electrochemical etching [15]. It is therefore apparent that previous efforts have been made to control the size and accuracy of micro tools. Considering production efficiency and process characteristics, fabrication of micro tool arrays is more difficult than that of single tools. Green prepared micro electrode arrays using electron beam lithography for biotechnological applications [16].
Hu
fabricated
high-aspect-ratio
electrode
arrays
by
combining
ultraviolet-lithographie, galanoformung and abformung (LIGA) process with micro electro discharge machining [17]. Kitamura fabricated a micro-needle array in two steps: arraying titanium wires and processing wires with electrochemical etching [18]. In this paper, WEMM, as an efficient and convenient approach, is introduced for micro square column tool array fabrication. 4
Bubbles produced in the machining area in WEMM can significantly affect the machining process. The influences of bubble behaviour on the slit width homogeneity and edge radius were studied through simulations of the electric field. In order to improve the diffusion of bubbles, electrode vibration was used to improve the machining accuracy and feedrate. The influences of cathode vibration speed and amplitude, anode vibration frequency and amplitude, electrical parameters, electrolyte concentration and feedrate on the slit width homogeneity and edge radius were experimentally investigated. Optimal machining parameters for micro square column tool array fabrication when using single wire were obtained through slit machining experiments, and micro square column tool arrays were prepared. In addition, multiple wire cathodes were used to prepare micro square column tool arrays in order to improve production efficiency. Finally, micro dimple arrays were fabricated using EMM by employing the micro square column tool array as the cathode. 2. Electrochemical machining of micro square column tool arrays To obtain good shape and accuracy of micro square column tool arrays, slit width homogeneity and a small edge radius are preferred, as shown in Fig. 1. In WEMM, bubbles and insoluble product produced and accumulated in the machining gap always lead to poor slit width homogeneity and a large edge radius. 2.1 Analysis of electric field and bubble behaviour during WEMM Bubble behaviour significantly influences current distribution in the machining gap and at the edge. In WEMM, bubbles produced in the machining area will move to the edge area and then out of the machining gap by diffusion. Therefore, a simplified 5
two-dimensional electrical model was set up as shown in Fig. 2. Fig. 2(a) shows a cross section of the machining area with the bubble 1# in the machining gap, close to point A1. In the side section, three bubbles in different locations are in the machining area as shown in Fig. 2(b). The bubble 2# is near to the edge and point B1. The bubble 3# is almost in the middle of the machining gap, and so may have an effect on the current distribution at point B2. Point B3 is far away from all the bubbles, and it is considered that the current distribution at this edge is not affected by bubbles. The bubble 4# is on the wire cathode just outside the machining gap, and the current distribution at point B4 will be affected by it. The study of current distribution in WEMM was carried out on the basis of three assumptions. The first was that the electrical parameters were stable. The second was that the insoluble product was evenly distributed in the machining gap and the conductivity of the electrolyte remained constant. The third was that the simulation was a static process without any deformation on the slit boundary. This meant that the accumulation of bubbles was a unique factor affecting the current distribution. According to the theory of static electric fields, the potential distribution in the domain satisfies Laplace’s equation:
2=0
(1)
The boundary conditions were as follows:
1 =0 V (at the wire cathode)
(2)
2 =5 V (at the anode surface)
(3)
6
3* 4 =0 (the boundary condition) n
(4)
The calculation was performed using the finite element method via COMSOL version 5.0. The parameters used in the calculations are listed in Table 1. Fig. 3 shows the current density distribution in the machining gap of WEMM, with bubbles present. The contours show that the bubbles significantly influence the current density distribution in the machining gap. As shown in Fig. 3(a), when a bubble remains in the machining gap, the current density on the anode surface near the bubble is much smaller than that on the other part of anode surface in the machining gap. For example, it was 2.66 A/mm2 at point A1 and 8.64 A/mm2 at point A2. Therefore, the dissolution rate of the anode at A1 was slower than that at A2, and the homogeneity of the slit width was poor. Moreover, accumulation of bubbles could even result in wire deformation and electrical short circuits, which also lead to poor homogeneity of the slit width. As shown in Fig. 3(b), the current density on the edge of the slit was larger than that on the surface in the machining gap. The current densities at points B1, B2, B3 and B4 were 19.844, 14.561, 14.371 and 14.055 A/mm2. The value at point B1 was the largest, which indicates that when the bubble was near the edge in the machining gap, the edge radius would be large. When the bubble was in the middle of the machining gap, the current density at point B2 was slightly larger than that at point B3, and the edge radius at point B2 would still be large. However, when the bubbles were driven out of the machining gap, the current density at point B4 was the lowest, and the edge radius would be small. Therefore, bubbles should be driven out of the 7
machining gap quickly in order to obtain a small edge radius. 2.2 Analysis of mass transport with electrode vibration Vibrations of electrodes obviously have an influence on mass transport in the machining gap. Cathode vibration was driven by the reciprocating movement of the z-axis, and a PZT (piezoelectric ceramics) was used to drive the anode vibration. Ignoring the transient acceleration process of the cathode, the process of electrode vibration is shown in Fig. 4. As shown in Fig. 5, the electrodes move upwards and downwards in the same path periodically, the pressure of the electrolyte in the machining gap first drops and then increases following vibration, and then micro bubbles are generated and collapse, which results in the enhanced convective mass transport of dissolved ions, insoluble product and supply of fresh electrolyte [14]. The insoluble product distributed evenly in the machining gap due to the anode vibration; the cathode vibration may drive bubbles and insoluble product out and bring fresh electrolyte into the machining gap in each vibration period, and the accumulation of insoluble product is thus reduced. As shown in Fig. 6(a), the process might not continue due to wire defamation without the use of vibration. When the cathode or anode vibrates, the machining process becomes stable, although the efficiency of mass transport still needs to be improved and large particles of insoluble product are produced because of the accumulation, especially with anode vibration only, which can be observed in Figs 6(b) and (c). With vibration of both cathode and anode, the mass transport was enhanced significantly, bubbles were driven out from the machining gap quickly and the 8
insoluble product was driven out before large particle formation, as shown in Fig. 6(d). Some insoluble product always adheres to the surface of the wire cathode after machining. As shown in Fig. 7, the surface of the wire cathode in the machining area becomes darker after machining, and insoluble product adheres to the wire cathode more with anode vibration only, which leads to poor machining results. The machining parameters were set to the following: 0.3 μm·s-1 feedrate, 5 V voltage, 50 ns pulse duration and 6 μs pulse period for electrical parameters, 400 μm·s-1 speed and 100 μm amplitude for cathode vibration (if applied), and 100 Hz frequency and 5 μm amplitude for anode vibration (if applied).
