Open Access
Add to BibSonomy
Abstract
Silicon carbide (SiC) ceramics have been widely used for microelectronics, aerospace, and other industrial fields due to their excellent chemical stability and thermal tolerance. However, hard machinability and low machining precision of SiC ceramics are the key limitations for their further applications. To address this issue, a novel method of underwater femtosecond laser machining was introduced in this study to obtain high precision and smooth surface of the microgrooves of SiC ceramics. The removal profiles were characterized in terms of width, depth, and surface morphology, which exhibited high dependence on the femtosecond laser processing parameters. The instability during the underwater processing affected by laser-induced gas bubbles and material deposition, however, limits the high surface accuracy of microgrooves and processing efficiency. The process condition transformation from a bubble-disturbed circumstance to a disturbance-free model was carefully investigated through a high speed camera for the femtosecond laser processing of SiC ceramics in water. The experiment results indicated that degree of disturbed effect was heavily dependent on size, distribution, and motion of laser-induced gas bubble. Furthermore, some typical evolution mechanisms of gas bubble and their influence on the removal profiles of microgrooves were discussed in detail. Bubble evolution has been proven to be mainly responsible for the behavior of laser propagation (focus model, total reflection, etc.), which notably affects microstructural characteristic of the microgrooves.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Silicon carbide (SiC) ceramics have many promising applications in petroleum, microelectronics, automotive, aerospace, and other industrial fields because of their excellent mechanical strength at high temperature, oxidation and corrosion resistance, low thermal expansion coefficient, and high thermal shock resistance [1–3]. However, owing to its brittleness and high hardness, it is difficult to machine SiC bulk material into desirable micro-components by conventional machining techniques, such as diamond tool [4] and electrical discharge machining [5]. These processes are associated with many unacceptable drawbacks such as serious tool wear, lower material removal rates, insufficient dimensional accuracy, and thermal damage of the work piece [6]. Therefore, it is necessary to develop a novel approach to achieve defect-free and high-precision processing of microstructures.
Recently, laser machining has been considered to be an innovative and potential technology for bulk material removal and fabrication of ceramics-based complex components [2,3,7]. However, some studies demonstrated that ceramics fabricated with nanosecond laser have many disadvantages including the existence of heat-affected zones (HAZs) [8], debris redeposition [9–11], and thermal stress-induced microcracks [12,13]. Sciti et al. [14] found the presence of HAZs and microcracks extended around the holes via nanosecond laser drilling of SiC in ambient air. Liu et al. [15] adopted nanosecond laser to fabricate the micro-scale textured grooves, and the particle redeposition and microcracks were observed nearby the edge of grooves. Ultra-short pulsed (picosecond ~10−12 s or femtosecond ~10−15 s) laser, as a promising alternative for machining ceramics, can effectively eliminate the thermal stress-induced microcracks and minimize the size of HAZs, as the thermal diffusion is highly restricted within the extreme short laser-material interaction time. Kurita et al. [16] found that the femtosecond laser could eliminate microcracks and reduce the HAZs, compared to nanosecond laser. Crawford et al. [17] obtained the microcracks-free microgrooves of Si through optimizing the femtosecond laser cutting parameters. However, the formed debris and recast still remained on the surface of microgrooves, which are the main obstacles for the femtosecond laser machining ceramics referring to micro-level depth. This is mainly attributed to the occurrence of the surface inductive effect in laser processing, and significant influence of the redeposition on the surface topography [18].
The underwater laser machining, as an advanced manufacturing method, can avoid or reduce the HAZs, microcracks, and debris redeposition [19]. Water can more effectively cool the workpiece and ejected material, because the heat convection coefficient of water is 2 – 3 orders higher than that of air [19]. The effect of water during laser material processing can convert the light energy into a mechanical impulse, producing higher plasma pressure and longer duration of shock waves due to confinement [20]. Subsequently, Water convection, explosive evaporation, and motion of laser-induced bubble are generated, which contribute to the removal of redeposited debris or particles from surfaces. Moreover, water may react chemically with target material or its generated particles under the effect of femtosecond laser irradiation in some cases. Therefore, a water film can significantly improve the smooth surface and dimensional precision of the machined microstructures. Dolgaev et al. [21] obtained a high precision machined cavity in SiC ceramics through underwater copper-vapor-laser etching compared to that in air. Chen et al. [22] reported a method referring to the underwater nanosecond laser processing of silicon nitride ceramics, and they found that the underwater machining reduced the oxidation reaction, providing debris-free and recast-free surface. However, there are several drawbacks in the processing of underwater nanosecond laser. For example, serious ionization of water and repeated high pressure by plasma confinement and thermal stress can lead to the accumulation of fatigue on the target surface, inducing the generation of surface microcracks [23]. It is difficult to control generation of cracks during processing. To solve this issue, some authors adopted femtosecond laser in the underwater machining [24,25]. This technique can promisingly ensure that the ablated material is suspended in water [24] and the machined surfaces are smooth without debris redeposition of the ablated material on the target surface [26]. Muhammad and Li [25] found that high machining quality with well-defined edges and debris-free and recast-free characteristics was achieved by the water-assisted femtosecond laser micromachining. However, they also observed that the shielding effect was induced by filamentation and optical breakdown of water, which reduced the amount of energy reaching the target. Furthermore, the laser-induced gas bubble could scatter the laser beam, accordingly affecting the precision of the processed surface [27].
However, the effect of evolution of laser-induced gas bubble on the microstructures of underwater laser machining has rarely been investigated. To fill this gap, in this study, with assistance of high-speed camera, underwater femtosecond laser microgrooving (UFLG) of SiC ceramics with various processing parameters was investigated. Dynamic behavior of the gas bubbles and their effect on structural characteristic of microgrooves were discussed. Moreover, the potential mechanism of microstructural and morphological evolution involved in UFLG process were elucidated in details.
