The effect of ns bursting on percussion drilling of metal is investigated experimentally and analytically, and compared with the efficiency and quality of drilling using single ns pulses. Key advantages are demonstrated, correlating well with the results from a thermal theoretical model. The 1064 nm bursts contain up to 14 pulses of various pulse widths and spacing, and at frequencies of tens of MHz within the burst. The individual pulses have pulse widths of 10 to 200 ns, and up to 12 kW peak power. Burst repetition frequency is single shot to 500 kHz.
©2011 Optical Society of America
A wide array of lasers-material interactions are described in the literature using a variety of lasers, each with its own pulse shape, pulse energy, pulse repetition frequency (PRF), wavelength, and average power [1–5]. The behavior of the material to incoming radiation depends on the mechanism for energy transfer from the laser beam to the material.
In the CW to nanosecond regime, this energy transfer is thermally based, and can be understood in terms of the increase of the temperature of the material as pulse energy is absorbed, balanced by the diffusion of heat into the substrate. This behavior influences the quality and speed of the machining process. Alternately, this provides for opportunities of significantly improving the machining process if such energy transfer is optimized.
It is also understood that pulse width, peak power and pulse energy, each can have significant influence on ablation depth, cutting, drilling or annealing speed. More generally, the quality of the end product is highly dependent on the type of laser used and how the energy is delivered. This has fuelled development of shorter pulses (ps and fs) for machining operation in the interest of improving the quality of the machining process, while minimizing processing time and post-machining processes [6–9]. The minimization of processing time includes increasing of the efficiency of material removal, while maintaining the quality of the end product.
Several techniques have been outlined in the literature regarding improvement of the machining operation by tailoring the shape of the delivered pulses so as to control the response of the material as the pulse is absorbed [10–17].
Other techniques involve enhancing the efficiency of material removal by the use of multiple pulses in rapid succession. These pulses are formed using multiple lasers delivering synchronized pulses that are a few tens to hundreds of ns apart. This approach demonstrates the delivery of two or more completely different pulses, delivered with an adjustable delay between them [18,19]. Other techniques rely on forming bursts of identical pulses by picking off and amplifying a series of pulses from a mode locked laser . An alternate approach is to generate a burst of pulses by a seed laser diode, and amplifying the burst using a MOPA (master-oscillator power-amplifier) fiber laser .
In this study, we analyze the effects of changing pulse width, peak power, and pulse energy on processing speed and machining quality for percussion drilled metal foil. Enhanced material removal is demonstrated using the MOPA laser configured to produce bursts of pulses of various number and shapes.
The experimental observations also correlate well with the results from a theoretical thermal model. This model is based on calculating the temperature rise of the substrate after absorption of incoming laser energy. We also show that the temperature of the surface after pulse absorption is highly dependent on the peak power and pulse duration of such incoming pulses.
2. Pulsed MOPA fiber laser
The pulsed MOPA fiber laser used in these tests is Model MOPA-DY, by Multiwave Photonics (shown in Fig. 1 ). The system’s configuration is based on a pulsed seed laser diode, whose output is amplified by a series of fiber amplifiers, ending with an optically isolated collimator. The laser can be externally triggered to emit pulses at pulse repetition frequency (PRF) from single shot up to 500 kHz. The externally selectable pulse widths range from 10 to 200 ns. Peak power is up to 12 kW, pulse energy is up to 0.6 mJ, average power is up to 20 W, output polarization is random, and operating wavelength is 1064 nm. The beam propagation factor M2 is measured and remains stable at 1.25 as the various laser’s parameters, such as peak power, PRF, average power, and/or pulse widths are adjusted.
In addition to the classical system operation, where the laser is triggered to emit individual pulses at the pulse width and PRF of choice, the laser can be triggered to emit a burst of preconfigured pulses at that same PRF. Each burst is composed of a series of pulses, with pulse widths, number and spacing configurable in any random arrangement (see for example Fig. 2 ). In addition, the pulses within each burst can be a mixture of different pulse widths, whereby the overall width for the group can be up to 1000 ns. Within the burst, the effective PRF can be tens of MHz, as shown in Fig. 2.
