Investigations into morphology and surface integrity of micro-hole during femtosecond laser drilling of titanium alloy

To increase an aircraft engine efficiency, thousands of cooling micro-holes should be drilled on the gas turbine blade. Because of the superior thermal and mechanical properties of titanium alloy, it is challenging to produce micro-holes with high dimensional and form accuracy using conventional methods. So, in this work, an attempt has been made to produce micro-holes on Ti–6Al–4V using a femtosecond laser. A detailed experiment is performed using full factorial design to comprehend the combined effect of laser process parameters and laser scanning strategies on micro-hole characteristics like hole circularity at entry, exit, taper hole, surface finish, and microstructure. A combination of higher laser fluence and a lower pulse repetition rate improves the entry and exit hole circularity. On the contrary, the taper angle is lowered by increasing laser fluence and pulse repetition rate. The zigzag scanning strategy reduces the hole taper, while the concentric circle scanning strategy improves the circularity of the hole at the entry and exit. It is found that the optimum process parameters for improving micro-hole geometry in Ti–6Al–4V include a laser fluence of 1.90 J/cm2, a pulse repetition rate of 20 kHz, and a concentric circle scanning strategy. The surface finish of the micro-hole deteriorates with increase in laser fluence and repetition rate, while concentric circle scanning yields lower surface roughness with fewer surface imperfections. Furthermore, the hole wall microstructure evolution exhibits deep craters at higher laser fluence and undulating grooves at higher repetition rates.