2. Experimental setup and methodology
The experiments were carried out with the Yb: KGW femtosecond laser system Pharos (Light Conversion Ltd., Vilnius, Lithuania). The scanning path of laser processing was controlled using a two-axis galvanometric scanner (PS1XY2 - S4, CTI Inc.), and the beam was focused with an F-theta lens (fF = 170 mm). The SiC ceramics (30 mm × 30 mm × 4 mm) were mounted at the bottom of the water container. Typical thickness of water layer above the workpiece surface was about 1 mm at room temperature (298 K for pure water). Furthermore, the container was mounted on a computer-controlled X–Y–Z stage, as shown in Fig. 1. The laser machining parameters are listed in Table 1. Furthermore, all samples were ultrasonically cleaned for 30 minutes before and after laser machining.
Fig. 1 (a) Schematic illustration of the experimental setup. (b) Schematic illustration for measuring the removal cavity characteristics: Line 1 and Line 2 represent the height data of the upper surface and the bottom of the measured microgroove, respectively. (c) Relationship between microgroove depth and focal position of UFLG samples with laser frequency of 10 kHz, pulse energy of 25 μJ, scanning speed of 0.5 mm / s, and scanning number of 5, where negative sign represents the focus position below the sample surface relative to the air environment.
Table 1. Femtosecond laser parameters for micro-grooving processing
Furthermore, a high-speed camera (Optronis, CP80 - 3 - M - 540) with a filter was used to record the dynamic behavior of gas bubbles during the UFLG processing. The surface morphologies of microgrooves were observed by scanning electron microscopy (SEM, HITACHI SU8010), and changes in the surface elements of processed materials were identified by energy dispersive spectroscopy (EDS). Prior to this, a thin gold layer was sputtered onto the ceramic surface to avoid the charging effect during the detection of microstructures. The three-dimensional (3D) profiles of the microgrooves were characterized by laser scanning confocal microscopy (Olympus, OLS4000).
In this study, the removal volume per pulse is described as Vpulse = Vrem / Npulse. The pulse number (Npulse) with certain removal volume was obtained as Npulse = L f Nscan / vscan, where Nscan is the scanning number per microgroove. The length L and the removal volume per length Vrem were measured using a laser scanning confocal microscope. The measurement method for the depth and width of the machined microgroove is shown in Fig. 1(b). The microgroove depth (H) is the average height of the upper surface Line 1 (Hupp) minus the average height of the bottom Line 2 (Hbot). Moreover, owing to the existence of bubbles resulting in scattering, reflection, and refraction of laser beam; calculation of the spot size based on the Gaussian beam was not accurate [28,29]. Hence, it is difficult to determine the focal position in water via the machined microgroove width compared with in air [30]. In this regard, to ensure the maximum ablation efficiency occurring near the interface of samples during underwater processing, the focal position was fixed at 1 mm under the surface of workpiece, as shown in Fig. 1(c).
3. Results and discussion
3.1. Comparison of the femtosecond laser-ablated microgrooves formed in air and in water
Figure 2 shows the surface morphologies of the SiC ceramic microgrooves machined with frequency of 100 kHz, pulse energy of 25 μJ, scanning speed of 2 mm / s, and scanning number of 1 in air and in water. The result indicated that the undesired phenomenon appeared during processing in air. The splash and debris residue are observed on top of the kerf. Besides, a 6-μm thick recast layer surrounded by redeposition of vaporized material was formed on the sidewall of microgroove [Fig. 2(a)]. The composition of the redeposition mainly included C, O, and Si as indicated by the EDS analysis. This is attributed to the occurrence of the thermal chemical reaction, as presented in Eqs. (1)–(3) [31,32]. As a result, the reaction product SiO2 particles covered the surface of sample and formed an oxide layer, developing an uneven and rough surface.
Fig. 2 Comparison of the machined SiC ceramic microgrooves with laser frequency of 100 kHz, pulse energy of 25 μJ, scanning speed of 2 mm / s, and scanning number of 1 corresponding to the cases of (a) air environment and (c) water environment. 3D profile of (b) and (d) corresponding to the cases of (a) air environment and (c) water environment, respectively.
A smooth surface of the microgrooves was obtained in water as depicted in Fig. 2(c), and almost no deposited debris and recast layer occurred. Compared to that in air, the interaction between laser and water is very complex [33,34]. This is attributed to the fact that the optical breakdown occurs in water, and vaporization, expansion, plasma shock wave, and gas bubbles are produced, which effectively carry away the deposited debris [19]. Owing to the existence of high temperature and high pressure of plasma shock wave, SiC oxidation is likely to be accelerated to form SiO2 as presented in Eqs. (4) and (5) [35,36]. When the temperature in the processing zone exceeds 1000 °C, SiO2 further reacts with H2O to form gaseous silicic acid (Si(OH)4) as shown in Eq. (6) [35].
Therefore, the composition of gas bubbles induced by femtosecond laser may contain water vapor, CO, CO2, H2, and Si(OH)4. The shrinking and collapse of gas bubbles may produce significant effect on the microstructural evolution. A high-pressure region near the bubble wall is produced at the moment of collapse due to the water confinement effect [37]. The impact pressure acting towards the sample is about 10 times greater than that in the ambient air, which ejects the molten material away from the cavity and prevents recast formation [38,39]. Compared to laser microgrooving in air, the amount of oxygen decreased when it was performed in water, as presented in Figs. 2(a)–2(c). The result indicated that water carried the discharged particles efficiently and controlled the deposition of the debris during processing.
Moreover, noteworthy, the microgroove depth was only about 6 μm with scanning number of 1 [Fig. 2(d)], which is far below that of the ultra-short pulsed laser ablation in air [Fig. 2(b)]. Therefore, the significant machining resolution of UFLG condition is more helpful for obtaining microgrooves with different shapes and high precision. For this purpose, the effects of femtosecond laser processing parameters on the surface morphologies and 3D profiles are discussed in the following sections of this paper.
