Two-side percussion microdrilling method for improved hole accuracy employing a reflective mirror lens Open Access

Laser drilling has established itself as the prime machining technique¹ for producing microholes among other conventional and nonconventional microdrilling processes.² And, this is not surprising as the technology has important advantages, i.e., it is a noncontact process without any physical force, vibration,³ and tool wear,² and any material can be processed, which are especially important in any micromachining applications. Therefore, laser drilling has become a primary process for producing holes with diameters ranging from a few millimeters down to 1 μm.^4,5 There are many applications of this technology in different industrial sectors, such as aerospace, automotive, energy, electronics, biomedical, jewellery, and home appliances.^6,7 These applications include, but are not limited to: cooling holes for combustors and turbines in aircraft jet engines^3,6 and gas turbines,^4,8 spray holes for fuel injection,^9–11 notches in diesel engines,⁹ holes in the distributors for liquid-fluidized-bed heat exchangers,³ automotive fuel filters, micronozzles, microinjection molds, vias on printed circuit board, high aspect ratio holes for chemical microreactors, optical displays, wireless and optical communications, microholes on probe cards, and surgical needles.⁶ 

Generally, there are three main laser drilling methods: percussion, helical, and trepanning drilling.^12,13 In percussion drilling, a train of pulses is used to irradiate a workpiece at the same spot while the laser beam is stationary relative to the workpiece^2,14 and the hole size is determined only by the beam spot size at focus and the pulse energy.⁴ In the case of trepanning drilling, the beam rotates continuously, around the circumference of the hole,^4,13 while in the helical one, there is an additional simultaneous shift of the focusing plane.^10,12

One of the main advantages of percussion drilling is the possibility to produce smaller size holes compared to those achievable through trepanning or helical drilling, while the processing time is dependent only on the number of irradiated pulses.^2,13 However, microholes produced with this drilling method have some limitations related to the achievable geometrical accuracy, i.e., the hole circularity and taper,^9,13 and the formation of recast together with some material spattering around the holes. Controlling and minimizing the hole taper, especially the differences between the hole entrance and exit diameters, is one of the key issues in this laser drilling method.⁸ The achievable geometrical accuracy can be an issue with other laser drilling methods, too; however, it is much more pronounced in percussion drilling.¹³ 

The main reasons for the taper formation in any laser drilling method are the beam divergence and the increase in the irradiated surface area with the hole depth.^13,15 As the hole depth increases, the divergence leads to energy absorption by a constantly increasing area along the hole walls and hence the fluence decreases, too.^8,16 Another mechanism discussed in the literature is the multiple beam reflections along the hole’s wall that reduce the absorbed energy with the depth increase.^3,4 Other contributing factors have been suggested, too, such as the plasma shielding and the energy absorption by the ejected vapor.¹⁵ 

Having zero-taper and producing holes with parallel walls is required or desirable in many applications,^7,8 e.g., in cooling holes where the cooling efficiency depends on the flow of the cooling air.¹⁶ It is also crucial in producing probe cards for testing electronic chips, particularly when the holes are densely packed into a small area. Ensuring that the hole diameters and their geometrical accuracy are within predefined ranges is essential to prevent any issues in the use of such probe cards, such as looseness or short circuits during chip testing.¹⁷ Therefore, many researchers have been investigating different ways to improve the hole quality, especially by reducing the holes’ taper angles.¹⁶ Generally, the use of trepanning or helical drilling significantly reduces the taper angles. However, such taper reduction can be achieved only on holes with diameters bigger than 100 μm, while in percussion drilling their size is closer to the beam spot diameter at focus. In the literature, most investigations were focused on reducing the taper angle by optimizing the laser parameters, i.e., pulse duration, spatial beam profile, pulse energy, focal position with respect to the workpiece, and pulse frequency.⁹ 

Dhaker and Pandey. studied the effect of different laser parameters on the taper angle of holes produced by laser trepanning and an optimal parameter domain was reported that reduced the hole taper angle.¹³ A similar study but for percussion drilling was conducted by Ghoreishi et al., where they statistically analyzed the effect of different laser parameters on the taper angle. It was reported that the pulse repetition rate and gas pressure did not have an effect on the taper angle, while the focal position, pulse width, pulse energy, and the number of pulses affected the taper angle significantly.¹⁸ 