2.3 Experimental set up Fig. 8 shows the experimental set up for slit machining experiments and micro tool fabrication using WEMM, which includes a PC, a three-axis stage with a 0.1μm resolution, a PZT and its controller, an electrolyte slot, a pulse generator and a charge-coupled device (CCD) camera. As the power supply, the pulse generator provided ultrashort pulses for WEMM. The wire cathode was attached to a three-axis stage, then cathode vibration was achieved when the z-axis moved up and down and the x–y-axis could drive the feed movement. Anode vibration took place when pulses produced by the PZT controller charged the PZT, and its direction was parallel to the z-axis. In order to observe the machining process and capture images, a CCD with 7.5Hz frame rate was used. 2.4 Experimental details 9
The designed experimental set up was used for micro tool fabrication and electrochemical micromachining purposes. In order to obtain micro tools with a good shape, slit machining experiments were developed, and the influences of predominant parameters such as vibration condition, pulse conditions, electrolyte concentration and feedrate on homogeneity of slit width and edge radius were studied. Small pieces of cobalt-based alloy of thickness 80 μm were used as the workpiece, and tungsten wire of diameter 10 μm diameter was used as the cathode electrode. The slit width and edge radius were measured using a digital microscope (DVM5000, Leica, Germany). Nine measurements were taken for each slit, and the standard deviation of the slit width was calculated. During machining, the process was observed and recorded by a CCD camera. Optimal machining conditions for the fabrication of micro square column tool arrays were obtained, and micro tools were prepared and were used as cathode tools in EMM operation for the machining of micro dimples. Dilute hydrochloric acid at a concentration of 0.05 mol·L-1 was preferred as the electrolyte because acid electrolytes usually produce less insoluble electrolysis products than common salt electrolytes. Ultrashort pulses of voltage 7 V, duration 90 ns and period 6 μs, and an anode vibration of frequency and amplitude of 500 Hz and 1 μm, were adopted in the machining process. 3. Results and discussions 3.1 Effects of electrode vibration Vibration of electrodes has a significant influence on the enhancement of mass 10
transport. Due to the different driving modes, different vibration parameters are discussed for the cathode and anode separately. The cathode vibration was driven by the z-axis, which could achieve large amplitudes and high speeds. On the other hand, high frequencies and small amplitudes were used for anode vibration, which was driven by a PZT. The following parameters were used for the slit experiments: 5 V voltage, 6 μs period and 50 ns duration for the pulses, 0.05 mol·L-1 electrolyte concentration and 0.5 μm·s-1 feedrate. 3.1.1 Influence of amplitude of cathode vibration
Cathode vibration produces hydrodynamic effects on mass transport, which are affected by the amplitude of vibration. Even though the vibration speed was fixed, an increase of amplitude led to a small standard deviation of the slit width and a small edge radius when the amplitude was less than 200 μm, as shown in Fig. 9, Fig. 10(a) and (b). As the wire cathode drives the viscous electrolyte flow when vibrating, the bubbles move, following the cathode and electrolyte. When the amplitude was less than twice the thickness of the anode, the larger amplitude meant that the bubbles and insoluble product had a greater chance of being removed out of the machining gap, and so the homogeneity of the slit width and edge radius would be improved. When the amplitude was greater than 200 μm, the standard deviation of the slit width and the value of the edge radius remained small, as shown in Fig.10(c). Thus, a large amplitude of cathode vibration is preferred for micro tool fabrication. 3.1.2 Influence of speed of cathode vibration 11
Cathode vibration speed significantly affects bubble behaviour, and has been utilized for effective removal of bubbles and insoluble product, and for refreshing electrolyte [14]. As shown in Fig. 11, Fig. 10(a) and Fig. 10(b), when the speed was less than 400 μm·s-1, the standard deviation of the slit width and the value of the edge radius decreased as the speed increased. With a continuous increase of speed, the values remained small but began to increase for edge radius, as shown in Fig. 10(d). As the amplitude was fixed, an increase of speed meant an increase of frequency, and too large a frequency could reduce the efficiency of mass transport. With a high frequency, the vibration period was short, and then the bubbles that were removed out of the machining gap had a short time available for diffusion and could be driven back into the machining gap. Thus, a speed of cathode vibration of 400 μm·s-1 is considered to be appropriate for micro tool fabrication when the amplitude of cathode vibration is fixed to 200 μm. 3.1.3 Influence of amplitude of anode vibrationAs it was driven by a PZT, the anode vibration achieved a higher frequency and smaller amplitude than the cathode vibration. The influence of anode vibration amplitude on the standard deviation of the slit width and on the value of the edge radius are shown in Fig. 12. As shown in Fig. 13(a) and (b), without anode vibration, the standard deviation of the slit width and the value of the edge radius were 0.74 μm and 3.35 μm, respectively, which then decreased to 0.28 μm and 2.34 μm when the amplitude increased from 0 to 5 μm. When the vibration frequency was fixed, a large amplitude meant a high speed and a high efficiency of mass transport. However, as bubbles were produced on the surface 12
of the wire cathode, anode vibration had a limited effect on bubble behaviour. The standard deviation of the slit width and the value of the edge radius remained almost constant regardless of the increase of amplitude when the amplitude was more than 5 μm, as shown in Fig. 13(c). Therefore, an amplitude of 5 μm was considered optimal for anode vibration. 3.1.4 Influence of frequency of anode vibration Fig. 14 demonstrates the effect of anode vibration frequency on the standard deviation of the slit width and the value of the edge radius. When the frequency was less than 100 Hz, both the standard deviation of the slit width and the value of the edge radius decreased with increasing frequency, as shown in Fig. 13(a) and (b). Higher frequency with a fixed amplitude meant a higher speed of anode vibration, which caused a greater pressure rise or fall in the machining gap, and the efficiency of mass transport was improved. As for the amplitude of anode vibration, the standard deviation of the slit width and the value of the edge radius remained small, regardless of the increase of frequency when it was greater than 100 Hz, as shown in Fig 13(d). 3.2 Effects of electrical parameters Electrical parameters, such as voltage, pulse period and pulse duration, have effects on the material removal rate and localization. Therefore, the investigation of electrical parameters is important and meaningful. The following parameters were used in the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 0.05 mol·L-1 electrolyte concentration and 0.5 μm·s-1 feedrate. 13
3.2.1 Influence of voltage Voltage is one of the important parameters for material removal rate and localization. A large voltage causes a high removal of material and a large amount of bubbles and insoluble product, which increase the difficulty of mass transport and lead to poor machining. As shown in Fig. 15, it is obvious that the standard deviation of the slit width and the value of the edge radius increased gradually with voltage varying from 5 V to 10 V. Small values of 0.28 μm for the standard deviation of the slit width and 2.34 μm for the value of the edge radius were obtained. When the voltage was increased to 10 V, the values increased to 1.34 μm and 3.01 μm, respectively; Fig 16(a) and (b) show the SEM photos of slits machined using 5V and 10V, it is obvious that the slit in Fig. 16(a) has smaller standard deviation value. 3.2.2 Influence of pulse period Pulse period has a significant effect on mass transport and material removal rate. When the pulse duration is constant, a long period leads to low material removal rate and plenty of time for diffusion of bubbles, which is helpful for obtaining sharp micro tools. Fig. 14 shows the influence of pulse period on the standard deviation of the slit width and the value of the edge radius. When pulse period varied from 2 μs to 6 μs, the standard deviation of the slit width and the value of the edge radius reduced, which could be observed in Fig. 16 (a) and (c). However, when pulse period varied from 6 μs to 8 μs, the standard deviation of the slit width and the value of the edge radius increased with increasing period because of short circuits caused by a too low material removal rate. 14
3.2.3 Influence of pulse duration Fig. 18 shows the effects of pulse duration on the standard deviation of the slit width and the value of the edge radius. Durations that were too short resulted in short circuits due to low material removal rate, and large values of 0.40 μs for the standard deviation of the slit width and 3.13 μs for the edge radius were obtained when the pulse duration was 40 ns. The two values were 0.28 μs and 2.34 μs when the pulse duration was 50 ns. With further increases of pulse duration, the values rose, as a long duration meant a high material removal rate and large amounts of bubbles produced in the machining gap. Fig 16 (d) shows the SEM photo of the slit machined using duration of 90ns, which has larger standard deviation value of the slit width than that of the slit in Fig. 16(a). 3.3 Effects of electrolyte concentration Fig. 19 describes the variation of the standard deviation of the slit width and the value of the edge radius with electrolyte concentration. The following machining parameters were used for the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse and 0.5 μm·s-1 feedrate. As shown in Fig. 19 and Fig. 20, the standard deviation of the slit width decreased with increasing concentration, but the edge radius increased with increasing concentration. A high concentration led to stable machining and wide machining gap, and then the homogeneity of the slit width improved. However, the electric field intensity outside the machining gap was strong and a large amount of bubbles was 15
produced in the machining gap with a high concentration, which resulted in a large value of the edge radius. Thus, an optimal electrolyte concentration of 0.05 mol·L-1 was chosen to fabricate the micro tools. 3.4 Effects of federate It is obvious from Fig. 21 and Fig. 22 that a high feedrate improves the homogeneity of the slit width and the edge radius. The following machining parameters were used in the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse and 0.05 mol·L-1 electrolyte concentration. A high feedrate reduces secondary corrosion, which is helpful for improving the machining accuracy and the homogeneity of the slit width and the edge radius. The standard deviation of the slit width decreased from 0.41 μm to 0.28 μm and the value of the edge radius decreased from 3.74 μm to 2.34 μm when the feedrate increased from 0.05 μm·s-1 to 0.5 μm·s-1. Thus, a high feedrate is preferred for micro tool fabrication.