Introduction
In the present modernization era, industrialists are giving more attention to miniaturization, cost reduction, and dimensional accuracy of manufactured parts due to the increasing demand in automobiles, aerospace, defense, and the biomedical industry.Micro-features such as microholes, micro-slots, micro-dimples, and other micro-shapes require greater dimensional accuracy in various components like fuel filters, fuel injection nozzles, coronary stents, micro-electro-mechanical systems, and gas turbine blades [1].
Nowadays, modern gas turbine engines operate at extremely high temperatures to improve fuel efficiency and engine performance [2].For that, engine components like turbine blades, combustion chambers, and compressor blades require high fracture strength and corrosion-resistant material [3].Among several alloys, titanium alloy is widely used to fabricate aero-engine parts due to its ability to withstand high temperatures, excellent corrosion resistance, and lower density [4].To increase the life of aeroengine parts, thousands of cooling micro-holes should be drilled on the turbine blade.These cooling holes allow cool air from the internal cooling channel to circulate through the blade surface and thereby it augments the service life of the component [2,5].However, drilling micro-holes in titanium alloy using conventional methods is difficult for machining industries due to low thermal conductivity and high chemical reactivity [5].This can be alleviated using unconventional techniques such as laser micro-drilling, electrochemical drilling, and electric discharge drilling [3,5].
Among several unconventional methods, laser micro-drilling is a suitable technique to drill micro-holes in conductive and non-conductive materials, regardless of their material properties [5].
In recent decades, long pulsed lasers such as milliseconds and microsecond lasers have been widely used in aircraft industries to produce micro-holes on various engine components [6].Due to the longer pulse duration, laser-induced defects such as micro-cracks, recast layer, and heat-affected zone are observed around the drilled hole, thus affecting the service life of the engine components [6,7].Recently, an ultra-short pulsed laser [Picosecond (ps) and Femtosecond (fs)] having pulse durations less than 10 ps is gaining attention among researchers and industrialists due to its ability to produce micro-holes with a less heat-affected zone (HAZ), minimal recast layer, and good surface quality.Nevertheless, ultra-short pulsed lasers can provide peak power in the gigawatt range within a short period, enabling them to ablate hard and brittle materials with less thermal distortion [4,7].
In spite of the remarkable benefits of an ultra-short pulsed laser during micro-drilling, several drawbacks, like low material removal rate, high processing time, circularity deviation, spatter, and presence of recast layer, are still challenging for engine manufacturers [8].Various researchers have conducted experiments on the ultra-short pulsed laser to understand the laser material interaction and to identify the significant parameters affecting the hole geometry.Chichkov et al. studied the ablation mechanism using a femtosecond laser (fs) on steel, copper, aluminum nitride, and silicon plates, showing that fs laser creates vapor and plasma phase, resulting in the improvement of hole geometry [9].Li et al. studied the influence of laser fluence on surface ablation and hole morphology using fs laser on nickel-based superalloy.It was noted that debris accumulation is observed on the surface of the hole at various fluence [10].A comparison of fs and ps laser percussion drillings using a pulse repetition rate and laser power was performed on copper and stainless steel, indicating that the fs laser is more efficient than the ps laser while drilling holes in metals [11].The influence of laser processing parameters on the micro-drilling of various metal foils using fs laser pulses was studied by Zhu et al., and they observed that fs laser is not proficient in drilling micro-holes in thin foils having a low melting point because of less material removal rate [12].Jeong et al. performed experiments to study the geometry of the hourglass-shaped drilled hole in a diamond using an ultra-short laser by varying the laser energy and the pulse number.The result shows that as the laser power is increased, the taper increases slightly along with the entrance and exit diameter due to the immense heat accumulation during the process [13].
Dhaker et al. performed experiments to optimize the laser parameters on nickel-based superalloy to maintain hole circularity and to reduce recast layer thickness using response surface methodology.The results conclude that the developed model shows an error percentage of 8.61% and 12.24% for circularity and recast layer thickness with experimental results [14].Tam et al. executed a laser drilling operation to generate a high aspect hole using Nd: YAG laser on nickel-based superalloy material to optimize the laser process parameters using the Taguchi design of experiments and observed that the laser energy, pulse duration, and pulse shape significantly affect the drill quality [15].Using Taguchi methodology, laser input parameters were optimized on nickel-based superalloy to achieve the lowest recast layer thickness.The ANOVA results conclude that the peak power, assist gas pressure, and focal position significantly influence the recast layer thickness [16].Optimizing the laser process parameters like pulse width, pulse frequency, focal length, current, and air pressure using response surface methodology on the titanium nitride-alumina was executed by Biswas et al.The optimum results for obtaining better hole circularity and lower taper are higher values of the lamp current, moderate air pressure, positive focal length, and reduced pulse width [17].Prior findings indicated that there is a need to optimize the laser process parameters for creating quality micro-holes on difficult-tocut materials using an ultra-short pulsed laser.
Wang et al. conducted experiments to enhance the surface quality of the micro-hole on the thermal barrier-coated nickel superalloy using various pulse shapes.The result shows that the ramp-down pulses could generate holes with less taper than ramp-up and uniform-shaped pulses [18].Romoli and Vallini proposed a micro-drilling cycle on martensitic stainless steel using an ultrafast fiber laser by dividing the drilling process into three different phases as a pilot through hole drilling, roughing phase, and finishing stage, obtained the micro-hole without burrs and debris within 3.25 s with 99% of repeatability [19].The method of a combined pulsed laser by coupling the nanosecond laser with a millisecond laser was proposed by Jia et al. to increase the quality of the drilled hole on alumina ceramic material.The result reveals that micro-holes of less than 200 µm on a 1-mm-thick alumina ceramic substrate were fabricated by coupled pulsed laser [20].Effect of scanning method on the roundness and taper of the nickel-based superalloy was investigated by Wang et al., indicating that the spiral scanning offers better roundness.Additionally, the result shows that the spiral strategy with different process parameters, such as scanning speed and average power, has improved the hole quality [21].
From the prior literature, it is inferred that only limited studies have been conducted on the morphology and surface integrity of micro-holes drilled on titanium alloy.The poor surface quality and geometry of cooling holes on gas turbine blades reduces airflow and affects the cooling efficiency [22].Therefore, a detailed investigation is required to understand the effect of fs laser parameters on microhole quality.Also, the effect of laser scanning path on hole morphology, surface roughness, and microstructure analysis remains a knowledge gap in the literature.The main objective of the work is to identify the significant laser process parameter and laser scanning methods affecting the microhole geometry and morphology.Furthermore, the surface roughness of the inner side hole wall, the morphology of the hole profile, and microstructures of the inner side wall of the drilled hole are examined to understand the microhole quality.