Low et al. proposed a method to reduce the hole taper by linearly increasing the pulse energy in percussion drilling with the hole depth to achieve a higher pulse energy and thus to compensate the decrease in laser intensity.^8,16 Other researchers investigated multistage drilling methods that had been combining two or more drilling methods to reduce the hole taper. Romoli and Vallini proposed a sequential drilling method in three steps, first forming a pilot hole with a smaller diameter with either spiral or trepanning drilling that was then followed by an enlargement step employing a spiral trajectory.¹⁰ Finally, the material left after the second step was removed through a trepanning step and thus produced a hole with an acceptable quality, but the decrease in the taper angle was not considered in this sequential drilling method. Similar research was conducted by Mincuzzi et al., where a sequence of percussion, spiraling, and trepanning steps was used to produce holes with zero-tapering angle in relatively thick substrates, i.e., 0.9–2 mm; however, relatively large holes were drilled, i.e., minimum diameters of 200 μm.⁷ Other reported approaches for minimizing the taper angle to improve the hole quality include beam shaping, i.e., by employing a top-hat profile,¹⁹ utilizing high numerical aperture lenses,¹⁷ and tilting the laser beam at a specific angle with respect to the hole axis.^7,10

As the amount of research published on laser processing increases, more creative techniques have to be utilized to see any further improvements in the machining quality. This has resulted in more examples of assisted laser machining as well as hybrid machining, which combine conventional machining techniques with advanced techniques. A few examples of the main directions of research are as follows: submerged drilling techniques are being explored to counteract the heat effects of the laser process, i.e., Tao et al. have explored machining under a solution of water and NaCl, which allows for geometrical improvements in the holes as well as reducing the heat effects of the drilling process, while also reducing the soot and recast formation during the process.²⁰ Hybrid techniques allow much easier machining on conventionally difficult-to-machine materials. Usually, this is done by utilizing the heat effects of the laser to soften the material before removal which preserves the cutting edge of the tool as well as enhances the cutting performance. Recently, a laser ultrasonically assisted turning has been gaining popularity due to significant improvements in the cutting force for a variety of materials, i.e., cemented carbide by Zhang et al.,²¹ titanium alloys by Dominguez-Caballero et al.²² and SiC/Al composites by Kim et al.²³ 

However, especially in the area of hard-to-machine materials, laser processing still has room for improvement. Gao et al. have recently utilized a two-side method in drilling hard-to-machine materials, i.e., a stacked structure of Ti6Al4V and C/SiC.²⁴ This resulted in structural as well as geometrical improvements, utilizing the method first reported by Nasrollahi et al. In this approach, the holes are initially drilled from one side, then the workpiece is flipped using a high precision rotary stage to continue drilling the holes from the opposite side.²⁵ The aspect ratio of produced holes was doubled; however, a high precision alignment step with a relatively expensive processing setup was required to improve the geometric accuracy. Another issue with this method was the time required to rotate and align the two sides of the substrates that made this approach practical only for processing arrays of holes. In addition, it is mostly applicable for processing relatively small workpieces.

The research reported in this paper focuses on developing an alternative method for two-sided percussion drilling that employs a reflective concave mirror to refocus the penetrating laser beam from the exit side. Thus, the necessity to rotate and realign the workpiece can be eliminated along with the use of a complex and relatively expensive processing setup in the two-side drilling method. By refocusing the beam on the holes’ exit side which results in some additional ablation, the hole cylindricity can be improved and the taper angles reduced in the percussion drilling method.

In percussion drilling with ultrashort lasers with pulse energies up to hundreds of micro joules, the laser beam can be focused on workpieces with a spot size down to a few micrometers.⁷ The high laser fluence achieved in this way is sufficient to drill relatively low aspect ratio microholes with satisfactory quality. However, when high aspect ratio (more than 10) holes are required, the hole geometrical accuracy is compromised, i.e., holes with higher taper angles are produced.