4. Fabrication of micro square column tool arrays using a single wire cathode From the earlier analysis, the following optimal machining parameters were used in the micro tool fabrication: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse, 0.05 mol·L-1 electrolyte concentration and 16
0.5 μm·s-1 feedrate. The machining process for micro tools was divided into two steps, as shown in Fig. 23. Micro beams were produced in step 1, as shown in Fig. 24(a) and (b), and then in step 2, micro square columnar tool arrays were fabricated, as shown in Fig. 24(c) and (d). Each micro tool was 140 μm long with a section size of 10 ×10 μm2. A micro tool was randomly selected to measure the homogeneity of the tool width and edge radius; 0.30μm and 2.54 μm was obtained for standard deviation of the tool width and the edge radius separately for the tool in Fig. 24(c).
5. Fabrication of micro tool arrays using multiple wire cathodes In order to improve production efficiency, multiple wire cathodes were used in micro tool array fabrication in WEMM, and the feed path changed in step 1 as shown in Fig. 25. As the machining process needed more energy, the electrical parameters were adjusted to 6 V for voltage, 5 μs for period and 90 ns for duration in step 1, and then micro beams were produced, as shown in Fig. 26(a). As the processing time was short, a single wire cathode was still used in step 2; a fabricated micro square columnar tool array is shown in Fig. 26(b). 6. EMM application of micro tools As shown in Fig. 27, the Micro dimple array were electrochemically fabricated using micro square columnar tool array as the cathode. As demonstrated in Fig. 27(b), measurements were taken on the dimples: the depths were 15.9 μm with entrance size of 25.5×24.6 μm. 7. Conclusions 17
On the basis of the results presented above, the following conclusions can be drawn: (i) Diffusion of electrolytic products such as bubbles is difficult in WEMM, and the influence of bubble behaviour on the electric field distribution in the machining area has been studied through simulations of the electric field. Bubbles in the machining gap always cause poor homogeneity of slit width and a large edge radius. (ii) The bubbles were driven out of the machining gap quickly, and accumulation of insoluble product reduced with vibration of electrodes. Large amplitudes and high speeds of cathode vibration, and large amplitudes and high frequencies of anode vibration, reduced the edge radius and improved the homogeneity of the slit width. (iii) For stable machining, pulses with low voltages, long periods and short durations and high feedrates improved the homogeneity of the slit width and reduced the edge radius. High concentrations of electrolyte improved the homogeneity of the slit width, and low concentrations of electrolyte reduced the edge radius. (iv) Using optimal machining parameters, micro square columnar tool arrays with a small edge radius and a good tool width homogeneity were fabricated. The micro dimple array was fabricated using EMM by employing the micro tool array as cathodes.
Acknowledgement 18
This work was conducted under the sponsorship of the National Natural Science Foundation of China (51375238), the Jiangsu Natural Science Foundation (BK20131361), the Funding of Jiangsu Innovation Program for Graduate Education (CXZZ13_0153) and a project funded by the China Postdoctoral Science Foundation (2015T80546).
References [1] Xue B, Yan Y, Li J, Yu B, Hu Z, Zhao X, Cai Q. Study on the micro-machining process with a micro three-sided pyramidal tip and the circular machining trajectory. J Mater Process Technol 2014; 217: 122–130. [2] Ghoshal B, Bhattacharyya B. Vibration assisted electrochemical micromachining of high aspect ratio micro features. Precis Eng –J Int Soc Precis Eng 2015; 42: 231–241. [3] Zeng YB, Yu Q, Wang SH, Zhu D. Enhancement of mass transport in micro wire electrochemical machining. CIRP Ann-Manuf Technol 2012; 61 (1): 195–198. [4] Rodríguez P, Labarga JE. Tool deflection model for micromilling processes. Int J of Adv Manuf Technol 2014; 76 (1–4): 199–207. [5] Koyano T, Kunieda M, Ultra-short pulse ECM using electrostatic induction feeding method. Procedia CIRP 2013; 6: 390–394. [6] Nguyen MD, Wong YS, Rahman M. Profile error compensation in high precision 3D micro-EDM 19
milling. Precis Eng –J Int Soc Precis Eng 2013; 37 (2): 399–407. [7] Chung DK, Hong SS, Min SP, Bo HK, Chong NC. Recent researches in micro electrical machining. Int J Precis Eng Man 2011; 12 (2): 371–380. [8] Kim BH, Na CW, Lee YS, Choi DK, Chu CN. Micro electrochemical machining of 3D micro structure using dilute sulfuric acid. CIRP Ann-Manuf Technol 2005; 54 (1): 191–194. [9] Zhu D, Wang K, Qu NS. Micro wire electrochemical cutting by using in situ fabricated wire electrode. CIRP Ann-Manuf Technol 2007; 56 (1): 241–244. [10] Sheu DY. High-speed micro electrode tool fabrication by a twin-wire EDM system. J Micromech Microeng 2008; 18 (10): 2067–2070. [11] Morgan C, Vallance R, Marsh E. Micro machining glass with polycrystalline diamond tools shaped by micro electro discharge machining. J Micromech Microeng 2004;14 (12): 1687-1692. [12] Fan ZW, Hourng LW. The analysis and investigation on the microelectrode fabrication by electrochemical machining. Int J Mach Tools Manuf 2009; 49 (49): 659–666. [13] Cabrera M, Dahmani R, Layouni Y, Semet V. Micro EDM milling with electrochemical fabrication of ultra-thin microtools and mapping of electrical microdischarges. Procedia CIRP 2016; 42: 650-655. [14] Ghoshal B, Bhattacharyya B. Influence of vibration on micro-tool fabrication by electrochemical machining, Int J Mach Tools Manuf 2013; 64 (1): 49–59. [15] Wang YF, Zeng YB, Qu NS, Zhu D. Note: electrochemical etching of cylindrical nanoprobes using a vibrating electrolyte. Rev Sci Instrum 2015; 86 (7): 106–109. [16] Green NG, Morgan H, Monaghan W, Robertson M. Fabrication of micro-electrode arrays for biotechnological applications. Microelectron Eng 1999; 46 (1–4): 397–400. 20
[17] Hu YY, Zhu D, Li HS, Zeng YB, Ming PM. Fabrication of high-aspect-ratio electrode array by combining UV-LIGA with micro electro-discharge machining. Microsyst Technol 2009; 15 (4): 519–525. [18] Kitamura N, Chim J, Miki N. Electrotactile display using microfabricated micro-needle array. J Micromech Microeng 2015; 25 (2): 25016–25023.