Experiment methods
In this work, a femtosecond laser is used to produce a microhole of 100 µm on a Ti-6Al-4V sheet (ASTM Grade 5) of 0.5 mm thickness [21,23].The photograph of the femtosecond micro-drilling center (Model-IR-50-40, Make: Huaray) having three-axis movement is utilized in this experimentation as shown in Fig. 1 a and close-up view in Fig. 1b.The pulse width, wavelength, and spot diameter of the femtosecond laser used in this work are 350 fs, 1.035 µm, and 40 µm, respectively.The femtosecond laser has Gaussian profiled laser energy distribution with an average power of 40 W and can vary pulse repetition rate from 1 to 800 kHz.
Based on an elaborated literature review and several pilot runs, repetition rate (RR) and laser fluence are the utmost influential femtosecond laser parameters on micro-hole geometry and morphology [24].So, detailed experiments have been carried out by varying laser fluence, repetition rate (RR), and scanning strategies (SS) such as linear continuous, concentric circle (inside to out), zigzag, and linear discontinuous, which is demonstrated in Fig. 2. In the linear continuous scanning strategy, the laser beam will move in a straight line in a bidirectional manner without any breaking in the scanning path, as seen in Fig. 2a. Figure 2b portrays a concentric circle in which the laser beam will move circularly from the inside to the outside.In the zigzag path, the laser beam follows a Z-shaped path to remove the material layer by layer from the surface, as shown in Fig. 2c. Figure 2d shows the linear discontinuous path in which laser beams move uni-directional, having discontinuity in the scanning route after each linear path.Scanning pitch of 0.8 µm and a pulse overlap of 50% are kept constant throughout the experiments.All the experiments were performed in ambient atmospheric conditions without any shielding gases.
The full factorial design of experiments is performed to analyze the effect of the main factor and its interaction with output responses.The full factorial design will allow the main factors with different levels and generate an experimental design that includes every possible combination of factors and their levels.Furthermore, the analysis of variance (ANOVA) is computed using Minitab software to determine the percentage contributions of each main factor and on the output responses.According to Table 1, each input parameter, such as laser fluence and pulse repetition rate, is varied at three levels, while scanning strategies are varied at four levels with three replications for each set of runs.The average value of each response is used as an output measure.The details of the experiment run with performance characteristics are given in Table 2.

Mounting table
Fig. 1 Femtosecond laser micro-drilling center After performing the experiments, the samples were first polished with several grades of emery paper, including 600, 800, and 1200 to remove the spatter on the surface of the drilled hole before taking the SEM micrographs [25].Then, the polished samples were ultrasonically cleaned using diluted ethanol for 10 min to remove the fragments of emery paper trapped inside the micro-hole.The microhole geometry, such as entry hole diameter and exit hole diameter, were measured by analyzing the SEM micrographs (Scanning electron microscope, Make: Carl Zeiss, model: EVO 18) using ImageJ software.For measuring the hole diameter, six diameters (i.e., D 1 , D 2 , D 3 , D 4 , D 5 , and D 6 ) are measured along the circumference of laser drilled hole at 30° interval for both the entrance and exit side as shown in Fig .3 [14].The circularity of the micro-hole at the entry and exit is obtained using Eqs.( 1) and (2), respectively.
(1) Entry hole circularity, where D min and D max are the minimum and maximum diameter of the hole at the entry side as shown in Fig. 3. Also, d min and d max are the minimum and maximum diameter of the hole at the exit side.The value of C entry and C exit equals one, indicating that the hole is perfectly circular [17].
Taper angle (Ɵ) of the drilled micro-hole is evaluated using Eq. ( 3) [17] After completing the geometrical measurement, the laser dilled micro-hole is sectioned using a shearing machine, then gradually grounded till the hole center, and finally polished using emery paper.Then, the polished samples were ultrasonically cleaned using diluted ethanol for 10 min.Then, the samples were mounted on the universal tribometer integrated with a three-dimensional profiler [(Make: Rtec instruments, Model: MFT-5000)], to measure the surface roughness on the inner hole wall of the micro-hole.Furthermore, the sectioned micro-hole is etched with 1 mL of HF, 3 mL of HNO 3 and 46 mL of H 2 O to reveal the hole wall microstructure [26].3 Results and discussion