The bursts are treated by the laser as single effective pulse, triggered and amplified at a rate from single shot to 500 kHz. This feature, together with the flexibility of organizing any sequence of pulses in a single burst, is referred to as Dynamic Pulsingtm.
The flexibility of the laser stems from the design concept of using a modulated laser diode as the seed of the laser system. Having direct access to the driving current of the seed allows one to configure the laser to emit pulses at a variety of pulse widths, sequence, pulse separation, and frequency, and can even switch to CW operation if desired. This is in contrast to Q-switched systems that depend heavily on operating at a particular resonance frequency of an intracavity element, and produces pulses having limited range of pulse widths and shape.
3. Experimental configuration
The system of Fig. 1 is configured for percussion drilling as shown in the schematic of Fig. 3 . The beam is focused to a 20 μm spot diameter at the surface of a stainless steel sheet of type AISI 316L, 25 μm thick. The number of pulses needed to percussion drill the metal sheet is counted by using a detector, a counter and additional circuitry to interrupt the laser’s emission. The PRF to trigger the laser’s emission is set at 100 Hz, to allow for accurate pulse detection and counting, although the MOPA laser itself can be triggered up to 500 kHz.
A series of tests are conducted where the number of pulses per burst is adjusted, while counting the number of bursts required for drilling clear holes through the stainless steel sheet. The detection threshold and the position of the detector are also adjusted to ensure that clear and not partial holes are measured. In addition, the peak power, pulse width and pulse energy are premeasured to enable accurate peak power and energy correlation as the burst parameters are adjusted.
4. Why pulse peak power is important
Ablation of material depends upon the temperature of the surface after pulse absorption. The temperature of the surface after absorption of an incident laser pulse is shown here to depend on peak power and duration of that pulse. Consequently, the efficiency of the drilling process depends on peak power and duration of the incident pulse.
To illustrate this behavior, an analytical model is built for the temperature increase after an incident Gaussian laser beam is absorbed. This temperature increase can be approximated for a high absorption coefficient material by [13,22,23]
Equation (1) is applied to a stainless steel substrate, and a suggested incident sequence of five pulses, of different peak powers, but all have the same 10 μJ pulse energy, as shown in Fig. 4(a) . These pulses are assumed to have widths of 10, 20, 50, 100 and 200 ns, respectively, and pulse separation of 25 ns. The focused spot diameter on the surface is set at 50 μm. The top trace of Fig. 4(b) shows the calculated temperature rise for the surface of the material, showing that the rate of temperature rise depends on the peak power of the incident laser pulse.
This dependence can be understood in terms of the rate of energy deposition, balanced by the coefficient of thermal conductivity of the substrate. Thereby, pulses of higher peak powers lead to substantially higher surface temperatures, as shown in Fig. 4(b), top trace. Subsequent traces in Fig. 4(b) represent the increase of the temperature of the material on axis and incrementally below the surface at depths of 1, 2, 3, 4, 5, 6 and 7 μm, respectively. The shaded area in Fig. 4 represents the melting region for stainless steel. The lower threshold is the melting point (about 1600 K), while the top level is its boiling point (about 3000 K).
The ablation process of material has several components. It includes material evaporation as its temperature exceeds its boiling point, a phase explosion, a shock wave and subsequent ejection of melt, as well as optical breakdown and plasma formation. Raising the surface temperature to above the evaporation temperature for the material is an important element of this process. The point where material is ejected from the substrate either by direct evaporation or by a combination of the above is referred to here as ablation threshold. Note that operating close to the ablation threshold may include excessive melting, and so is not often desired.
As seen in Fig. 4(b), the 10 ns pulse is the only one of these five pulses that drove the surface temperature above the boiling point for steel, despite the accumulation of energy and substrate temperature increase as succeeding pulses are deposited. The last pulse, which is 200 ns wide, has almost no temperature increase. Note also that the temperature of the material below the surface depends on the accumulation of energy.