21
Fig. 1 Schematic diagram of the slit width and edge radius Fig. 2 Electric potential domain in the machining gap Fig. 3 Current density distribution around the wire electrodes Fig. 4 Process of electrode vibration Fig. 5 Schematic diagram of the machining process in WEMM Fig. 6 CCD images of the machining process Fig. 7 CCD images of wire cathodes after machining Fig. 8 Experimental set up for micro tool fabrication using WEMM Fig. 9 Cathode vibration amplitude versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 10 SEM photos of slits (anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 11 Cathode vibration speed versus standard deviation of the slit width and the value of edge radius (cathode vibration amplitude: 200 μm, anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 12 Anode vibration amplitude versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, cathode vibration amplitude: 200 μm, anode vibration frequency: 100 Hz) Fig. 13 SEM photos of slits (cathode vibration amplitude: 200 μm, cathode vibration speed: 400 μm·s-1)
22
Fig. 14 Anode vibration frequency versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, cathode vibration amplitude: 200 μm, anode vibration amplitude: 5 μm) Fig. 15 Voltage versus standard deviation of the slit width and the value of edge radius (period: 6 μs, duration: 50 ns) Fig. 16 SEM photos of slits Fig. 17 Pulse period versus standard deviation of the slit width and the value of edge radius (voltage: 5 V, duration: 50 ns) Fig. 18 Pulse duration versus standard deviation of the slit width and the value of edge radius (voltage: 5 V, period: 6 μs) Fig. 19 Electrolyte concentration versus standard deviation of the slit width and the value of edge radius Fig. 20 SEM photos of slits Fig. 21 Feedrate versus standard deviation of the slit width and the value of edge radius Fig. 22 SEM photos of slits Fig. 23 Machining process for micro tools Fig. 24 Micro tools fabricated using a single wire cathode Fig. 25 Step 1 in machining process for micro tools using three wire cathodes Fig. 26 Micro tools fabricated using multiple wire cathodes in step 1 Fig. 27 The micro dimple array fabricated using the micro tool array
23
Figure
Width
Edge radius Anode
Fig. 1
Γ2
A2 y
Electrolyte x Wire
Γ3 Γ1 Γ4 (bubble 1#)
A1
Ω
Anode
(a) cross section Γ4 (bubble 2#)
Γ1 Electrolyte
B1
B2 Anode
Γ2 y Γ4 (bubble 4#) B4 x
Wire
Γ4 (bubble 3#) B3 Ω
(b) side section Fig. 2
Γ3
A/mm2
Y/μm
A2
A1 X/μm
Area-B
B1 Area-A
Area-C
Area-D
Area-A
Y/μm
(a) cross section
B2 Area-B
X/μm Area-C B4
B3 Area-D (b) side section
Fig. 3 .
Wire cathode
Cathode displacement One period
Cathode velocity One period
Amplitude Workpiece anode
Anode displacement One period
Anode velocity One period
Amplitude
Fig. 4
Bubbles
Anode
Wire cathode Fig. 5
Insoluble product
Wire cathode Wire cathode Insoluble product
Bubbles
Anode Anode Bubbles
(a) Without vibration
Bubbles
(b) Only anode vibration
Wire cathode
Wire cathode
Anode
Bubbles
Insoluble product
Anode
Insoluble product
(c) Only cathode vibration Fig. 6
(d) With electrodes vibration
Wire cathode
Wire cathode Wire cathode
Machining area
Insoluble product
(a) only anode vibration Fig. 7
Machining area
Machining area
(b) only cathode vibration
(c) with electrodes vibration
Pulse generator Z
Industrial PC
CCD camera
Y X Electrolyte
PZT controller PZT
Vibration isolation platform Fig. 8
Cathode vibration amplitude / μm Fig. 9
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Without cathode vibration
(b) Cathode vibration speed: 400 μm·s-1, amplitude:200 μm
(c) Cathode vibration speed: 400 μm·s-1, (d) Cathode vibration speed: 600 μm·s-1, amplitude:800 μm μm amplitude:200 Fig. 10 Hz)
Cathode vibration speed / μm Fig. 11
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Anode vibration amplitude / μm Fig. 12
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Without cathode vibration
(c) Anode vibration amplitude: 9 μm, frequency: 100 Hz Fig. 13
(b) Anode vibration amplitude: 5 μm, frequency: 100 Hz
(d) Anode vibration amplitude: 5 μm, frequency: 200 Hz
Anode vibration frequency / μm Fig. 14
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Voltage / V Fig. 15
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Voltage: 5v, period: 6 μs, duration: 50 ns
(b) Voltage: 10v, period: 6 μs, duration: 50 ns
(c) Voltage: 5v, period: 2 μs, duration: 50 ns Fig. 16
(d) Voltage: 5v, period: 6 μs, duration: 90 ns
Period /μs Fig. 17
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Duration / ns Fig. 18
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Concentration / mol·L-1 Fig. 19
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Concentration: 0.02 mol·L-1
(b) Concentration: 0.05 mol·L-1 Fig. 20
(c) Concentration: 0.07 mol·L-1
Feedrate / μm·s-1 Fig. 21
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Feedrate: 0.05 μm·s-1
(b) Feedrate: 0.3 μm·s-1 Fig. 22
(c) Feedrate: 0.5 μm·s-1
Anode
Wire cathode
Feed path Step 1
Step 2 Fig. 23
Micro tools
a
b
c
d
Fig. 24
Wire cathode
Anode
Feed path Fig. 25
a
b
Fig. 26
a
b
Fig. 27
Table 1 Parameters applied in electric field simulation Parameter
Value
Wire cathode diameter/μm
10
Slit width/μm
12
Anode thickness/μm
40
Bubble diameter/μm
0.6/0.6/0.6/2
Electrolyte electrical conductivity/S·m−1
1.83
Applied voltage/V
5
24
S0141-6359(16)30264-1 http://dx.doi.org/doi:10.1016/j.precisioneng.2016.10.004 PRE 6472
To appear in:
Precision Engineering
Received date: Revised date: Accepted date:
20-6-2016 27-9-2016 7-10-2016
Please cite this article as: Xu Kun, Zeng Yongbin, Li Peng, Zhu Di.Vibration assisted wire electrochemical micro machining of array micro tools.Precision Engineering http://dx.doi.org/10.1016/j.precisioneng.2016.10.004 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Vibration assisted wire electrochemical micro machining of array micro tools
Kun Xu, Yongbin Zeng*, Peng Li, Di Zhu
(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics & Astronautics, Nanjing 210016) TEL:86-25-84896601 Fax:86-25-84895912 E-mail:
[email protected],
[email protected](corresponding
[email protected], [email protected]
1
author),
Highlights
The influence of bubble behaviour on electric field was studied through simulations.
By using electrodes vibration, bubbles were driven out of the machining gap quickly.
Optimal machining parameters were obtained though experimental study.
Micro square columnar tool arrays with good shapes were fabricated.
The micro dimple array was fabricated by employing tool array as the cathode.