Computation of ablation threshold
Ablation threshold is the minimum amount of laser fluence needed to perform efficient ablation.To produce a good quality micro-hole, laser fluence must be equal to or greater than the material ablation threshold.The threshold fluence is evaluated by forming a relationship between hole diameter (D), material threshold fluence ( th ), laser spot radius (ω o ), and the peak laser fluence ( o ) [27].
The peak laser fluence ( o ) and pulse energy (Ep) are obtained by Eq. ( 4) Pulse energy (Ep) is evaluated using Eq. ( 5) where P is the average laser power (i.e., 40 W) and f is the RR.
Finally, the ablation threshold ( th ) is evaluated using Eq. ( 6) Figure 4 depicts the correlation between ablation threshold and RR for various laser fluence which is investigated in the prior work by Deepu et al. (2023).The saturation of threshold fluence is observed after 10 kHz irrespective of variation in laser fluence, termed as a critical repetition rate (CRR) [24].Based on the analysis, it is determined that the laser fluence can vary from 1.78 to (5) 2.04 J/cm 2 , and RR from 10 to 20 kHz for effective drilling of the micro-hole.

Circularity of the micro-hole at the entry side
The main and interaction effects of the laser process parameters on the hole circularity at the entry side were studied using the main effect plot and the interaction plot.The main effect plot in Fig. 5 elucidates the entry hole circularity of the drilled micro-hole.The main effect plot represents the means of entry hole circularity for each parameter and its corresponding levels.It is evident from the main effect plot that hole circularity increases with an increase in laser fluence.The mean of entry hole circularity at a fluence of 1.78, 1.90, and 2.04 J/cm 2 results in 0.74, 0.75, and 0.82, respectively.At a lower fluence of 1.78 J/cm 2 , the laser energy is insufficient to remove the material uniformly from the workpiece.Melting is more predominant than vaporization resulting in the uneven distribution of laser energy around the hole edge, thereby reducing the hole circularity.Increase in fluence from 1.90 to 2.04 J/cm 2 , an improvement in the hole circularity is observed due to gentle ablation.Here, complete vaporization of the material from the machining spot results in uniform distribution of laser energy along the hole periphery [28].In addition, the high laser fluence prevents molten material from accumulating near the hole entry and effectively discharges it from the hole, resulting in reduced spatter buildup near the hole entrance.A similar trend is observed during the micro-hole drilling of alumina ceramics using the carbon dioxide laser [29].
The main effect plot shown in Fig. 5 demonstrates that an increase in pulse repetition rate (RR) causes a decrease in hole circularity at the entrance side.The mean hole circularity of 0.80, 0.78, and 0.74 is obtained for 10 kHz, 15 kHz, and 20 kHz, respectively.At a lower RR of 10 kHz, the highest hole circularity of 0.80 is obtained due to high time separation gap between the subsequent laser pulse resulting in better hole geometry.Also, saturation in the ablation threshold is obtained at a critical RR of 10 kHz resulting in higher hole circularity.Increase in RR from 15 to 20 kHz; the hole circularity diminishes gradually.Because the inter-pulse-off time reduces, the pulse repetition rates increase; thereby, more laser energy is accumulated on the material surface, resulting in higher material ablation.Hence, hole circularity reduces.In addition, the molten material gets agitated when the pulse repetition rate increases because of less time for the molten material to get solidified.As a result, the molten material is settled unevenly near the edges of the microdrilled hole resulting in a decrease in hole circularity at the entry side.Similar behavior is reported during micro-drilling TiN-Al 2 O 3 composites using a pulsed Nd: YAG laser [30].The main effect plot in Fig. 5 indicates that the circularity of the entry hole is affected by several laser scanning strategies.The better entry hole circularity of 0.82 and 0.81 is obtained for linear continuous and concentric scanning strategies, followed by linear discontinuous having a circularity of 0.77.The zigzag method results in a poor entry hole circularity of 0.70.While using a linear continuous scanning method, the laser scanning path is bidirectional parallel lines having start and end points of each scanning path located at the edge of the hole, as given in Fig. 2a, and it continues to move in a bidirectional parallel manner without any discontinuity