3.2. Influences of frequency on the microgrooves
The depth and width of microgroove change with different frequencies as shown in Figs. 3(a) and 3(b). Figure 3(a) demonstrates that the microgroove depth increases with the increase of the frequency from 10 to 50 kHz because of increasing pulse number per area, resulting in higher removal depth. When the frequency increases to 100 kHz, the microgroove depth with pulse energy of 25 μJ tends to be saturated, while the depth with pulse energy of 55 μJ decreases. Figure 3(b) illustrates that the microgroove width with pulse energy of 25 μJ increases with increasing frequency. However, the microgroove width with pulse energy of 55 μJ increases with increasing frequency from 10 to 25 kHz, and then decreases when the frequency continuously increases to 100 kHz. The experimental results indicated that the higher pulse energy was more sensitive to the increase of frequency than the lower pulse energy in term of the influence on the microgroove depth and width. Therefore, the relationship between the ablation rate (defined as the removal volume per pulse) and laser frequencies at pulse energy of 55 μJ was investigated herein, as presented in Fig. 3(c). Clearly, the removal volume per pulse almost remains constant with the increase in frequency from 10 to 50 kHz, and gradually drops when laser frequency continues to increase. It is suspected that the presence of gas bubbles may lead to this phenomenon. Previous pulse produces the bubble, and the subsequent laser pulse leads to the growth and deformation of the bubble and small bubbles are sometimes generated on the top of the bubble [30]. In addition, the bubbles also lead to the scattering of the subsequent laser pulse, thus reducing the laser ablation efficiency [40].
Fig. 3 (a) Depth and (b) width of UFLG samples with different frequencies. (c) Relationship between the removal volume per pulse and laser frequencies at pulse energy of 55 μJ. Surface morphologies and 3D profiles of UFLG samples with different frequencies at pulse energy of 55 μJ. (d) 10 kHz. (e) 25 kHz. (f) 50 kHz. (g) 100 kHz. The input scanning speed and scanning number are set as 1 mm / s and 5, respectively.
Figures 3(d)–3(g) exhibit the surface morphologies along with 3D profiles, revealing the relationship of the laser frequency with the microgroove. Obviously, high frequency brings in the low ablation rate, but the discontinuous and non-uniform ablated bottom, accompanying with holes. Moreover, the distortion of the bottom edge of the microgroove also emerges. Higher number of pulses per area led to the increase in the interaction opportunity between laser pulses and gas bubbles, thus affecting the propagation and energy distribution of laser pulses. Furthermore, interestingly, the microgroove shape changes from the inverted trapezoidal microgroove to the V-shaped microgroove with the increase in the frequency. Figure 3(d) exhibits the formation of the inverted trapezoidal microgroove with defect-free and smooth surface at frequency of 10 kHz. With the increase in the frequency to 25 kHz, the inverted trapezoidal microgroove depth increases as shown in Fig. 3(e). When the frequency continuously increases to 50 kHz, a microgroove with non-symmetrical kerf that is tilted to the left side is formed as presented in Fig. 3(f). The effect of shock wave enhanced with increasing frequency, and the non-symmetrical distribution of water flow may also increases along moving direction of laser beam in the processing area, inducing a change in the direction of laser beam propagation. In addition, owing to the increased number of pulses per area, the laser-induced gas bubbles inside the microgroove become more unstable. And Scattering at interface of water-bubble deflects to the left under the effect of water flow. Moreover, the microgroove was continuously processed five times. The previous beam deflection further significantly affected the energy distribution of subsequent laser pulses.
In order to explore the case involving undesired microgroove formation mentioned above in laser processing in water, a high speed camera shooting was performed during UFLG process. The laser-water-workpiece interaction zone was recorded using the high-speed camera, as shown in Fig. 4. Figure 4(a) reveals that the interaction zone is labeled by a lower gas bubble activities at frequency of 10 kHz, where the gas bubbles are small and rapidly dissipate, which does not affect the machining area for the subsequent pulses. During processing, the camera reveals that the gas bubbles move forward and backward under the action of the laser-induced shock wave with respect to the scanning direction (see Visualization 1). Owing to the different density of water and gas bubbles, the moving direction of gas bubbles is opposite to that of the water flow. The debris may be carried away, leaving a smooth surface [see Figs. 3(d)]. When the frequency is 50 kHz, the generation rate increases and size of gas bubble enlarges. Some of them adhere around the material surface [Fig. 4(b)] owing to the surface tension of water and the others float in water. In addition, the vortex also occurs round the laser beam in the processing area, where the gas bubbles rotate with the flow of water [see Visualization 2]. When the gas bubbles randomly move under the vortex effect and happen to collide with the laser beam in water, the interesting phenomenon of gas bubble explosion takes place and then the gas bubble clusters abruptly emerge. As a possible explanation, the laser beam produced an effect on second focus inside the bubble due to the different refractive indices of water and bubbles, causing the bubble to absorb energy sharply which leads to its blasting. This may be the main reason for the change in the shape of the microgroove from the inverted trapezoid to V shape. Furthermore, this may also bring about the uncontrollable reflection, refraction, and scattering of laser beam, unpredictably changing the position and intensity of beam aiming at the workpiece surface [27]. Moreover, the gas bubbles would change the displacement of focus point for subsequent laser pulses [29,40], which can lead to the poorly finished surface [see Fig. 3(g)]. Therefore, to avoid the effect of high frequency on gas bubble, the frequency of femtosecond laser should be less than 50 kHz during the underwater machining.
Fig. 4 The dynamic behavior of the laser-water-workpiece interaction zone from a disturbance-free to a disturbed removal under different frequencies: (a) 10 kHz (see Visualization 1). (b) 50 kHz (see Visualization 2). Fs-laser and the white arrow indicate the moving direction of femtosecond laser. The input pulse energy and scanning speed are set as 55 μJ and 1 mm / s, respectively.
3.3. Influences of pulse energy on the microgrooves
Apart from laser frequency, previous studies have shown that pulse energy also significantly influences the microgroove depth and width. Therefore, the machined depth and width of UFLG samples with different pulse energy were studied as shown in Figs. 5(a) and 5(b). Obviously, the curve of the machined microgroove depth shows a slow increasing trend with increasing pulse energy below 55 μJ. Further increasing pulse energy, the microgroove depth increases rapidly, but the fluctuation of microgroove depth becomes larger. The microgroove width increase with pulse energy of below 55 μJ, and then decreases with pulse energy from 55 to 75 μJ due to the nonlinear effect of underwater laser processing. With the continuous increase in the pulse energy to 85 μJ, the microgroove width increases dramatically. The removal volume per pulse increases as pulse energy increases up to 45 μJ, and then shows a descending with increasing pulse energy to 55 μJ. The self-focusing effect leads to the maximum ablation position to shift up and the shielding effect of bubble is induced [29], affecting laser beam propagation to reduce the ablated efficiency. Further increasing pulse energy, removal volume per pulse increases again. This is because the self-focusing effect significantly enhances and the as-generated higher shock wave reduces the shielding effect of bubble, increasing the microgroove depth.