It is well known that the hole’s depth increases with the number of pulses until a saturation point and a “steady state” in regard to the hole shape is reached.^15,26 After this point, the hole exit diameter remains the same even if the number of pulses continues to increase. Also, shifting the focus of the beam down into the substrate will not be beneficial because part of the beam will be clipped at the hole entrance. Besides the tapering effect, there are some other negative effects on the hole quality in percussion drilling of high aspect ratio holes, such as displacements of the exit hole position in regard to the hole entrance/axis, cylindricity deviations, and formations of multihole exits.²⁶ 

As it was discussed in Sec. I, the main reasons for the hole taper are the increase in the irradiated surface area with the increase in the hole depth that results in a continuous decrease in laser fluence. At the same time, this tapering on the side walls leads to significant deviations of the incident beam angle from normal and thus reduces the laser absorption. Alongside the change of irradiated area, this further decreases the amount of laser light absorbed by the substrate and foremost exponentially decreases the ablation efficiency with the increase in the hole depth. Despite this fluence decrease with the hole depth, the peak fluence is still sufficient to continue ablating material but with declining efficiency and ultimately to penetrate the workpiece.

The experiments were performed on a laser microprocessing platform, LASEA LS-4, depicted in Fig. 3. The platform integrates diode-pumped ultrashort laser source Yuja from Amplitude Systems with a pulse width of 310 fs, a central wavelength of 1030 nm, maximum repetitions of 2 MHz, a maximum pulse energy of 100 μJ, and a maximum average power of 10 W. The beam delivery system includes a scan head (LS-XY20) from Lasea and a 100 mm telecentric focusing lens to achieve a spot size of 25 μm on the top surface of the workpiece. A concave reflectance (mirror) lens from Thorlabs is integrated into the setup by using a three-axis stage over a motorized XY one for mounting the workpiece, and its focal length and diameter are 100 and 25.4 mm, respectively, and the parameters of the mirror were selected to achieve a spot size of 25 μm, to match the spot size of the main beam. The lens is positioned laterally in such a way that the reflected beam is aligned to the main laser beam in Z and its focus is on the exit plane of the sample. This alignment should be done only once when the lens is fixed on the three-axis manual stage while the workpiece can be repositioned precisely with the motorized XY stage for drilling the holes. Since the proposed method requires the penetrating beam to be refocused back exactly at the exit of the drilled hole, the laser source and the optical components in the beam line should be protected from any back reflections, in this particular case, the laser source used has an integrated optical isolator; thus, no further modification of the beamline was necessary as all other components are transparent to the laser wavelength used.

A stainless steel (SS304L) substrate with a 0.45 mm thickness was used in all experiments and they were conducted in a temperature-controlled environment. The beam polarization was circular while pulse repetition rate and pulse energy were fixed at 100 kHz and 100 μJ, respectively, in all drilling trials to make the comparison and analysis of the results easier.

The samples were cleaned ultrasonically, first, for 10 min in water followed by 10 min in acetone and then dried with hot air prior to laser processing. To inspect the opening and exit side of the holes, an SEM (JCM6000 system) was used and for assessing their morphology and size, a high-resolution x-ray tomography (XCT) system, ZEISS Xradia 620 Versa, was used. In all XCT scans, the exposure voltage, power, and time of each projection were set to 140 kV, 21 W, and 2.5 s, respectively, while the resolution/voxel size was approximately 3.7 μm. The XCT scans were segmented and analyzed employing VG studio Max 2022.1 from Volume Graphics and the hole 3D model was created by applying the VG’s advanced surface determination. This 3D model of the hole is obtained to quantify various holes’ parameters, including depth, diameter at different depths, surface area, and volume, and thus to characterize them.

In this study, each measurement was taken from three different holes produced under identical parameters. Error bars in all figures represent ±1 standard deviation based on these three measurements, indicating the variability between the holes.