in the path.Because of the continuous laser scanning path, the material gets sufficient laser energy to remove the material from the allotted cross section, thereby improving the circularity of the hole at entry.A similar result is obtained while adopting the concentric circle scanning path from inside to outside, as shown in Fig. 2b.The laser beam is concentrated initially at the center of the hole, resulting in higher material removal at the center.In addition, the ablated material is easily flushed out from the entrance and exit sides of the hole, contributing to an increase in the circularity of the entry hole [6].In addition, the hatching space considered in this experiment is less than the laser spot diameter, resulting in larger scanning space, enhancing the material removal rate, and increasing the hole circularity [31].
In the linear discontinuous, as shown in Fig. 2d, the starting and end points of the laser beam are located on the edges of the hole and move in a uni-directional path consuming a longer time for the beam to complete the scanning path.Moreover, the discontinuity between the scanning path reduces the laser energy in the desired cross-sectional, thereby reducing the material removal rate, due to which hole circularity reduces [31].In the case of the zigzag scanning path demonstrated in Fig. 2c, in addition to the linear path movement of the laser beam from the start to the end point located at the edge of the hole, the beam is also moving in an inclined manner connecting the end point of first path to the start point of the following linear path.Due to this additional path movement, more laser energy is infused onto the material surface, leading to strong ablation, thereby reducing the circularity.
Figure 6 demonstrates the interaction plot for the entry hole circularity.The interaction effect between the laser fluence and repetition rate (i.e., laser fluence and RR) is the dominant factor affecting the entry hole circularity.When the laser fluence is at 1.78 and 2.04 J/cm 2 , there is a significant difference in hole circularity for various pulse repetitions.At 1.78 J/cm 2 , the pulse repetition rate of 15 kHz provides better hole circularity, followed by 10 kHz.In contrast, no significant difference in the entry hole circularity is observed for various repetition rates at 1.90 J/cm 2 .Higher entry hole circularity is observed at 2.04 J/cm 2 fluence and 10 kHz pulse repetition rate.This is because, at higher fluence and lower repetition rate, the complete melting and vaporization of the material take place.From the experimental results, the effect of laser fluence and RR on the entry hole circularity for various laser scanning strategies is correlated using a linear regression analysis and is given in Eqs. 7, 8, 9 and 10.
Entry hole circularity for, Fig. 5 The main effect plot for the entry hole circularity The ANOVA test is used to determine whether laser process factors significantly impact the output responses.The P value was used to determine the significance of the parameters and their interaction.A 95% confidence level (P = 0.05) was used while conducting the ANOVA.Therefore, if the P value is less than 0.05, the input parameter and their interaction are significant.No third-order interaction   3 illustrates that the P value obtained for laser fluence and interaction of laser fluence and repetition rate is 0.033 and 0.031, which is less than the 'α' level of 0.05 means the laser fluence and interaction effect of laser fluence and repetition rate are significant.The ANOVA indicated that the laser fluence, scanning strategy, and the interaction between the laser fluence and repetition rate are the most significant parameters affecting the entry hole circularity, with the contribution of 13.68%, 20.04%, and 22.91%, respectively.In addition, the RR has a lower effect on the entry hole circularity.
Figure 7 shows the SEM images of micro-hole entry side morphology processed at different laser fluence and scanning strategies.It is clear from the images that micro-cracks are visible at higher laser fluence of 2.04 J/cm 2 .Figure 8 demonstrates the morphology at the entry side at various pulse repetition rates and scanning strategies.It shows that at higher repetition rate, slight increase in heat-affected zone is observed.The similar result has been reported for both the observation during laser percussion drilling of titanium alloy [32].The concentric laser scanning strategy gives better hole circularity with minimal surface defects.Linear discontinuous and zigzag method leads to micro-crack near the hole periphery.