Surface ablation, therefore, is more sensitive to peak power, or irradiance, while substrate melting is influenced by the deposited energy, or fluence. This behavior influences the process of percussion drilling as ns laser pulses are applied.
5. Percussion drilling with single-pulsed MOPA laser
The percussion drilling setup of Fig. 3 is used to count the number of pulses required to drill clear holes as pulse widths and peak powers are adjusted, while the MOPA laser is set to emit individual pulses. This implies that the MOPA laser is operated such that a single pulse is emitted for every input trigger, at the PRF of choice from single shot to 500 KHz, instead of the burst of five pulses shown in Fig. 2(a). This arrangement is also equivalent to bursts that contain one pulse only.
Figure 5 shows the percussion drilling result using individual pulses, where the pulse width is incrementally adjusted from 10 ns to 200 ns. Three peak power levels are used: 2, 5 and 10 kW. These correspond to irradiance levels of 0.6, 1.6 and 3.2 GW/cm2, respectively.
One observation of Fig. 5(a) is that for percussion drilling of metal in the ns regime, it is more efficient to use longer pulses of lower peak powers. This is the consequence of the thermal processes that dominate the laser-material interaction in this regime, as well as the type of substrate used, i.e. metal. Note that this percussion drilling experiment is performed at low repetition frequency to allow for accurate counting of the number of pulses. This lower PRF diminishes the effect of cumulative heating as subsequent pulses are deposited. It is expected that operating at higher PRF would enhance the process of material removal by ejection of liquid material due to the generated shock waves.
Increasing peak power or pulse energy is also seen not to accelerate drilling speed. The increase of peak power is, however, seen to contribute more thermal energy, melt more material, as well as provides a stronger phase explosion and shock wave to violently disperse the ablated material. This process is shown in the SEM photos of Figs. 5(b)-5(d), for the same 100 ns pulse widths, but for three different fluence levels. Hence, it is often preferable to operate with lower irradiance levels to improve the quality of the drilling operation. On the other hand, operating with higher irradiance and fluence levels lead to removal of higher volume of material.
Pulse emission from the system of Fig. 1 is measured and calibrated in terms of its pulse width, peak power and pulse energy. Consequently, the vertical axis of Fig. 5(a) can be rescaled in terms of the total energy required to drill clear holes in that 25 μm thick stainless steel foil. This is shown in Fig. 6(a) , which highlights that pulses with lower peak power require less total energy to penetrate the same metal foil. For example, 200 ns pulses that are 0.68 kW in peak power require only 20% of the energy of a 10 kW pulse to penetrate to a same depth. Consequently, a high peak power laser is better utilized for drilling metal foils if its output is divided into several parallel beams that drill simultaneous holes, thereby increasing throughput significantly.
6. Percussion drilling with bursts of pulses
The flexibility imbedded in the MOPA system of Fig. 1 allows one to change the number of pulses in each burst, without changing the peak power of emitted pulses. The percussion drilling experiment of Fig. 3 is repeated whereby the number of pulses per burst is incrementally adjusted from N = 1 to 14, while recording the number of pulses required to drill through the 25 μm stainless steel sheet. Additionally, the peak power of the emitted pulses is calibrated and the test is repeated where the peak power of the first pulse is adjusted to 1, 2, 5, and 10 kW, sequentially. Note that the test is conducted such that incrementally adding pulses to the burst, i.e. N = 2 to 14, does not alter the peak power of the first pulse, or subsequent pulses within the burst.
The individual pulses for this test are set for 30 ns wide, and pulse spacing at 50 ns, corresponding to an effective frequency of about 12.5 MHz. The choice of pulse width and spacing can be arbitrary. Although for this case, the pulse width is chosen to provide sufficient energy for material removal, while maintaining uniform peak power within the individual pulse.