Fig. 5 (a) Depth and (b) width of UFLG samples with different pulse energy. (c) Relationship between the removal volume per pulse and pulse energy. Surface morphologies and 3D profiles of UFLG samples with different pulse energy: (d) 15 μJ. (e) 35 μJ. (f) 45 μJ. (g) 75 μJ. The input frequency, scanning speed, and scanning number are set as 25 kHz, 1 mm / s, and 5, respectively.
The surface morphologies and 3D profiles of UFLG samples with different pulse energy are presented in Figs. 5(d)–5(g). Noteworthy, the inverted trapezoidal microgrooves are obtained with pulse energy of 15 and 35 μJ, where the edge surfaces are smooth without debris redeposition and recast layer and the bottoms are relatively flat due to the periodic pulsation of micro gas bubbles. Although the inverted trapezoidal microgroove with pulse energy of 45 μJ, the edge surfaces is obtained, some holes emerge on the sidewall of microgroove shown in Fig. 5(f). The width of microgroove bottom becomes wider compared with 35μJ. However, when the pulse energy is 75 μJ, the microgroove shape changes to the V shape as shown in Fig. 5(g). The sawtooth-like stripes appear on the sidewall of microgroove, and some holes distinctly appear at the bottom. Unsatisfying edge retention of the microgroove bottom was more curved and rough with increasing pulse energy to 75 μJ.
Based on the above mentioned analysis, the pulse energy may determine the formation and growth of gas bubbles. Therefore, the relationship between pulse energy and gas bubble was also studied herein using the high speed camera. The size of gas bubbles is small at the pulse energy of below 35 μJ, as presented in Figs. 6(a) and 6(b). The gas bubbles float and move away from the processing area along the microgroove direction at 35 μJ [see Visualization 3], due to the effect of shock wave. The transverse movement of gas bubbles, during the shrinking and collapse process near the target, flattens the bottom of microgrooves as shown in Figs. 5(d) and 5(e). With increasing pulse energy to 45 μJ, laser beam is extremely unstable during underwater movement. Therefore, the nonlinear optical effects including self-focusing and plasma defocusing easily occurred [40], leading to an increase of microgroove depth slowly. Once the self-focusing effect occurs during processing, the local laser energy density increases dramatically, resulting in the water absorbing a large amount of laser energy to cause gas bubble explosion [see Visualization 4]. Consequently, the size and quantity of gas bubble instantaneously enlarged and increased, respectively, adhering to the workpiece surface as presented in Fig. 6(c). Therefore, the removal regime was disturbed and then the propagation of laser beam was affected, which led to the formation of holes on the microgroove sidewall [Fig. 5(f)]. In addition, the bubble cavitation occurring at the microgroove bottom widened the width of microgroove bottom, which was beneficial to increase ablation efficiency [Fig. 5(c)]. With the further increase in the pulse energy to 75 μJ, the local energy density is more concentrated due to stronger self-focusing effect, decreasing the laser beam diameter. Therefore, the deeper and narrower microgroove was obtained. In addition, the gas bubbles continuously enlarged and the quantity of gas bubble increased exponentially, as presented in Fig. 6(d). The gas bubbles exploded dramatically and the collision between the gas bubbles and laser beam became violent, which markedly enhanced the influence of reflection, refraction, and scattering [see Visualization 5] compared to that at 45 μJ. This unfavorable phenomenon could lead to the appearance of holes in the bottom followed by the distortion of the bottom edges owing to the presence of gas bubble cavitation as shown in Fig. 5(g), seriously affecting the dimensional accuracy of microgrooves. With the continuous increase in the pulse energy, the enlarging radius of the bubble becomes comparable with the thickness of the liquid layer. The cavitation effect of bubbles further gets enhanced, and then these bubbles may break the liquid surface [29]. In this situation, the liquid pins to the target surface, so that the subsequent pulses ablate the target in air. This is the reason for the extreme increase in the microgroove depth and width.
Fig. 6 The dynamic behavior of the laser-water-workpiece interaction zone from a disturbance-free to a disturbed removal under different pulse energy: (a) 15 μJ. (b) 35 μJ (see Visualization 3). (c) 45 μJ (see Visualization 4). (d) 75 μJ (see Visualization 5). Fs-laser and the white arrow indicate the moving direction of femtosecond laser. The input frequency and scanning speed are set as 25 kHz and 1 mm / s, respectively.
3.4. Modulation of depth and removal profile with scanning speed and scanning number
Figure 7(a) shows the snapshots of the machined microgrooves with different scanning speed and scanning number obtained using scanning confocal microscope. It is distinctly found that increasing the scanning speed results in the discontinuity and non-uniformity of the machined microgrooves at scanning number of 1 and 3. To avoid this unfavorable phenomenon, increasing scanning number to five provides the continuous microgroove geometries. However, the non-uniform ablation during microgrooving is not completely eliminated with increasing scanning number. It is indicated that scanning speed during underwater femtosecond laser processing has a significant influence on the geometry and surface quality of microgroove. Dynamic behaviors of the laser-water-workpiece interaction zone with respect to scanning speed are depicted in Figs. 7(b)–7(e), revealing that the processing state changes from the disturbance-free to disturbance condition with increasing scanning speed from 0.5 to 2 mm / s. With further increase in the scanning speed, the agglomeration of small gas bubbles results into bigger gas bubble due to the surface tension of water, thus adhering to the workpiece surface when the size of the bubble enlarges to a certain extent as shown in Figs. 7(d) and 7(e). On the other hand, the high scanning speed causes waves at the air-water interface, which affects the stability of laser beam propagation. The above mentioned two factors may be the main reasons for the inhomogeneous distribution of laser energy, leading to the continuous and uniform geometries of the microgroove.