Circularity of the micro-hole at the exit side
The main and the interaction effects of the laser process parameters on the exit hole circularity are studied using the main effect and interaction plot.Figure 9 elucidates that as the laser fluence increases, the hole circularity at the exit hole increases.The mean of exit hole circularity of 0.80, 0.85, and 0.86 is obtained for 1.78.1.90, and 2.04 J/cm 2 , respectively.At a lower fluence of 1.78 J/cm 2 , an exit hole circularity of 0.80 is obtained.It is because, as penetration depth increases, the side wall of the hole absorbs a part of laser energy.Therefore, the laser energy available at the exit of the hole is not enough to remove the material uniformly from the hole edge resulting in lower exit hole circularity.
Increased laser fluence from 1.90 to 2.04 J/cm 2 produced a hole with better circularity.Because at higher laser fluence, the energy available at the hole exit is high enough for effective ablation, thereby preventing the accumulation of debris at the exit side [25,30].
The main effect plot for hole circularity at the exit side shown in Fig. 9 demonstrates that as the pulse repetition rate increases, the circularity of the exit hole decreases.The mean exit hole circularity of 0.85, 0.84, and 0.83 is obtained at 10, 15, and 20 kHz, respectively.Because at a lower RR of 10 kHz, the cooling time between the upcoming pulses is longer, preventing the accumulation of laser energy onto the material surface, and causing lower material removal, thereby increasing the hole circularity.Moreover, molten material is getting more time to cool and get closer to the solid state resulting in higher exit hole circularity of 0.85 [28,33].As the RR increases from 15 to 20 kHz, saturation in the ablation threshold is noticed, and more laser energy is infused on the machining spot, resulting in high thermal accumulation, producing the hole with less exit hole circularity.
It is observed from the mean plot that the concentric circle scanning strategy gives a higher exit hole circularity of 0.91.This is because the laser beam is initially concentrated at the center of the hole and moves in a concentric manner from inside to outside.This will prevent the laser beam from getting deflected from the spot because of multiple reflections caused by the entrapped plasma plume in the hole.As a result, the energy loss is much less, providing better hole circularity at the exit [6].The zigzag scanning method delivers poor exit hole circularity of 0.75 because of discontinuity in the scanning route, resulting in the oval-shaped hole exit.
As seen in Fig. 10, the interaction effect between the laser fluence and repetition rate, repetition rate, and scanning strategy is the dominant factor affecting the exit hole circularity of the through micro-hole.During the interaction between laser fluence and repetition rate, a significant difference in exit hole circularity is observed among various repetition rates at 1.90 J/cm 2 .The highest exit hole circularity is observed at 1.90 J/cm 2 and 10 kHz repetition rate.While in the case of interaction between the repetition rate and scanning strategies, a significant difference in exit hole circularity is observed at 20 kHz.Moreover, the concentric circle scanning method provides better exit hole circularity at 20 kHz.The ANOVA table (Table 4) demonstrates that the scanning strategy and the interaction between the laser fluence and repetition rate, repetition rate, and scanning strategy are considerably influencing the exit hole circularity, with a P value of 0.002, 0.050, and 0.028, which is less than the α value of 0.05.The scanning strategy contributes 30.47%, followed by interaction among laser fluence and repetition rate, repetition rate, and scanning strategy having 13.85% and 22.97%.
The SEM images in Fig. 11 show the micro-hole morphology at the exit side of the holes processed at different laser fluence and scanning strategies.Zigzag and linear continuous method results in a micro-crack at the exit side of the hole at laser fluence of 1.78 J/cm 2 and 1.90 J/cm 2 .Moreover, debris accumulation is observed at 2.04 J/cm 2 while using the zigzag strategy.No surface defect is observed for the concentric circle and linear discontinuous method with increased laser fluence.Figure 12 illustrates the morphology of the exit side of the micro-hole at various pulse repetition rates and scanning strategies.Micro-cracks are observed for zigzag at a repetition rate of 15 kHz and 20 kHz.No surface defects is noticed for concentric, linear continuous, and linear discontinuous at various repetition rates.