The ablated material absorb and reflect some of the incident pulse energy, thereby shield the substrate and prevent some of the incoming pulse energy from reaching the surface. Two expanding fronts, indicated in , are associated with material ablation: a shock and a vapor wave front. The first, which accounts for most of the shielding, is the faster of the two. It dissipates after a few tens of ns. The second front become effective after 100 ns and gradually dissipate in a few microseconds. The pulse spacing (50 ns) is chosen since it is longer than the dissipation time constant for the expanding shock wave. This 50 ns pulse spacing is however still short enough when compared with the thermal diffusion for 25 μm stainless steel foil.
The thermal diffusion time constant for the material is experimentally verified by adjusting the pulse spacing from 50 to 300 ns, while looking for the onset of surface oxidation and discoloration. The change of surface texture and color suggests the onset of thermal damage by overheating. The area that shows these thermal effects is termed heat-affected zone (HAZ), and becomes visible when pulse spacing is more than 150 ns.
Figure 7 shows SEM photos of percussion-drilled holes performed using bursts of pulses of different pulse number and peak power. All the holes are clear holes. However, as is shown in these photos, the entrance and exit diameters depend on the test parameters.
The peak power indicated in Figs. 7(a), 7(b) and 7(c) corresponds to the peak power of the first pulse, and those correlate to irradiance levels of 1590, 636, and 318 MW/cm2, respectively. All the photos shown in Figs. 7(a)-7(c) are of the same magnification, as indicated by the scale shown at the bottom left of all photos, which is for a distance of 70 μm.
Figure 7 shows that changing the number of pulses N in a burst does not substantially change the quality of the drilled hole. For the same peak power, as N increases, the size of the emitted particles increase, and they become less in quantity. This is because of the enhanced melting and ejection of liquid particles. The lower peak power group of Fig. 7(c) shows far less dispersion of the material than the higher peak power levels. Figure 7(c), at N = 10, shows the onset of substrate melting, which occurs due to operation close to the ablation threshold.
The most dominant observation of Fig. 7 is the strong ablation and material dispersion at high peak powers. This behavior mirrors the same observation as shown in Figs. 5(b)-5(d) and the comments associated with section 4. Note that the increase of peak power does not lead to acceleration of drilling speed. Although, the additional pulse energy as peak power is increased contributes to melting of the substrate and ejection of more material. This is evident by the substantially larger diameter of the entrance hole (see for example Fig. 7(a)).
Figure 5(a) suggests that this drilling speed does not have a strong dependence on peak power, but requires a certain number of pulses, provided that these pulses have irradiance and fluence levels above the ablation threshold. This concept is directly applicable to the results of Fig. 7, where the number of pulses per burst is increased; thereby accelerating the drilling process by accelerating the delivery of pulses to the material (for a similarly applied trigger frequency). This is illustrated in Fig. 8 , which uses the setup of Fig. 3 and the counted bursts of Fig. 7 to show that increasing the number of pulses per burst accelerates drilling.
Figure 8 shows the decrease of the required number of bursts to drill through the metal sheet as the number of pulses in a burst is incrementally increased from N = 1 to 10. Note the geometrical progression reduction of the required number of bursts as N is increased. For example, at N = 1, the number of required pulses of 1 kW peak power is 43. At N = 2, the number of required bursts, of two pulses per burst, is 21, giving a total number of pulses of 42, which is roughly equal to that at N = 1. The difference is that drilling is completed at half the required number of trigger points for that same PRF.
Similarly to Fig. 5, Fig. 8 does not show a strong dependence on peak power, as it is adjusted to 1, 2, 5 and 10 kW. Although, at N > 10, the required number of 1 kW peak power bursts is seen to substantially rise. That’s because the 1 kW bursts of N > 10 do not produce a strong enough shock wave that efficiently ejects the liquid metal from the hole, which causes the hole to close on itself, thereby increasing the required bursts to penetrate the foil.