To demonstrate the feasibility of the proposed laser drilling method and investigate its capabilities, the concave mirror lens was intentionally misaligned laterally regarding the main laser beam through some manual adjustments with the kinematic mount. In this way, the reflected beam was shifted in regard to the penetrating beam axis and refocused back just beside the exit hole as illustrated in Fig. 4. To determine the number of pulses required to penetrate the sample plate and to determine if it can be penetrated by the reflected beam, an array of holes was produced by increasing the pulse number from 500 to up to 60 000 pulses. The beam polarization was circular while the pulse repetition rate and pulse energy were fixed at 100 kHz and 100 μJ, respectively.

The number of pulses required to penetrate the sample with the main beam was determined by analyzing the XCT image of an array of holes as shown in Fig. 5 and was found to be approximately 9400 pulses or the drilling time was 95 ms. The cross section of the produced holes clearly depicts their evolution with the increase in the pulse number.

In the same way, the XCT image of the holes produced with the misaligned reflected beam was analyzed, and it was determined that 45 600 pulses or 456 ms would be required to penetrate the sample. Figure 6 clearly illustrates the evolution of the drilled holes with varying relative numbers of pulses by excluding the initial 9400 pulses used to penetrate from the top side of the sample.

A comparison of hole’s depths achieved with the main and reflected beam while increasing the pulse numbers is shown in Fig. 7. As can be seen in the figure, the relative number of pulses required to penetrate the workpiece with the reflected beam is approximately fivefold higher than the number of pulses necessary with the main beam. This is due to the lower pulse energy of the reflected beam, because of the absorption of the penetrating beam by the walls of the main hole, and to a lesser extent to some other factors such as scattering and plasma shielding.

The extent of the pulse energy attenuation of the main beam due to these factors can be assessed by measuring its power at the hole entrance and exit, i.e., 9.8 and 5.5 W, respectively. In this way, it can be judged that over 40% of the pulse energy is absorbed by the sidewall of the main hole without any additional ablation by the penetrating beam. It would be informative to compare the fluence on the side walls with that achieved with the refocused beam at the hole exit after delivering 9400 pulses.

In the conducted feasibility study, A_re is calculated to be 0.29 by taking into account the constant optical coefficient of stainless steel for 1030 nm wavelength,²⁹ the circular polarization of the beam, and its normal incident angle. The measured average power of the beam at the hole exit was 5.5 W at 100 kHz repetition rate and, thus, E_re is 55 μJ. At the same time, S_re has been calculated in the conducted drilling experiments to be 1935 μm² by assuming the reflected beam spot size to be the same as the original beam diameter. It is important to note that E_re varies during the drilling process and thus this is just a conservative estimate at the time when the machining with the refocused beam starts.

In the conducted empirical study, A_wall is 0.21 as there is a tapering angle (and incident angle) of 2.5° as shown in Fig. 5, while S_wall is 41 315 μm² as assessed based on the created 3D hole’s model, depicted in Fig. 5, by using the VG software.

The conducted quantitative analysis clearly shows the impact that the proposed method has on the drilling efficiency, i.e., by utilizing the pulse energy of the penetrating beam. This improvement of processing efficiency can be explained by the achieved high absorption due to the normal incident angle of the refocused beam and its higher fluence.

In this section, the experiments conducted to investigate the capabilities of the main and reflected beams are discussed, especially when they were aligned and used in “tandem” to drill holes in the sample. The reflected beam was aligned coaxially with the main beam, and arrays of holes were produced to investigate what would be the impact on the holes’ quality. As a reference, holes produced with the conventional percussion drilling method were used to discuss the capabilities of the proposed drilling method.

Specifically, to investigate how the proposed drilling method affects the holes’ quality, three sets of holes were drilled using only the main beam and three different pulse numbers, 10k, 120k, and 1000k, respectively. The trials with each pulse number were repeated three times to judge about the repeatability of the conventional percussion drilling process. The XCT profile data of the holes produced in this way have been analyzed and the results obtained are provided in Fig. 8, i.e., how the holes’ diameters decreased with the increase in the holes’ depths.

As it is depicted in Fig. 8, the entrance and exit diameters of holes produced with 10k pulses or just above the minimum required to penetrate the sample, were 52.5 and 11 μm, respectively.