Abstract: Micro structures and components are widely used in modern industries, and micro machining has therefore become a popular research topic. As micro tools are essential in micro machining, wire electrochemical micro machining is introduced in the fabrication of micro tools in this paper, and micro square column tool arrays are fabricated using wire cathodes by two steps. In order to improve the machining efficiency and quality, an electrode vibration technique is used, and the effects of bubble behaviour on slit width homogeneity and edge radius are studied through simulations of the electric field. The influences of various machining parameters such as vibration conditions, electrical properties, electrolyte concentration and feedrate on the standard deviation of the slit width and on the value of the edge radius are investigated. In addition, the micro dimple array is fabricated using electrochemical
2
micro machining by employing the micro square column tool array as the cathode. Keywords: Micro tool array; Electrodes vibration; Bubble behavior; Wire electrochemical micro machining 1. Introduction Micro machining technology has been widely used in information technology [1], aerospace industries, automotive engines, biomedical industries, micro-fluidic systems and micro electromechanical systems [2]. To meet different continuously improving requirements, including higher production efficiency, better machining accuracy, smooth and burr free surface and so on, many micro machining approaches have been developed, such as laser beam machining [3], micro milling [4], micro electrochemical machining [5], micro electro discharge machining [6] and micro electrochemical discharge machining [7]. Among them, micro electrochemical machining is receiving considerable attention as it can be thought of a controlled anodic dissolution at atomic level, and can produce smooth and burr-free surfaces without tool wear. Different process methods have been studied, such as micro electrochemical milling, micro electrochemical drilling [8] and wire electrochemical micro machining (WEMM) [9]. In order to obtain high machining accuracy, micro tools with small sizes and high precisions are necessary. As the fabrication of micro tools is essential, it has been a popular research topic in recent years. Efforts have been made to fabricate micro tools using wire electro discharge grinding (WEDG), micro electro discharge machining and micro 3
electrochemical machining. Sheu developed a twin-wire electro discharge machining system, and fabricated Ø20 μm micro tools with rough and finish machining on the same machine [10]. Morgan produced polycrystalline diamond tools made using micro EDM [11]. Fan and Hourng discussed the effects of rotation of an anodic tool on the diffusion layer and rate of dissolution with increasing rotational speed, and cylindrical microelectrodes of diameter 100 μm were prepared [12]. Cabrera fabricated ultra-thin microtools with a final diameter typically ranging from 3 to 30 μm using electrochemical etching [13]. Ghoshal and Bhattacharyya studied the influence of vibration on micro tool fabrication by electrochemical machining, and a micro tool of tip diameter 1 μm and tip angle of 11.59° was fabricated [14]. Wang prepared
cylindrical
nanoprobes
using
a
vibrating
liquid
membrane
by
electrochemical etching [15]. It is therefore apparent that previous efforts have been made to control the size and accuracy of micro tools. Considering production efficiency and process characteristics, fabrication of micro tool arrays is more difficult than that of single tools. Green prepared micro electrode arrays using electron beam lithography for biotechnological applications [16].
Hu
fabricated
high-aspect-ratio
electrode
arrays
by
combining
ultraviolet-lithographie, galanoformung and abformung (LIGA) process with micro electro discharge machining [17]. Kitamura fabricated a micro-needle array in two steps: arraying titanium wires and processing wires with electrochemical etching [18]. In this paper, WEMM, as an efficient and convenient approach, is introduced for micro square column tool array fabrication. 4
Bubbles produced in the machining area in WEMM can significantly affect the machining process. The influences of bubble behaviour on the slit width homogeneity and edge radius were studied through simulations of the electric field. In order to improve the diffusion of bubbles, electrode vibration was used to improve the machining accuracy and feedrate. The influences of cathode vibration speed and amplitude, anode vibration frequency and amplitude, electrical parameters, electrolyte concentration and feedrate on the slit width homogeneity and edge radius were experimentally investigated. Optimal machining parameters for micro square column tool array fabrication when using single wire were obtained through slit machining experiments, and micro square column tool arrays were prepared. In addition, multiple wire cathodes were used to prepare micro square column tool arrays in order to improve production efficiency. Finally, micro dimple arrays were fabricated using EMM by employing the micro square column tool array as the cathode. 2. Electrochemical machining of micro square column tool arrays To obtain good shape and accuracy of micro square column tool arrays, slit width homogeneity and a small edge radius are preferred, as shown in Fig. 1. In WEMM, bubbles and insoluble product produced and accumulated in the machining gap always lead to poor slit width homogeneity and a large edge radius. 2.1 Analysis of electric field and bubble behaviour during WEMM Bubble behaviour significantly influences current distribution in the machining gap and at the edge. In WEMM, bubbles produced in the machining area will move to the edge area and then out of the machining gap by diffusion. Therefore, a simplified 5
two-dimensional electrical model was set up as shown in Fig. 2. Fig. 2(a) shows a cross section of the machining area with the bubble 1# in the machining gap, close to point A1. In the side section, three bubbles in different locations are in the machining area as shown in Fig. 2(b). The bubble 2# is near to the edge and point B1. The bubble 3# is almost in the middle of the machining gap, and so may have an effect on the current distribution at point B2. Point B3 is far away from all the bubbles, and it is considered that the current distribution at this edge is not affected by bubbles. The bubble 4# is on the wire cathode just outside the machining gap, and the current distribution at point B4 will be affected by it. The study of current distribution in WEMM was carried out on the basis of three assumptions. The first was that the electrical parameters were stable. The second was that the insoluble product was evenly distributed in the machining gap and the conductivity of the electrolyte remained constant. The third was that the simulation was a static process without any deformation on the slit boundary. This meant that the accumulation of bubbles was a unique factor affecting the current distribution. According to the theory of static electric fields, the potential distribution in the domain satisfies Laplace’s equation:
2=0
(1)
The boundary conditions were as follows:
1 =0 V (at the wire cathode)
(2)
2 =5 V (at the anode surface)
(3)
6
3* 4 =0 (the boundary condition) n
(4)
The calculation was performed using the finite element method via COMSOL version 5.0. The parameters used in the calculations are listed in Table 1. Fig. 3 shows the current density distribution in the machining gap of WEMM, with bubbles present. The contours show that the bubbles significantly influence the current density distribution in the machining gap. As shown in Fig. 3(a), when a bubble remains in the machining gap, the current density on the anode surface near the bubble is much smaller than that on the other part of anode surface in the machining gap. For example, it was 2.66 A/mm2 at point A1 and 8.64 A/mm2 at point A2. Therefore, the dissolution rate of the anode at A1 was slower than that at A2, and the homogeneity of the slit width was poor. Moreover, accumulation of bubbles could even result in wire deformation and electrical short circuits, which also lead to poor homogeneity of the slit width. As shown in Fig. 3(b), the current density on the edge of the slit was larger than that on the surface in the machining gap. The current densities at points B1, B2, B3 and B4 were 19.844, 14.561, 14.371 and 14.055 A/mm2. The value at point B1 was the largest, which indicates that when the bubble was near the edge in the machining gap, the edge radius would be large. When the bubble was in the middle of the machining gap, the current density at point B2 was slightly larger than that at point B3, and the edge radius at point B2 would still be large. However, when the bubbles were driven out of the machining gap, the current density at point B4 was the lowest, and the edge radius would be small. Therefore, bubbles should be driven out of the 7
machining gap quickly in order to obtain a small edge radius. 2.2 Analysis of mass transport with electrode vibration Vibrations of electrodes obviously have an influence on mass transport in the machining gap. Cathode vibration was driven by the reciprocating movement of the z-axis, and a PZT (piezoelectric ceramics) was used to drive the anode vibration. Ignoring the transient acceleration process of the cathode, the process of electrode vibration is shown in Fig. 4. As shown in Fig. 5, the electrodes move upwards and downwards in the same path periodically, the pressure of the electrolyte in the machining gap first drops and then increases following vibration, and then micro bubbles are generated and collapse, which results in the enhanced convective mass transport of dissolved ions, insoluble product and supply of fresh electrolyte [14]. The insoluble product distributed evenly in the machining gap due to the anode vibration; the cathode vibration may drive bubbles and insoluble product out and bring fresh electrolyte into the machining gap in each vibration period, and the accumulation of insoluble product is thus reduced. As shown in Fig. 6(a), the process might not continue due to wire defamation without the use of vibration. When the cathode or anode vibrates, the machining process becomes stable, although the efficiency of mass transport still needs to be improved and large particles of insoluble product are produced because of the accumulation, especially with anode vibration only, which can be observed in Figs 6(b) and (c). With vibration of both cathode and anode, the mass transport was enhanced significantly, bubbles were driven out from the machining gap quickly and the 8
insoluble product was driven out before large particle formation, as shown in Fig. 6(d). Some insoluble product always adheres to the surface of the wire cathode after machining. As shown in Fig. 7, the surface of the wire cathode in the machining area becomes darker after machining, and insoluble product adheres to the wire cathode more with anode vibration only, which leads to poor machining results. The machining parameters were set to the following: 0.3 μm·s-1 feedrate, 5 V voltage, 50 ns pulse duration and 6 μs pulse period for electrical parameters, 400 μm·s-1 speed and 100 μm amplitude for cathode vibration (if applied), and 100 Hz frequency and 5 μm amplitude for anode vibration (if applied).