Hole taper of the micro-hole
The main effects of the laser process parameters on the hole taper were studied using the main plot.The main effect plot in Fig. 13 depicts that the hole taper angle decreases with the increase in laser fluence.The mean taper angle of 0.077, 0.070, and 0.056 radians is obtained for 1.78, 1.90, and 2.04 J/cm 2 , respectively.At a lower fluence, a weak plasma plume resulting in the generation of less recoil pressure

Micro-cracks
Fig. 12 SEM image at different repetition rates and scanning strategy for exit hole Fig. 13 The main effect plot for the taper which is not strong enough to eject debris from the hole leads to the scattering and reflection of the upcoming laser beam.As a result, fluence available at the exit is less resulting in small exit diameter which leads to higher taper angle [34].A substantial decrease in the taper angle is observed as the laser fluence is raised from 1.90 to 2.04 J/cm 2 .A higher laser fluence of 2.04 J/cm 2 produces a large diameter hole at the entry and exit sides.Because after the laser beam penetrates through the work material, the energy available at the exit side of the hole is high enough to remove the accumulated debris from the hole.This increases the ablation rate to form a larger diameter hole at the exit side, resulting in less taper angle.A similar phenomenon was observed during laser percussion drilling on stainless steel and nickel alloy [33,35].
It is observed from the mean effect plot that the taper angle of 0.074, 0.066, and 0.064 radians is obtained for 10, 15, and 20 kHz, respectively.The molten material solidified as dross near to the hole exit due to the less time intervals between the pulses at a lower repetition rate of 10 kHz.As a result, the hole exit side diameter decreases and the taper angle increases.An increase in RR from 15 to 20 kHz results in a lower taper angle of 0.066 and 0.065, respectively.Due to less pulse-off time between the pulses, more laser pulse energy is available at the exit side, resulting in a high ablation rate.This leads to a higher exit hole diameter and a lower taper angle.The same trend is observed during the laser drilling of microholes in titanium alloy and nickel-based alloy using Nd: YAG laser [32].The main effect plot demonstrates that the least taper of 0.058 is obtained for the zigzag strategy.In the zigzag strategy, as shown in Fig. 2 c, in addition to the parallel scanning path, the laser beam travels from the endpoint of the first scanning path located at the hole edge to the start point of the following scanning path in Z-shaped.Due to the additional path traveled by the laser beam between the two parallel scanning routes, more heat energy is supplied throughout the thickness of the material, which will enhance the diameter of the hole at the exit side.As a result, the taper angle decreases.While in the case of the linear discontinuous method, a higher taper angle is observed because of the interruption in the scanning path.As a result, less laser energy is supplied to the work material in the desired location, reducing the material removal rate as the penetration depth increases.Hence, a small diameter hole at the exit is formed.
The ANOVA table given in Table 5 elucidates that the P value for the laser fluence is 0.030, which is less than the 'α' level of 0.05 means the laser fluence is significant.And no  The optimum levels of processing parameters needed to generate a perfect hole geometry by femtosecond laser micro-hole drilling are identified.The optimal values of the femtosecond laser process parameters are identified with their predicted responses by keeping the confidence level of 95% for all the intervals.The optimal process parameter setting and estimated output response are given in Table 6.
The confirmation test is conducted at laser fluence of 1.83 J/cm 2 and RR of 15 kHz for different scanning strategies, as given in Table 7.The regression model is validated with experimental results and percentage of error is calculated using Eq. ( 7) [36].Table 7 shows that average absolute percentage of error for the entry hole circularity, exit hole circularity, and taper angle are 2.71%, 2.19%, and 3.93%, respectively.

Influence of laser fluence on the surface roughness of micro-hole wall
The entrance, middle, and exit areas are three separate zones where the inner wall surface roughness (Ra) is measured, and these areas are highlighted in Fig. 14.Regardless of variations in laser fluence, the surface roughness in the entrance area is higher than that at the middle and exit areas.This is a result of the micro-cracks, ravines, re-solidified layer, and fish scale structure that are present, as seen in Fig. 16a-c.Multiple axial feed for drilling a through hole have also affected the surface finish at the entrance area [37].The thermal energy provided by the laser pulses is sufficient to completely melt and vaporize the material at a low laser fluence of 1.78 J/cm 2 , which results in better surface roughness (Ra) of 1.78, 1.67, and 1.59 µm at the entrance, middle, and exit areas, respectively, as shown in Fig. 15.With the laser fluence increased further to 1.90 J/cm 2 , there is an increase in the surface roughness of 1.90, 1.73, and 1.69 µm at the entrance, middle, and exit area due to the existence of ravines and debris, as shown in Fig. 16b.With a further increase in laser fluence (2.04 J/cm 2 ), the inner hole wall surface quality deteriorates because of immense pulse energy as well as due to the presence of ravines at entrance area and crater at exit portion.A similar trend was observed during the micro-cutting of alumina/aluminum composites using pulsed Nd: YAG [38].

Influence of pulse repetition rate on the surface roughness of micro-hole wall
It is observed from Fig. 17, the surface roughness in the entrance region is higher for all the repetition rate as explained in Sect.3.4.At repetition rate of 10 kHz, better surface roughness (Ra) of 1.52, 1.49, and 1.48 µm is observed at the entrance, middle, and exit area, respectively.This is because, at low repetition rate the time gap between the successive pulses is long enough for the molten material to solidify results less surface defects as shown in Fig. 18a.Moreover, the laser pulse energy supplied is sufficient for the gentle ablation with any surface defects.When the repetition rate is increased to 15 kHz, the surface roughness (Ra) is increased to 1.73, 1.65, and 1.67 µm at entrance, middle, and exit area, respectively.This is due to the presence of ravines, recast layer, and micro-cracks, as shown in Fig. 18b.The surface finish deteriorates as the repetition rate is increased to 20 kHz because there is less space between pulses, which leads to an increase in laser pulse energy resulting in surface distortion.Moreover, the less spatial time between the pulses prevents the ablated material from being completely ejected.As a result, a re-solidified layer forms on the hole wall, which causes micro-cracks, as seen in Fig. 18c.A similar result was observed during the micromachining of a silicon wafer using an ultrafast pulse laser [39].This phenomenon is in line with the result obtained during laser micro-drilling of nickel-based super alloy using a picosecond ultra-short pulse laser [40].