The use of a higher number of pulses, i.e., 120k and 1 M, led to an increase in entrance diameter to approximately 55 μm in both cases, while exit diameters increased to 15 and 17 μm, respectively. This suggests that the ablation at the exit side of the hole has reached the saturation point, as there was no increase in the entrance diameter and only a marginal increase in the exit diameter. Overall, there were only marginal dimensional improvements.

Then, arrays of holes were produced with the proposed method, where the main and refocused penetrating beams were used in “tandem.” The same laser settings were applied as those used for the conventional percussion drilling method, while the pulse numbers were set to 20k, 50k, 120k and 1000k pulses. Again, three holes were drilled with each of these four different pulse numbers. As shown in Fig. 8, Fig. 9 shows how the holes’ diameters evolved with the increase in the holes’ depths when the proposed two-side drilling method was used. Comparing the hole’s profiles in Figs 8 and 9, it is evident that the entrance sides of the holes produced with both methods are almost identical up to a depth of approximately 150 μm and only afterward the differences are distinguishable.

As depicted in Fig. 9, the diameter of the exit holes increased to 38, 41, 43, and 44 μm with the increase in pulse numbers, i.e., when 20k, 50k, 120k, and 1000k pulses were used, respectively. At the same time, the exit hole diameter was only 11 μm when the drilling was done with the main beam and 10k pulses. Collectively, there were only minor changes in the diameter of the entrance holes, i.e., they increased from 52.5 to 55 μm with the increase in pulse number from 10k to 1000k. The analysis of the hole profiles shows that the holes’ minimum (neck) diameters of 17, 21.5, 22.5, and 23.5 μm were achieved at hole depths of 350, 280, 250, and 275 μm when pulse number increased from 20k to 1000k, respectively. Thus, these results suggest that the saturation point in the two-side drilling method was achieved with 120k pulses and the further increase to 1000k pulses did not bring any noticeable improvements. The increase in the minimum hole diameters and the decrease in the depth at which the hole “neck” occurred should be attributed to the additional ablation at the hole exit due to the refocused beam.

Figure 10 depicts the differences in the shape of the exit holes produced by the two methods. The exit diameters of the holes produced with the two-side drilling method are much bigger than those achieved with the conventional drilling method, and they are similar to their respective entrance ones, due to the additional ablation with the refocused beam.

Regarding the geometrical accuracy of the holes at the exit side, the measurements indicate that the deviation from the circularity of the holes produced with the conventional drilling and two-sided drilling methods is approximately 2 and 6.7 μm, respectively. This suggests that the beam shape of the refocused beam deviates from its axisymmetric Gaussian profile due to light diffraction and imperfections in the original hole circularity as shown in Fig 10(a). Another reason for this decrease in accuracy is the accumulation of material spatter on the surface of the concave mirror used, which impacts the quality of the reflected beam. No mitigation techniques were used for these trials; however, this would need to be considered in further development of the technique. Consequently, the intensity distribution is no longer axisymmetric for the reflected beam. However, if the circularities are normalized with respect to the hole sizes achieved with both methods, the respective deviations from circularity are 11.7% and 15.2%, respectively, which can be considered comparable.

A comparison of the holes’ profiles achieved with the conventional (main beam only) and the two-side percussion drilling methods and the same number of pulses, i.e., 20k and 1000k, is provided in Fig. 11. The improvement of the hole profiles is only marginal when the pulse number increases to 1000k in the case of the conventional drilling method, as shown in Figs. 11(a) and 11(b). On the contrary, in the case of the two-side drilling method, the increase in pulse number from 20k to 1000k led to clear improvements of the hole quality as shown in Figs. 11(c) and 11(d), respectively.

A quantitative analysis of the dimensional and geometrical accuracy of the produced holes with both percussion drilling methods was carried out, too, following the methodology in Sec. II B, and the results are presented in Table I.