2.3 Experimental set up Fig. 8 shows the experimental set up for slit machining experiments and micro tool fabrication using WEMM, which includes a PC, a three-axis stage with a 0.1μm resolution, a PZT and its controller, an electrolyte slot, a pulse generator and a charge-coupled device (CCD) camera. As the power supply, the pulse generator provided ultrashort pulses for WEMM. The wire cathode was attached to a three-axis stage, then cathode vibration was achieved when the z-axis moved up and down and the x–y-axis could drive the feed movement. Anode vibration took place when pulses produced by the PZT controller charged the PZT, and its direction was parallel to the z-axis. In order to observe the machining process and capture images, a CCD with 7.5Hz frame rate was used. 2.4 Experimental details 9
The designed experimental set up was used for micro tool fabrication and electrochemical micromachining purposes. In order to obtain micro tools with a good shape, slit machining experiments were developed, and the influences of predominant parameters such as vibration condition, pulse conditions, electrolyte concentration and feedrate on homogeneity of slit width and edge radius were studied. Small pieces of cobalt-based alloy of thickness 80 μm were used as the workpiece, and tungsten wire of diameter 10 μm diameter was used as the cathode electrode. The slit width and edge radius were measured using a digital microscope (DVM5000, Leica, Germany). Nine measurements were taken for each slit, and the standard deviation of the slit width was calculated. During machining, the process was observed and recorded by a CCD camera. Optimal machining conditions for the fabrication of micro square column tool arrays were obtained, and micro tools were prepared and were used as cathode tools in EMM operation for the machining of micro dimples. Dilute hydrochloric acid at a concentration of 0.05 mol·L-1 was preferred as the electrolyte because acid electrolytes usually produce less insoluble electrolysis products than common salt electrolytes. Ultrashort pulses of voltage 7 V, duration 90 ns and period 6 μs, and an anode vibration of frequency and amplitude of 500 Hz and 1 μm, were adopted in the machining process. 3. Results and discussions 3.1 Effects of electrode vibration Vibration of electrodes has a significant influence on the enhancement of mass 10
transport. Due to the different driving modes, different vibration parameters are discussed for the cathode and anode separately. The cathode vibration was driven by the z-axis, which could achieve large amplitudes and high speeds. On the other hand, high frequencies and small amplitudes were used for anode vibration, which was driven by a PZT. The following parameters were used for the slit experiments: 5 V voltage, 6 μs period and 50 ns duration for the pulses, 0.05 mol·L-1 electrolyte concentration and 0.5 μm·s-1 feedrate. 3.1.1 Influence of amplitude of cathode vibration
Cathode vibration produces hydrodynamic effects on mass transport, which are affected by the amplitude of vibration. Even though the vibration speed was fixed, an increase of amplitude led to a small standard deviation of the slit width and a small edge radius when the amplitude was less than 200 μm, as shown in Fig. 9, Fig. 10(a) and (b). As the wire cathode drives the viscous electrolyte flow when vibrating, the bubbles move, following the cathode and electrolyte. When the amplitude was less than twice the thickness of the anode, the larger amplitude meant that the bubbles and insoluble product had a greater chance of being removed out of the machining gap, and so the homogeneity of the slit width and edge radius would be improved. When the amplitude was greater than 200 μm, the standard deviation of the slit width and the value of the edge radius remained small, as shown in Fig.10(c). Thus, a large amplitude of cathode vibration is preferred for micro tool fabrication. 3.1.2 Influence of speed of cathode vibration 11
Cathode vibration speed significantly affects bubble behaviour, and has been utilized for effective removal of bubbles and insoluble product, and for refreshing electrolyte [14]. As shown in Fig. 11, Fig. 10(a) and Fig. 10(b), when the speed was less than 400 μm·s-1, the standard deviation of the slit width and the value of the edge radius decreased as the speed increased. With a continuous increase of speed, the values remained small but began to increase for edge radius, as shown in Fig. 10(d). As the amplitude was fixed, an increase of speed meant an increase of frequency, and too large a frequency could reduce the efficiency of mass transport. With a high frequency, the vibration period was short, and then the bubbles that were removed out of the machining gap had a short time available for diffusion and could be driven back into the machining gap. Thus, a speed of cathode vibration of 400 μm·s-1 is considered to be appropriate for micro tool fabrication when the amplitude of cathode vibration is fixed to 200 μm. 3.1.3 Influence of amplitude of anode vibrationAs it was driven by a PZT, the anode vibration achieved a higher frequency and smaller amplitude than the cathode vibration. The influence of anode vibration amplitude on the standard deviation of the slit width and on the value of the edge radius are shown in Fig. 12. As shown in Fig. 13(a) and (b), without anode vibration, the standard deviation of the slit width and the value of the edge radius were 0.74 μm and 3.35 μm, respectively, which then decreased to 0.28 μm and 2.34 μm when the amplitude increased from 0 to 5 μm. When the vibration frequency was fixed, a large amplitude meant a high speed and a high efficiency of mass transport. However, as bubbles were produced on the surface 12
of the wire cathode, anode vibration had a limited effect on bubble behaviour. The standard deviation of the slit width and the value of the edge radius remained almost constant regardless of the increase of amplitude when the amplitude was more than 5 μm, as shown in Fig. 13(c). Therefore, an amplitude of 5 μm was considered optimal for anode vibration. 3.1.4 Influence of frequency of anode vibration Fig. 14 demonstrates the effect of anode vibration frequency on the standard deviation of the slit width and the value of the edge radius. When the frequency was less than 100 Hz, both the standard deviation of the slit width and the value of the edge radius decreased with increasing frequency, as shown in Fig. 13(a) and (b). Higher frequency with a fixed amplitude meant a higher speed of anode vibration, which caused a greater pressure rise or fall in the machining gap, and the efficiency of mass transport was improved. As for the amplitude of anode vibration, the standard deviation of the slit width and the value of the edge radius remained small, regardless of the increase of frequency when it was greater than 100 Hz, as shown in Fig 13(d). 3.2 Effects of electrical parameters Electrical parameters, such as voltage, pulse period and pulse duration, have effects on the material removal rate and localization. Therefore, the investigation of electrical parameters is important and meaningful. The following parameters were used in the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 0.05 mol·L-1 electrolyte concentration and 0.5 μm·s-1 feedrate. 13
3.2.1 Influence of voltage Voltage is one of the important parameters for material removal rate and localization. A large voltage causes a high removal of material and a large amount of bubbles and insoluble product, which increase the difficulty of mass transport and lead to poor machining. As shown in Fig. 15, it is obvious that the standard deviation of the slit width and the value of the edge radius increased gradually with voltage varying from 5 V to 10 V. Small values of 0.28 μm for the standard deviation of the slit width and 2.34 μm for the value of the edge radius were obtained. When the voltage was increased to 10 V, the values increased to 1.34 μm and 3.01 μm, respectively; Fig 16(a) and (b) show the SEM photos of slits machined using 5V and 10V, it is obvious that the slit in Fig. 16(a) has smaller standard deviation value. 3.2.2 Influence of pulse period Pulse period has a significant effect on mass transport and material removal rate. When the pulse duration is constant, a long period leads to low material removal rate and plenty of time for diffusion of bubbles, which is helpful for obtaining sharp micro tools. Fig. 14 shows the influence of pulse period on the standard deviation of the slit width and the value of the edge radius. When pulse period varied from 2 μs to 6 μs, the standard deviation of the slit width and the value of the edge radius reduced, which could be observed in Fig. 16 (a) and (c). However, when pulse period varied from 6 μs to 8 μs, the standard deviation of the slit width and the value of the edge radius increased with increasing period because of short circuits caused by a too low material removal rate. 14