Fig. 7 (a) Snapshots of the machined microgroove by scanning confocal microscope at different scanning speed and scanning number (Nscan). The dynamic behavior of the laser-water-workpiece interaction zone under different scanning speed: (b) 0.5 mm / s. (c) 2 mm / s. (d) 5 mm / s. (e) 7 mm / s (see Visualization 6). Fs-laser and the white arrow of Figs. (b-e) indicate the moving direction of femtosecond laser. The input frequency and pulse energy are set as 25 kHz and 35 μJ, respectively.
Therefore, Fig. 8 presents the depth, width, and removal volume per pulse of UFLG samples through modulating the scanning speeds and scanning numbers. Figure 8(a) demonstrates that the microgroove depth increases with increasing scanning number, and decreases with increasing scanning speed. The growth rate of depth at low scanning speeds is higher than that at high scanning speeds with the increase in the scanning number. Moreover, Fig. 8(b) exhibits that increase in scanning number and scanning speed can enhance the width of the as-formed microgroove. When the scanning number increases to 10, the microgroove width tends to be saturated. As scanning number with different scanning speed increases up to a certain point, the microgroove width will tends to be saturated. Therefore, when the microgroove width is constant, the microgroove depth can be controlled by modulating scanning speed. Therefore, when the microgroove width is constant, the microgroove depth can be controlled by modulating the scanning speed. In particular, starting from scanning number of 15, the microgroove width becomes larger at the scanning speed of 2 mm / s compared to that at 0.5 and 1 mm / s. During longer processing time, the gas bubbles may adhere on the wall side of kerf [see Fig. 11(d)] [27], or the moving gas bubbles hit the laser beam, which deflects the direction of laser propagation. This is a drawback when the underwater laser processing is performed in still water [41].
Fig. 8 (a) Depth and (b) width of UFLG samples with different scanning speed and scanning number. (c) Relationship between the removal volume per pulse and scanning number. The input frequency and pulse energy are set as 25 kHz and 35 μJ, respectively.
Figure 8(c) shows the relationship between the removal volume per pulse and scanning numbers and scanning speeds. At the scanning speed of 0.5 mm / s, the removal volume per pulse decreases with increasing scanning number. However, when the scanning speed is 1 and 2 mm / s, the removal volume per pulse gradually drops with increasing scanning number from 1 to 5, and remains constant with the continuous increase in the scanning number. With the further increase in the scanning speed to 5 mm / s, the removal volume per pulse hardly changes with scanning numbers. Considering the accumulated effect and the amount of deposited energy per area by fixing laser frequency and pulse energy, the ablation rate of two cases were studied, where case 1 is the microgroove with scanning speed of 0.5 mm / s and scanning number of 1; and case 2 is the microgroove with scanning speed of 5 mm / s and scanning number of 10. The removal volume per pulse for case 1 is 30.40 ± 0.93 μm3 per pulse; and the removal volume per pulse for case 2 is 2.67 ± 0.15 μm3 per pulse. The experimental results indicate that the low scanning speed has a high ablation rate than the high scanning speed. The presence of laser-induced gas bubbles during underwater processing reduces the ablation rate [24,42]. The super-continuum and plasma generation may attenuate the laser beam. The laser beam quality and energy density tremendously reduces, which decrease the machining efficiency.
According to the above mentioned results, the effect of scanning speed on surface morphologies and 3D profiles of UFLG samples with scanning number of 15 was first studied, as shown in Fig. 9. Clearly, with the increase in the scanning speed, the microgroove depth becomes shallower and the bottom profile becomes wider and uneven, seriously reducing surface dimensional accuracy. The V-shaped and U-shaped microgrooves were obtained at the scanning speed of 0.5 and 2 mm / s, respectively. Further increase in the scanning speed to 7 mm / s led to the emergence of the inverted trapezoidal microgrooves. The result indicated that when the scanning speed increased from 0.5 to 7 mm / s, the number of pulses per area decreased 14 times theoretically. However, the depths of the microgrooves changed from 149 ± 4.32 μm [Fig. 9(a)] to 12.52 ± 1.41 μm [Fig. 9(d)], only reducing by about 12 times. In fact, the removal geometry significantly influenced the propagation and the energy density distribution of the laser pulse [43]. The inclination angle of removal geometry increased with the increase of ablation depth during processing. Therefore, the multi-reflection on the microgroove wall and the presence of gas bubbles led to the shifting of the position of the maximum laser intensity until the ablation depth reached the saturation [43]. Moreover, the increase of scanning speed weakened the effect of shock wave in the ablation area. Consequently, the gas bubbles could not be timely washed away, resulting in more possibility for the agglomeration of small gas bubbles. The size of gas bubble enlarged obviously at the scanning speed of 7 mm / s [see Visualization 6]. The vortex appeared on both sides of laser processing area. The larger gas bubble and water flow may be another factor affecting the propagation of laser beam.
Fig. 9 Surface morphologies and 3D profiles of UFLG samples with different scanning speed: (a) 0.5 mm / s. (b) 2 mm / s. (c) 5 mm / s. (d) 7 mm / s. The input frequency, pulse energy, and scanning number were set as 25 kHz, 35 μJ and 15, respectively.
Second, the influence of different scanning numbers on the surface morphologies and 3D profiles of UFLG samples was studied as presented in Fig. 10. It is obviously found that the surface quality of the bottom and wall side of the microgrooves tends to be uniform and smooth with increasing scanning number. The microgroove shape is approximately the shallow rectangle [Fig. 10(a)] at scanning number of 1 and changes to the inverted trapezoidal [Fig. 10(b)] at scanning number of 5. With the further increase in the scanning number from 10 to 20, the V-shaped microgrooves [Figs. 10(c) and 10(d)] emerge. The experimental results indicated that the surface dimensional accuracy and removal profiles of SiC ceramic microgrooves strongly depend on the scanning number. The gas bubbles are effectively eliminated or pushed away from the interaction zone with increasing scanning number. Furthermore, the accumulation effect of multi-pulse during the interaction between laser, SiC ceramic, and water leads to the increase in the microgroove depth.