The minimum diameter achieved with both drilling methods is distinctly different, i.e., 9.5 and 17 μm with 20k pulse and 15 and 23.5 μm with 1000 pulses. This is a direct result of refocusing the laser at the exit hole on the substrate, which leads to additional ablation up to almost half of the hole depth. This allows an increase in the minimum diameter, which is otherwise impossible to achieve without using another approach for two-side drilling. A detrimental effect on the standard deviation across the diameter measurements can be seen and this is deemed to be due to the accumulation of material spatter on the surface of the mirror which should be mitigated with further development of the technique. Also, the taper angle decreased significantly when the holes produced with the conventional and two-side percussion drilling methods are compared, i.e., from 2.64° to 1.08° and from 2.41° to 0.63° with 20k and 1000k pulses, respectively. This represents a decrease in taper angle by 59% and 74% when 20k and 1000k pulses were used, respectively. As the taper angle is one of the key quality parameters in laser percussion drilling, this substantial improvement clearly demonstrates the advantages that the proposed two-side drill method offers.

Cylindricity of the holes has only a marginal improvement when the results achieved with both methods and 20 K pulses are compared. At 20k pulses, the ablation is still far from its saturation point and a noticeable improvement in cylindricity is achieved when the pulse number increased to 1000 K, i.e., the deviation from cylindricity decreased by 24%, from 20 to 15.2 μm.

As it was the case with the taper angles, the volumetric ratio $( η v)$ achieved with both drilling methods was substantially different, i.e., it increased from 0.24 to 0.30 and from 0.26 to 0.41 when 20k and 1000k pulses were used, respectively. This represents an increase of 57% when comparing the volumetric ratios achieved on holes produced with both methods and 1000k pulses. Also, it is worth noting that the increase in pulse number from 20 to 1000 K pulse numbers led to an increase of only 8% in the volumetric ratio when the holes were produced with the conventional drilling method, i.e., an increase from 0.24 to 0.26. On the contrary, the increase in volumetric ratio is much higher, almost 37%, when the holes were drilled with the two-side method, i.e., an increase from 0.30 to 0.41. This confirms that the available pulse energy is not utilized effectively during the latter stages of the drilling process, due to the constantly changing surface area and incident angles with the increase in the hole depth.

Finally, it is worth stressing that the obtained improvements of dimensional and geometrical accuracy of the holes produced with the proposed two-side method were not at the expense of processing efficiency, especially the specific processing energy required to drill the holes decreased by 29% and 34% when 20k and 1000k pulses were used, respectively (see Table I).

The research reports a novel two-side method that addresses a critical issue associated with the laser percussion drilling method, specifically the achievable dimensional and geometrical accuracy. The conducted drilling trials demonstrate the effectiveness of the proposed two-side percussion drilling method. The results show that the penetrating beam retains a sufficient pulse energy at the hole exit, which can be beneficially deployed to improve the holes’ quality by refocusing it at the hole exit with a reflective concave mirror lens. In this way, the main beam works in tandem with the refocused penetrating beam to ablate the holes from two sides.

The improvements of microholes’ accuracy achievable with the proposed method have been analyzed and they are significant regarding the achievable improvements of the hole minimum (neck) diameter, taper angle, and volumetric ratio, i.e., 44%, 74%, and 57%, respectively, in the conducted experimental trials. These improvements can be attributed, directly, to the better utilization of the available pulse energy by employing the penetrating beam in the drilling process. It is important to stress that these improvements are achieved without compromising the total drilling time while also reducing the specific processing energy. The saturation points in drilling the holes with the two-side method are achieved with the same pulse trains as those employed in percussion drilling.

Compared to the other two-side laser drilling methods that require the workpiece to be flipped, the proposed method does not necessitate the use of time-consuming rotational and realignment procedures, and the laser processing setup is much simpler and more cost-effective to implement. Additionally, this method could be implemented for other laser processing operations, e.g., laser cutting.

The authors have no conflicts to disclose.

Tahseen Jwad: Conceptualization (equal); Data curation (equal); Investigation (equal); Methodology (equal); Writing – original draft (lead). Vahid Nasrollahi: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Validation (equal); Writing – review & editing (equal). Aurimas Turkus: Data curation (equal); Formal analysis (equal); Investigation (equal); Visualization (equal); Writing – review & editing (equal). Ali Gökhan Demir: Validation (equal); Writing – review & editing (equal). Stefan Dimov: Methodology (equal); Project administration (equal); Resources (equal); Supervision (equal); Writing – review & editing (equal).