3.2.3 Influence of pulse duration Fig. 18 shows the effects of pulse duration on the standard deviation of the slit width and the value of the edge radius. Durations that were too short resulted in short circuits due to low material removal rate, and large values of 0.40 μs for the standard deviation of the slit width and 3.13 μs for the edge radius were obtained when the pulse duration was 40 ns. The two values were 0.28 μs and 2.34 μs when the pulse duration was 50 ns. With further increases of pulse duration, the values rose, as a long duration meant a high material removal rate and large amounts of bubbles produced in the machining gap. Fig 16 (d) shows the SEM photo of the slit machined using duration of 90ns, which has larger standard deviation value of the slit width than that of the slit in Fig. 16(a). 3.3 Effects of electrolyte concentration Fig. 19 describes the variation of the standard deviation of the slit width and the value of the edge radius with electrolyte concentration. The following machining parameters were used for the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse and 0.5 μm·s-1 feedrate. As shown in Fig. 19 and Fig. 20, the standard deviation of the slit width decreased with increasing concentration, but the edge radius increased with increasing concentration. A high concentration led to stable machining and wide machining gap, and then the homogeneity of the slit width improved. However, the electric field intensity outside the machining gap was strong and a large amount of bubbles was 15
produced in the machining gap with a high concentration, which resulted in a large value of the edge radius. Thus, an optimal electrolyte concentration of 0.05 mol·L-1 was chosen to fabricate the micro tools. 3.4 Effects of federate It is obvious from Fig. 21 and Fig. 22 that a high feedrate improves the homogeneity of the slit width and the edge radius. The following machining parameters were used in the slit experiments: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse and 0.05 mol·L-1 electrolyte concentration. A high feedrate reduces secondary corrosion, which is helpful for improving the machining accuracy and the homogeneity of the slit width and the edge radius. The standard deviation of the slit width decreased from 0.41 μm to 0.28 μm and the value of the edge radius decreased from 3.74 μm to 2.34 μm when the feedrate increased from 0.05 μm·s-1 to 0.5 μm·s-1. Thus, a high feedrate is preferred for micro tool fabrication.
4. Fabrication of micro square column tool arrays using a single wire cathode From the earlier analysis, the following optimal machining parameters were used in the micro tool fabrication: 200 μm amplitude and 400 μm·s-1 speed for cathode vibration, 5 μm amplitude and 100 Hz frequency for anode vibration, 5 V voltage, 6 μs period and 50 ns duration for the pulse, 0.05 mol·L-1 electrolyte concentration and 16
0.5 μm·s-1 feedrate. The machining process for micro tools was divided into two steps, as shown in Fig. 23. Micro beams were produced in step 1, as shown in Fig. 24(a) and (b), and then in step 2, micro square columnar tool arrays were fabricated, as shown in Fig. 24(c) and (d). Each micro tool was 140 μm long with a section size of 10 ×10 μm2. A micro tool was randomly selected to measure the homogeneity of the tool width and edge radius; 0.30μm and 2.54 μm was obtained for standard deviation of the tool width and the edge radius separately for the tool in Fig. 24(c).
5. Fabrication of micro tool arrays using multiple wire cathodes In order to improve production efficiency, multiple wire cathodes were used in micro tool array fabrication in WEMM, and the feed path changed in step 1 as shown in Fig. 25. As the machining process needed more energy, the electrical parameters were adjusted to 6 V for voltage, 5 μs for period and 90 ns for duration in step 1, and then micro beams were produced, as shown in Fig. 26(a). As the processing time was short, a single wire cathode was still used in step 2; a fabricated micro square columnar tool array is shown in Fig. 26(b). 6. EMM application of micro tools As shown in Fig. 27, the Micro dimple array were electrochemically fabricated using micro square columnar tool array as the cathode. As demonstrated in Fig. 27(b), measurements were taken on the dimples: the depths were 15.9 μm with entrance size of 25.5×24.6 μm. 7. Conclusions 17
On the basis of the results presented above, the following conclusions can be drawn: (i) Diffusion of electrolytic products such as bubbles is difficult in WEMM, and the influence of bubble behaviour on the electric field distribution in the machining area has been studied through simulations of the electric field. Bubbles in the machining gap always cause poor homogeneity of slit width and a large edge radius. (ii) The bubbles were driven out of the machining gap quickly, and accumulation of insoluble product reduced with vibration of electrodes. Large amplitudes and high speeds of cathode vibration, and large amplitudes and high frequencies of anode vibration, reduced the edge radius and improved the homogeneity of the slit width. (iii) For stable machining, pulses with low voltages, long periods and short durations and high feedrates improved the homogeneity of the slit width and reduced the edge radius. High concentrations of electrolyte improved the homogeneity of the slit width, and low concentrations of electrolyte reduced the edge radius. (iv) Using optimal machining parameters, micro square columnar tool arrays with a small edge radius and a good tool width homogeneity were fabricated. The micro dimple array was fabricated using EMM by employing the micro tool array as cathodes.
Acknowledgement 18
This work was conducted under the sponsorship of the National Natural Science Foundation of China (51375238), the Jiangsu Natural Science Foundation (BK20131361), the Funding of Jiangsu Innovation Program for Graduate Education (CXZZ13_0153) and a project funded by the China Postdoctoral Science Foundation (2015T80546).
References [1] Xue B, Yan Y, Li J, Yu B, Hu Z, Zhao X, Cai Q. Study on the micro-machining process with a micro three-sided pyramidal tip and the circular machining trajectory. J Mater Process Technol 2014; 217: 122–130. [2] Ghoshal B, Bhattacharyya B. Vibration assisted electrochemical micromachining of high aspect ratio micro features. Precis Eng –J Int Soc Precis Eng 2015; 42: 231–241. [3] Zeng YB, Yu Q, Wang SH, Zhu D. Enhancement of mass transport in micro wire electrochemical machining. CIRP Ann-Manuf Technol 2012; 61 (1): 195–198. [4] Rodríguez P, Labarga JE. Tool deflection model for micromilling processes. Int J of Adv Manuf Technol 2014; 76 (1–4): 199–207. [5] Koyano T, Kunieda M, Ultra-short pulse ECM using electrostatic induction feeding method. Procedia CIRP 2013; 6: 390–394. [6] Nguyen MD, Wong YS, Rahman M. Profile error compensation in high precision 3D micro-EDM 19
milling. Precis Eng –J Int Soc Precis Eng 2013; 37 (2): 399–407. [7] Chung DK, Hong SS, Min SP, Bo HK, Chong NC. Recent researches in micro electrical machining. Int J Precis Eng Man 2011; 12 (2): 371–380. [8] Kim BH, Na CW, Lee YS, Choi DK, Chu CN. Micro electrochemical machining of 3D micro structure using dilute sulfuric acid. CIRP Ann-Manuf Technol 2005; 54 (1): 191–194. [9] Zhu D, Wang K, Qu NS. Micro wire electrochemical cutting by using in situ fabricated wire electrode. CIRP Ann-Manuf Technol 2007; 56 (1): 241–244. [10] Sheu DY. High-speed micro electrode tool fabrication by a twin-wire EDM system. J Micromech Microeng 2008; 18 (10): 2067–2070. [11] Morgan C, Vallance R, Marsh E. Micro machining glass with polycrystalline diamond tools shaped by micro electro discharge machining. J Micromech Microeng 2004;14 (12): 1687-1692. [12] Fan ZW, Hourng LW. The analysis and investigation on the microelectrode fabrication by electrochemical machining. Int J Mach Tools Manuf 2009; 49 (49): 659–666. [13] Cabrera M, Dahmani R, Layouni Y, Semet V. Micro EDM milling with electrochemical fabrication of ultra-thin microtools and mapping of electrical microdischarges. Procedia CIRP 2016; 42: 650-655. [14] Ghoshal B, Bhattacharyya B. Influence of vibration on micro-tool fabrication by electrochemical machining, Int J Mach Tools Manuf 2013; 64 (1): 49–59. [15] Wang YF, Zeng YB, Qu NS, Zhu D. Note: electrochemical etching of cylindrical nanoprobes using a vibrating electrolyte. Rev Sci Instrum 2015; 86 (7): 106–109. [16] Green NG, Morgan H, Monaghan W, Robertson M. Fabrication of micro-electrode arrays for biotechnological applications. Microelectron Eng 1999; 46 (1–4): 397–400. 20
[17] Hu YY, Zhu D, Li HS, Zeng YB, Ming PM. Fabrication of high-aspect-ratio electrode array by combining UV-LIGA with micro electro-discharge machining. Microsyst Technol 2009; 15 (4): 519–525. [18] Kitamura N, Chim J, Miki N. Electrotactile display using microfabricated micro-needle array. J Micromech Microeng 2015; 25 (2): 25016–25023.