Fig. 10 Surface morphologies and 3D profiles of UFLG samples with different scanning number: (a) Nscan = 1. (b) Nscan = 5. (c) Nscan = 10. (d) Nscan = 20. The input frequency, pulse energy and scanning speed are set as 25 kHz, 35 μJ and 0.5 mm / s, respectively.
3.5. Discussions on the evolution mechanisms of gas bubble on the removal profiles of microgrooves during the underwater femtosecond laser machining of SiC ceramics
The femtosecond laser irradiation in the underwater processing is used for the localized micro-machining of SiC ceramics surface. Certain suitable underwater laser processing parameters impact on ceramics material removal and then a thermochemistry machining can be activated. Therefore, there are two reasons for the bubble generation. One is the generation of bubble of CO, CO2, H2, and Si(OH)4 by the thermochemical reaction, and the other is the water vapor formed by the laser heating water, similar to the bubble generated by boiling water, which is the main factor. Furthermore, laser-induced plasma shock wave and the gas bubbles escaping from the processing area cause the flow of water. Therefore, the evolution of gas bubble significantly influences the laser beam propagation and the removal profiles of microgrooves.
According to the previous experiments, it is concluded that different removal profiles of microgrooves are achieved by adjusting laser processing parameters. Therefore, schematic illustrations of gas bubble evolution were provided to explain their occurrence of some typical removal profile of microgroove as shown in Fig. 11. There are two states of the disturbance-free removal [Fig. 11(a)] and the disturbed removal [Figs. 11(b)–11(d)]. For the disturbance-free regime [Fig. 11(a)], the inverted trapezoidal, U-shaped, and V-shaped microgrooves with smooth surface and high precision are obtained through enhancing the scanning numbers. Under these conditions, i.e., laser frequency < 50 kHz, pulse energy < 45 uJ, scanning speed < 2 mm / s, the gas bubbles slightly affect the propagation of incident laser beam owing to their sizes being in submicron range as shown in Figs. 4(a), 4(b), 6(a), 6(b), 7(b), and 7(c). Under this condition, for instance, the inverted trapezoidal microgroove with high precision is obtained at the scanning number of 5 as shown in Figs. 3(d), 3(e) and 5(d)–5(f). This is attributed to the fact that the periodic pulsation of micro gas bubbles flattens the bottom of microgrooves during the shrinking and collapse process. Furthermore, the low scanning speed ensures a nearly continuous contact between laser and workpiece surface. With the increase in the scanning number above 10, V-shaped microgrooves are obtained [see Figs. 9(a), 10(c), and 10(d)]. With increasing microgroove depth, the position of the maximum laser intensity shifts up because of the multi-reflection on the microgroove wall and gas bubble defocusing, leading to the deeper V-shaped microgrooves.
Fig. 11 Characteristic removal profiles of the possible resulting underwater femtosecond laser machining SiC ceramics as well as schematic illustrations of the supposed dominating mechanisms showing (a) a disturbance-free removal and (b-d) a disturbed removal
When the size of gas bubble grew to be larger or approximately equal to the diameter of laser beam as illustrated in Figs. 11(b)–11(d), change in the propagation path occurred, which significantly affected the absorption of laser energy by the material [27]. The formation and evolution of these bubbles could be related to the high frequency, high pulse energy, or high scanning speed as shown in Figs. 4(b), 6(c), 6(d), 7(d), and 7(e), respectively.
The gas bubble size increases dramatically at the high frequency or high pulse energy. When the gas bubbles happen to collide with the laser beam in water, the bubble becomes equivalent to a lens. Therefore, the re-focusing of laser beam may be generated as shown in Fig. 11(b), which leads to change in the microgroove shape from the inverted trapezoid to V shape, as shown in Figs. 3(f), 3(g), and 5(g). Then the explosion of gas bubble occurs because the gas bubble absorbs a large amount of energy instantaneously, causing the generation of clusters of many small-size bubble around the laser beam. The small-size bubble clusters are more likely to cause total reflection of the laser and deflect the laser beam in random direction as shown in Fig. 11(c), accordingly causing certain holes at the surface of microgrooves and distortion of microgroove contour. Gas bubbles adhere to the SiC ceramics surface or wall side of microgroove as shown in Fig. 11(d) [27], which significantly leads to the irregular reflection of the laser beam or shields the workpiece surface against the propagation of laser beam. This unfavorable phenomenon is shown in Figs. 7(a) and 9(b). The transformation from re-focus effect to the irregular reflection effect is actually responsible for such unfavorable characteristics such as porous microstructure [Fig. 7(a)], unevenness surface [Figs. 9(c) and 9(d)], and asymmetrical microgroove structure [Fig. 10(b)], formed in underwater laser microgrooving.
4. Conclusion
In this study, the proposed method of underwater femtosecond laser processing could effectively reduce the debris redeposition around the peripheral region of SiC ceramic microgrooves, thus improving the machining quality of SiC ceramics with femtosecond laser. The underwater femtosecond laser machining is subjected to a dynamic process behavior which is significantly influenced by the laser processing parameters including frequency, pulse energy, scanning speed, and scanning number. With unsuitable parameters, the removal disturbances may occur. High frequency, high pulse energy, and high scanning speed can increase the size or generation rate of gas bubble, leading to the development of the disturbance of underwater processing. Furthermore, the result indicated that the vortex motion emerged in the machining area, and the micro explosion of gas bubbles due to the occurrence of re-focusing effect when the moving gas bubbles happened to collide with the underwater laser beam. The agglomeration of smaller gas bubbles lead to the generation of a large bubble, adhering to the bottom or wall side of microgroove. This significantly affected the uncontrollable reflection, refraction, and scattering of laser beam. The disturbance-free removal occurred when the condition corresponded to low pulse energy, low frequency, and low scanning speed. The gas bubbles were washed away along both the sides of the moving direction of laser beam, which had no effect of the transmission of laser energy. Furthermore, high scanning number eliminated the discontinuity and non-uniformity during the underwater processing. Therefore, the characteristics of laser-induced gas bubbles may be fully utilized by reasonably controlling underwater femtosecond laser processing parameters, and then the different depth of the inverted trapezoidal, U-shaped, and V-shaped microgrooves with high quality and high precision can be obtained.