7. Analytical model
In this section, we derive a formula for the number of pulse bursts needed to percussion drill a hole to a certain depth. The laser beam incident on the material is defined by a Gaussian spatial profile and a temporal profile represented by g(t). The intensity is specified by , where P is the power, w is the Gaussian beam radius (1/e pulse amplitude), and r is the radial position. The temperature rise for a single pulse at any radial point r and depth z in the irradiated material can be approximated accurately by ,Equation (2) neglects laser energy absorption by generated plasma plume. Such absorption becomes more important for more energetic pulses that lead to denser plasma or electron density.
Equation (2) gives the temperature rise due to an input pulse of rectangular temporal form. Naturally, for a Gaussian beam, the center of the beam at r = 0, has the highest temperature rise. Equation (2) can be inverted to calculate the ablation depth, z = h, by setting the temperature rise to the boiling temperature (3000 K for stainless steel), and setting r to zero, which results in the ablation depth for a single pulse,
Figure 8 shows the correlation between experimental and theoretical results, showing very good agreement for N < 10. This figure shows the results for the number of bursts required to drill through a 25-mm thin metal sheet, while increasing the number of pulses per burst, and as peak power is increased. The pulse profiles and images of the drilled holes are shown in Fig. 7.
In applying this model, we used an empirical value for the parameter for the absorbed energy fraction , where x is the average pulse power in units of kW. The reduction of this parameter as the pulse average power increases, reflects the fact that the ablated material carries away with it part of the absorbed energy, consequently, less energy participates in the heating process. For example, Fig. 7(a) shows that higher peak powers lead to substantially more material removal by the ablation process, causing the entrance aperture to be much larger than the exit hole.
The effect of pulse bursting, also referred to as Dynamic Pulsing  or multipulsing, is investigated for the case of percussion drilling in the ns regime of a metal foil. A pulsed MOPA fiber laser is configured to emit a variety of pulses of well-controlled pulse widths, peak power and energy, are used to show that bursting accelerates the material removal process, by accelerating the delivery of pulses to the surface. This enhancement is particularly visible while operating relatively close to the ablation threshold.
A developed theoretical thermal model calculates the temperature of the material in the presence of thermal diffusion and using an input beam of Gaussian spatial profile, showing good agreement with the experimental observations. For simplicity, a square temporal pulse profile is used. The process of drilling is explained in terms of temperature increases of the substrate due to absorption of the incoming pulse energy, and ejection of all material that is above the boiling point. Agreement is demonstrated between theory and experiment for ablation depth values. Since the model is thermal in nature, we expect it to be limited to pulse widths where thermal processes dominate, such as with pulse widths and spacing of longer than 10ns.
Dynamic Pulsing offers many advantages to the user, some of which are outlined above. Other possibilities include rapidly increasing the laser’s maximum PRF. For example, consider the inset photo in Fig. 2. By setting the number of pulses per burst to 10, and the width of each pulse to 200 ns, while operating at a burst repetition frequency of 500 kHz, one can effectively produce a continuous stream of pulses at an effective PRF of 5 MHz. Effectively increasing the laser’s maximum PRF by an order of magnitude. In machining and thin film scribing, this significantly higher repetition frequency is desirable as it produces finer scribed edges that resemble those from a mode-locked laser.
References and links
1. J. F. Ready, Effects of High-Power Laser Radiation (Academic Press, 1971), pp. 79.
2. R. E. Russo, “Laser ablation,” Appl. Spectrosc. 49(9), 14A–28A (1995).
3. J. C. Miller and R. F. Hagland, Jr., eds., Laser Ablation and Desorption (Academic Press, 1998).
4. B. C. Stuart, M. D. Feit, S. Herman, A. M. Rubenchik, B. W. Shore, and M. D. Perry, “Optical ablation by high-power short-pulse lasers,” J. Opt. Soc. Am. B 13(2), 459–468 (1996).
5. D. Breitling, A. Ruf, and F. Dausinger, “Fundamental aspects in machining of metals with short and ultrashort laser pulses,” Proc. SPIE 5339, 49–63 (2004).
6. X. Zeng, X. I. Mao, R. Greif, and R. Russo, “Experimental investigation of ablation efficiency and plasma expansion during femtosecond and nanosecond laser ablation of silicon,” Appl. Phys., A Mater. Sci. Process. 80(2), 237–241 (2005).