21
Fig. 1 Schematic diagram of the slit width and edge radius Fig. 2 Electric potential domain in the machining gap Fig. 3 Current density distribution around the wire electrodes Fig. 4 Process of electrode vibration Fig. 5 Schematic diagram of the machining process in WEMM Fig. 6 CCD images of the machining process Fig. 7 CCD images of wire cathodes after machining Fig. 8 Experimental set up for micro tool fabrication using WEMM Fig. 9 Cathode vibration amplitude versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 10 SEM photos of slits (anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 11 Cathode vibration speed versus standard deviation of the slit width and the value of edge radius (cathode vibration amplitude: 200 μm, anode vibration amplitude: 5 μm, anode vibration frequency: 100 Hz) Fig. 12 Anode vibration amplitude versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, cathode vibration amplitude: 200 μm, anode vibration frequency: 100 Hz) Fig. 13 SEM photos of slits (cathode vibration amplitude: 200 μm, cathode vibration speed: 400 μm·s-1)
22
Fig. 14 Anode vibration frequency versus standard deviation of the slit width and the value of edge radius (cathode vibration speed: 400 μm·s-1, cathode vibration amplitude: 200 μm, anode vibration amplitude: 5 μm) Fig. 15 Voltage versus standard deviation of the slit width and the value of edge radius (period: 6 μs, duration: 50 ns) Fig. 16 SEM photos of slits Fig. 17 Pulse period versus standard deviation of the slit width and the value of edge radius (voltage: 5 V, duration: 50 ns) Fig. 18 Pulse duration versus standard deviation of the slit width and the value of edge radius (voltage: 5 V, period: 6 μs) Fig. 19 Electrolyte concentration versus standard deviation of the slit width and the value of edge radius Fig. 20 SEM photos of slits Fig. 21 Feedrate versus standard deviation of the slit width and the value of edge radius Fig. 22 SEM photos of slits Fig. 23 Machining process for micro tools Fig. 24 Micro tools fabricated using a single wire cathode Fig. 25 Step 1 in machining process for micro tools using three wire cathodes Fig. 26 Micro tools fabricated using multiple wire cathodes in step 1 Fig. 27 The micro dimple array fabricated using the micro tool array
23
Figure
Width
Edge radius Anode
Fig. 1
Γ2
A2 y
Electrolyte x Wire
Γ3 Γ1 Γ4 (bubble 1#)
A1
Ω
Anode
(a) cross section Γ4 (bubble 2#)
Γ1 Electrolyte
B1
B2 Anode
Γ2 y Γ4 (bubble 4#) B4 x
Wire
Γ4 (bubble 3#) B3 Ω
(b) side section Fig. 2
Γ3
A/mm2
Y/μm
A2
A1 X/μm
Area-B
B1 Area-A
Area-C
Area-D
Area-A
Y/μm
(a) cross section
B2 Area-B
X/μm Area-C B4
B3 Area-D (b) side section
Fig. 3 .
Wire cathode
Cathode displacement One period
Cathode velocity One period
Amplitude Workpiece anode
Anode displacement One period
Anode velocity One period
Amplitude
Fig. 4
Bubbles
Anode
Wire cathode Fig. 5
Insoluble product
Wire cathode Wire cathode Insoluble product
Bubbles
Anode Anode Bubbles
(a) Without vibration
Bubbles
(b) Only anode vibration
Wire cathode
Wire cathode
Anode
Bubbles
Insoluble product
Anode
Insoluble product
(c) Only cathode vibration Fig. 6
(d) With electrodes vibration
Wire cathode
Wire cathode Wire cathode
Machining area
Insoluble product
(a) only anode vibration Fig. 7
Machining area
Machining area
(b) only cathode vibration
(c) with electrodes vibration
Pulse generator Z
Industrial PC
CCD camera
Y X Electrolyte
PZT controller PZT
Vibration isolation platform Fig. 8
Cathode vibration amplitude / μm Fig. 9
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Without cathode vibration
(b) Cathode vibration speed: 400 μm·s-1, amplitude:200 μm
(c) Cathode vibration speed: 400 μm·s-1, (d) Cathode vibration speed: 600 μm·s-1, amplitude:800 μm μm amplitude:200 Fig. 10 Hz)
Cathode vibration speed / μm Fig. 11
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Anode vibration amplitude / μm Fig. 12
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Without cathode vibration
(c) Anode vibration amplitude: 9 μm, frequency: 100 Hz Fig. 13
(b) Anode vibration amplitude: 5 μm, frequency: 100 Hz
(d) Anode vibration amplitude: 5 μm, frequency: 200 Hz
Anode vibration frequency / μm Fig. 14
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Voltage / V Fig. 15
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Voltage: 5v, period: 6 μs, duration: 50 ns
(b) Voltage: 10v, period: 6 μs, duration: 50 ns
(c) Voltage: 5v, period: 2 μs, duration: 50 ns Fig. 16
(d) Voltage: 5v, period: 6 μs, duration: 90 ns
Period /μs Fig. 17
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Duration / ns Fig. 18
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
Concentration / mol·L-1 Fig. 19
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Concentration: 0.02 mol·L-1
(b) Concentration: 0.05 mol·L-1 Fig. 20
(c) Concentration: 0.07 mol·L-1
Feedrate / μm·s-1 Fig. 21
/ μm
Fillet radius / μm
Standard deviation / μm
Standard deviation Fillet radius
(a) Feedrate: 0.05 μm·s-1
(b) Feedrate: 0.3 μm·s-1 Fig. 22
(c) Feedrate: 0.5 μm·s-1
Anode
Wire cathode
Feed path Step 1
Step 2 Fig. 23
Micro tools
a
b
c
d
Fig. 24
Wire cathode
Anode
Feed path Fig. 25
a
b
Fig. 26
a
b
Fig. 27
Table 1 Parameters applied in electric field simulation Parameter
Value
Wire cathode diameter/μm
10
Slit width/μm
12
Anode thickness/μm
40
Bubble diameter/μm
0.6/0.6/0.6/2
Electrolyte electrical conductivity/S·m−1
1.83
Applied voltage/V
5
24