However, there are still many challenges to be deemed mature and reliable for the industry to fully adopt it [44]. In order to study the effect of water layer on microgrooves of SiC ceramics with femtosecond laser, the detailed research will be carried out in future. Moreover, the effects of other solutions (such as KOH, ethanol, methanol, etc.) on femtosecond laser machining of SiC ceramic will be also studied in future.
Funding
National Natural Science Foundation of China (NSFC) (51735010); Shenzhen Science and Technology Innovation Commission (JCYJ20180306170821261); Program for Changjiang Scholars and Innovative Research Team in University (IRT_15R54).
References
1. S. Baunack, S. Oswald, H. K. Tonshoff, F. von Alvensleben, and T. Temme, “Surface characterisation of laser irradiated SiC ceramics by AES and XPS,” Fresenius J. Anal. Chem. 365(1-3), 173–177 (1999). [CrossRef]
2. A. N. Samant and N. B. Dahotre, “Laser machining of structural ceramics—A review,” J. Eur. Ceram. Soc. 29(6), 969–993 (2009). [CrossRef]
3. B. S. Yilbas and B. Bhushan, “Laser Treatment of Sintered Silicon Carbide Surface for Enhanced Hydrophobicity,” JOM 66(1), 87–94 (2014). [CrossRef]
4. H. Suzuki, M. Okada, W. Asai, H. Sumiya, K. Harano, Y. Yamagata, and K. Miura, “Micro milling tool made of nano-polycrystalline diamond for precision cutting of SiC,” CIRP. Ann. Manuf. Technol. 66(1), 93–96 (2017). [CrossRef]
5. Y. H. Zhao, M. Kunieda, and K. Abe, “EDM mechanism of single crystal SiC with respect to thermal, mechanical and chemical aspects,” J. Mater. Process. Technol. 236, 138–147 (2016). [CrossRef]
6. H. D. Vora and N. B. Dahotre, “Surface topography in three-dimensional laser machining of structural alumina,” J. Manuf. Process. 19, 49–58 (2015). [CrossRef]
7. D. T. Pham, S. S. Dimov, and P. V. Petkov, “Laser milling of ceramic components,” Int. J. Mach. Tools Manuf. 47(3-4), 618–626 (2007). [CrossRef]
8. B. Xie, G. Yuan, and D. Zhang, “Experimental Study of Laser Etching on Al2O3-SiC Composite Ceramics,” Procedia CIRP. 42, 444–447 (2016). [CrossRef]
9. F. Neri, F. Barreca, and S. Trusso, “Excimer laser ablation of silicon carbide ceramic targets,” Diamond Related Materials 11(2), 273–279 (2002). [CrossRef]
10. Y. Xing, J. Deng, Y. Lian, K. Zhang, G. Zhang, and J. Zhao, “Multiple nanoscale parallel grooves formed on Si3N4/TiC ceramic by femtosecond pulsed laser,” Appl. Surf. Sci. 289, 62–71 (2014). [CrossRef]
11. A. Pan, W. Wang, X. Mei, Q. Lin, J. Cui, K. Wang, and Z. Zhai, “Fractal titanium oxide under inverse 10-ns laser deposition in air and water,” Appl. Phys., A Mater. Sci. Process. 123(4), 253 (2017). [CrossRef]
12. Z. Fan, K. Wang, X. Dong, W. Duan, R. Wang, X. Mei, W. Wang, J. Cui, S. Zhang, and C. Xu, “Evaluation of microstructural evolution and corrosion types in ultrasonic assisted laser re-melted thermal barrier coatings under exposure to molten salts,” Mater. Lett. 188, 145–148 (2017). [CrossRef]
13. Z. Fan, K. Wang, X. Dong, R. Wang, W. Duan, X. Mei, W. Wang, J. Cui, S. Zhang, and C. Xu, “The role of the surface morphology and segmented cracks on the damage forms of laser re-melted thermal barrier coatings in presence of a molten salt (Na2SO4+V2O5),” Corros. Sci. 115, 56–67 (2017). [CrossRef]
14. D. Sciti and A. Bellosi, “Laser-induced surface drilling of silicon carbide,” Appl. Surf. Sci. 180(1-2), 92–101 (2001). [CrossRef]
15. Y. Liu, L. Liu, J. Deng, R. Meng, X. Zou, and F. Wu, “Fabrication of micro-scale textured grooves on green ZrO2 ceramics by pulsed laser ablation,” Ceram. Int. 43(8), 6519–6531 (2017). [CrossRef]
16. T. Kurita, K. Komatsuzaki, and M. Hattori, “Advanced material processing with nano- and femto-second pulsed laser,” Int. J. Mach. Tools Manuf. 48(2), 220–227 (2008). [CrossRef]
17. T. H. R. Crawford, A. Borowiec, and H. K. Haugen, “Femtosecond laser micromachining of grooves in silicon with 800 nm pulses,” Appl. Phys., A Mater. Sci. Process. 80(8), 1717–1724 (2005). [CrossRef]
18. T. Chen, W. Wang, T. Tao, X. Mei, and A. Pan, “Deposition and melting behaviors for formation of micro/nano structures from nanostructures with femtosecond pulses,” Opt. Mater. 78, 380–387 (2018). [CrossRef]
19. Y. Yan, L. Li, K. Sezer, W. Wang, D. Whitehead, L. Ji, Y. Bao, and Y. Jiang, “CO2 laser underwater machining of deep cavities in alumina,” J. Eur. Ceram. Soc. 31(15), 2793–2807 (2011). [CrossRef]
20. A. Kruusing, “Underwater and water-assisted laser processing: Part 1—general features, steam cleaning and shock processing,” Opt. Lasers Eng. 41(2), 307–327 (2004). [CrossRef]
21. S. I. Dolgaev, V. V. Voronov, G. A. Shafeev, C. Fauquet-Ben Ammar, J.-M. Themlin, A. Cros, and W. Marine, “Laser-induced fast etching and metallization of SiC ceramics,” Appl. Surf. Sci. 109, 559–562 (1997). [CrossRef]