7. B. Luther-Davies, A. V. Rode, N. R. Madsen, and E. G. Gamaly, “Picosecond high-repition-rate pulsed laser ablation of dielectrics: the effect of energy accumulation between pulses,” Opt. Eng. 44(5), 051102 (2005).
8. J. König, S. Nolte, and A. Tünnermann, “Plasma evolution during metal ablation with ultrashort laser pulses,” Opt. Express 13(26), 10597–10607 (2005). [PubMed]
9. D. M. Karnakis, G. Rutterford, and M. R. H. Knowles, “High power DPSS laser micromachining of silicon and stainless steel,” Proceedings of the Third International WLT-Conference in Manufacturing, Munich (June 2005).
10. H. Herfurth, R. Patwa, T. Lauterborn, S. Heinemann, and H. Pantsar, “Micromachining with tailored nanosecond pulses,” Proc. SPIE 6796, 67961G, 67961G-8 (2007).
11. S. T. Hendow, S. A. Shakir, and J. M. Sousa, “MOPA fiber laser with controlled pulse width and peak power for optimizing micromachining applications,” Proc. SPIE 7584, 758417, 758417-6 (2010).
12. P. Deladurantaye, D. Gay, A. Cournoyer, V. Roy, B. Labranche, M. Levesque, and Y. Taillon, “Material micromachining using a pulsed fiber laser platform with fine temporal nanosecond pulse shaping capability,” Proc. SPIE 7195, 71951S, 71951S-12 (2009).
13. S. T. Hendow and S. A. Shakir, “Structuring materials with nanosecond laser pulses,” Opt. Express 18(10), 10188–10199 (2010). [PubMed]
14. S. T. Hendow, P. T. Guerreiro, N. Schilling, and J. Rabe, “Pulse shape control of a MOPA fiber laser for marking of stainless steel and other materials,” M604, ICALEO Proceedings (2010).
15. M. Rekow, R. Murison, T. Panarello, C. Dunsky, C. Dinkel, and S. Nikumb, “Application of a pulse programmable fiber laser to a broad range of micro-processing applications,” M603, ICALEO Proceedings (2010).
16. H. Pantsar, T. Eisenbeis, M. Rekow, R. Murison, H. Herfurth, and S. Heinemann, “process optimization for improving drilling efficiency in WET solar cell manufacturing,” M1102, ICALEO Proceedings (2010).
17. S. Hendow, “Influence of Peak Power and ns Pulse Duration on Micromachining,” in Fiber Laser Applications, OSA Technical Digest (CD) (Optical Society of America, 2011), paper FThC4.
18. A. C. Forsman, P. S. Banks, M. D. Perry, E. M. Campbell, A. Dodell, and M. Armas, “Double pulse machining as a technique for the enhancement of material removal rates in laser machining of metals,” J. Appl. Phys. 98(3), 033302 (2005).
19. B. R. Campbell, T. M. Lehecka, J. Thomas, and V. Semak, “A study of material removal rates using the double pulse format with nanosecond pulse laser on metals,” in Proceedings of ICLEO, paper 401 (2008).
20. R Knappe, “Scaling ablation rate for picosecond lasers using burst-pulse micromachining,” ICALEO Proceedings, M602 (2010).
21. S. T. Hendow, J. R. Salcedo, R. Romero, and P. T. Guerreiro, “Dynamic pulsing of a MOPA fiber laser for enhanced material processing,” Proc. SPIE 7914, 791405, 791405-6 (2011).
22. A. Ganino, Korn and Theresa M. Korn, Mathematical Handbook for Scientists and Engineers (Dover Edition, 2000), pp. 119.
23. B. Luther-Davies, A. V. Rode, N. R. Madsen, and E. G. Gamaly, “Picosecond high-repition-rate pulsed laser ablation of dielectrics: the effect of energy accumulation between pulses,” Opt. Eng. 44(5), 051102 (2005).