Influence of scanning strategies on the surface roughness of micro-hole wall
The variation of surface roughness with different scanning strategies is shown in Fig. 19.Compared to other scanning strategies, the concentric circle has a lower surface roughness at the entrance, middle, and exit areas.This is due to the laser beam continuously moving in a circular motion along    the case of linear discontinuous, surface roughness obtained is of higher magnitude due to discontinuity in the scanning path and stair-step effect.

Microstructure evolution of micro-drilled hole wall
Figure 21 illustrates the microstructure evolution of microhole walls at various laser fluence.At low laser fluence of 1.78 J/cm 2 , the undulated grooves are observed with no evident micro-cracks and craters.When laser fluence is increased to 1.90 J/cm 2 , shallow craters are formed, resulting in an uneven surface.At higher laser fluence of 2.04 J/ cm 2 , deep craters are observed on the hole wall, implying intense ablation due to severe heat accumulation.The above observation is in line with femtosecond laser ablation on titanium alloy [42,43].
The microstructural evolution of micro-hole drilled at different repetition rates is shown in Fig. 22 Stalagmite structure is observed at a lower repetition rate of 10 kHz, whereas at 20 kHz undulating groove structure is formed.This is because, at a lower repetition rate, the complete re-solidification of melt occurs before the next laser pulse arrives.In contrast, at a higher repetition rate, the melt layer generated by the first laser pulse interacts with the subsequent pulse, leading to undulated groove microstructure on the hole wall surface.Moreover, no significant micro-cracks are observed for various repetition rates.A similar observation is noticed during the micro-machining of titanium alloy using a femtosecond laser at various repetition rates [44].
Figure 23 depicts the microstructure of the hole wall at various scanning strategies.In the case of linear continuous and linear discontinuous, ripples and micro-cracks are observed on the hole wall, whereas the concentric circle scanning strategy results in an undulated groove microstructure [40].While for the zigzag method, severe microcracks and debris accumulation is noticed on the hole wall microstructure.

Conclusion
This study investigated the effects of laser scanning paths and process parameters on the morphology and surface quality of micro-holes drilled on titanium alloy using a femtosecond laser.The following is a list of the conclusions.
• The hole circularity is better for concentric circle scanning path at a combination of higher laser fluence (2.04 J/ cm 2 ) and a lower pulse repetition rate (10 kHz).• When comparing different scanning paths, the zigzag scanning method results in less taper because the longer distance that the laser beam travels increase the amount of heat at the hole exit side.• An analysis of four different scanning paths reveals that the concentric circle scanning path has a better surface finish while the linear continuous, linear discontinuous, and zigzag scanning surfaces have defects such as stairstep effect, micro-cracks, crater, and ravines.• Variation in the microstructure is observed for different laser fluence and repetition rates.Undulating groove type microstructure changed to deep crater type structure with increased laser fluence, while stalagmite type gets modified into an undulating groove structure as repetition rate increased.

Fig. 3 Fig. 4
Fig. 3 Measurement of hole diameter at the entry side of fs laser drilled hole

Fig. 7 Fig. 8 Fig. 9
Fig. 7 SEM image at different laser fluence and scanning strategy for entry hole

Fig. 11
Fig. 11 SEM image at different laser fluence and scanning strategy for exit hole

Table 2
Experimental results of femtosecond laser micro-hole drilling process

Table 3
ANOVA table for the hole circularity at the entry df degree of freedom, SS sum of squares, MS mean sum of squares significantly impacted any output response in this situation; hence, it is not considered in the analysis.The ANOVA table given in Table

Table 4
ANOVA table for the hole circularity at the exit df degree of freedom, SS sum of squares, MS mean sum of squares

Table 7
Confirmation test and validation of model