22. Q. Chen, H. J. Wang, D. T. Lin, F. Zuo, Z. X. Zhao, and H. T. Lin, “Characterization of hole taper in laser drilling of silicon nitride ceramic under water,” Ceram. Int. 44(11), 13449–13452 (2018). [CrossRef]
23. N. Iwatani, H. D. Doan, and K. Fushinobu, “Optimization of near-infrared laser drilling of silicon carbide under water,” Int. J. Heat Mass Transfer 71, 515–520 (2014). [CrossRef]
24. G. Daminelli, J. Krüger, and W. Kautek, “Femtosecond laser interaction with silicon under water confinement,” Thin Solid Films 467(1-2), 334–341 (2004). [CrossRef]
25. N. Muhammad and L. Li, “Underwater femtosecond laser micromachining of thin nitinol tubes for medical coronary stent manufacture,” Appl. Phys., A Mater. Sci. Process. 107(4), 849–861 (2012). [CrossRef]
26. R. An, M. D. Hoffman, M. A. Donoghue, A. J. Hunt, and S. C. Jacobson, “Water-assisted femtosecond laser machining of electrospray nozzles on glass microfluidic devices,” Opt. Express 16(19), 15206–15211 (2008). [CrossRef] [PubMed]
27. S. Mehrafsun and H. Messaoudi, “Dynamic Process Behavior in Laser Chemical Micro Machining of Metals,” J. Manuf. Mater. Process. 2(3), 54 (2018). [CrossRef]
28. A. Menéndez-Manjón, P. Wagener, and S. Barcikowski, “Transfer-Matrix Method for Efficient Ablation by Pulsed Laser Ablation and Nanoparticle Generation in Liquids,” J. Phys. Chem. C 115(12), 5108–5114 (2011). [CrossRef]
29. J. S. Hoppius, S. Maragkaki, A. Kanitz, P. Gregorčič, and E. L. Gurevich, “Optimization of femtosecond laser processing in liquids,” Appl. Surf. Sci. 467–468, 255–260 (2019). [CrossRef]
30. A. Tamura, T. Sakka, K. Fukami, and Y. H. Ogata, “Dynamics of cavitation bubbles generated by multi-pulse laser irradiation of a solid target in water,” Appl. Phys., A Mater. Sci. Process. 112(1), 209–213 (2013). [CrossRef]
31. R. Yao, Y. Zheng, L. Liao, R. Zhou, and Z. Feng, “Surface oxidation behavior in air and O2-H2O-Ar atmospheres of continuous freestanding SiC films derived from polycarbosilane,” Ceram. Int. 44(17), 20974–20983 (2018). [CrossRef]
32. Z. Zhai, W. Wang, X. Mei, M. Li, J. Cui, F. Wang, and A. Pan, “Effect of the surface microstructure ablated by femtosecond laser on the bonding strength of EBCs for SiC/SiC composites,” Opt. Commun. 424, 137–144 (2018). [CrossRef]
33. J. Noack and A. Vogel, “Laser-induced plasma formation in water at nanosecond to femtosecond time scales: Calculation of thresholds, absorption coefficients, and energy density,” IEEE J. Quantum Electron. 35(8), 1156–1167 (1999). [CrossRef]
34. A. Vogel and V. Venugopalan, “Mechanisms of pulsed laser ablation of biological tissues,” Chem. Rev. 103(2), 577–644 (2003). [CrossRef] [PubMed]
35. A. Mohan, A. Udayakumar, and A. S. Gandhi, “High temperature oxidation behaviour of CVD β-SiC seal coated SiCf/SiC composites in static dry air and combustion environment,” Ceram. Int. 43(12), 9472–9480 (2017). [CrossRef]
36. Y. Liu, J. Wan, X. Zuo, L. Cheng, and L. Zhang, “Oxidation behavior of 2D C/SiC composites coated with multi-layer SiC/Si–B–C/SiC coatings under wet oxygen atmosphere,” Appl. Surf. Sci. 353, 214–223 (2015). [CrossRef]
37. N. Dabir-Moghaddam, Z. Liu, and B. X. Wu, “Modeling of the shrinking process of a bubble induced by laser metal ablation in water and experimental verification,” J. Appl. Phys. 121(4), 044908 (2017). [CrossRef]
38. R. Fabbro, J. Fournier, P. Ballard, D. Devaux, and J. Virmont, “Physical study of laser-produced plasma in confined geometry,” J. Appl. Phys. 68(2), 775–784 (1990). [CrossRef]
39. V. Tangwarodomnukun, J. Wang, and P. Mathew, “A Comparison of Dry and Underwater Laser Micromachining of Silicon Substrates,” Key Eng. Mater. 443, 693–698 (2010). [CrossRef]
40. J. Y. Xu, H. Hu, and Y. L. Lei, “Morphological features of silicon substrate by using different frequency laser ablation in air and water,” Appl. Surf. Sci. 317, 666–671 (2014). [CrossRef]
41. J. Koch, S. Taschner, O. Suttmann, and S. Kaierle, “Surface functionalization under water using picosecond and femtosecond laser pulses - first observations and novel effects,” Procedia CIRP 74, 381–385 (2018). [CrossRef]
42. K. L. Choo, Y. Ogawa, G. Kanbargi, V. Otra, L. M. Raff, and R. Komanduri, “Micromachining of silicon by short-pulse laser ablation in air and under water,” Mater. Sci. Eng. A 372(1-2), 145–162 (2004). [CrossRef]
43. L. S. Jiao, E. Y. K. Ng, H. Y. Zheng, and Y. L. Zhang, “Theoretical study of pre-formed hole geometries on femtosecond pulse energy distribution in laser drilling,” Opt. Express 23(4), 4927–4934 (2015). [CrossRef] [PubMed]
44. C. H. Tsai and C. C. Li, “Investigation of underwater laser drilling for brittle substrates,” J. Mater. Process. Technol. 209(6), 2838–2846 (2009). [CrossRef]
