Bright future for fibre lasers?

Professor Michalis Zervas, from the Optoelectronics Research Centre, charts the rise of high-power fibre lasers and asks what next for the technology?

Modern manufacturing has been revolutionised by lasers used in sectors such as automotive, aerospace, and electronics to weld cars, drill holes in turbine blades, fabricate 3D parts, and mark metals and plastics. According to Strategies Unlimited, industrial laser revenues in material processing alone amount to $4.5 billion, accounting for 30 per cent of total laser revenues. Over the last decade, fibre lasers – the latest entry in the manufacturing arena – have shown spectacular progress in power scaling and performance, making them the fastest growing business, currently worth $2 billion (CAGR +13 per cent) at the expense of traditional technologies such as CO₂ lasers (CAGR -1 per cent) and DPSS lasers (CAGR -3 per cent). According to an Allied Market Research report, the fibre laser market, which was valued at $1.8 billion in 2017, is projected to reach $4.4 billion by 2025, growing at a CAGR of 12 per cent from 2018 to 2025. This growth is expected to be dominated by increasing demand in high power fibre lasers to cater for existing applications, as well as emerging sectors such as additive manufacturing and 3D printing.

Modern manufacturing has been revolutionised by using fibre lasers in sectors such as automotive, aerospace, medical and electronics

The success of fibre lasers in manufacturing is attributed to a number of properties and performance characteristics that set them apart from competing solid-state and other laser technologies [1]. Fibre lasers comprise monolithic, all-fibre, compact cavities without moving parts or free-space optics, resulting in small footprint, maintenance-free manufacturing tools with fast turn-key operation. They show stable output and excellent beam quality over extended power ranges. This is a result of the superb inherent beam control, thanks to their uninterrupted light guiding properties, and efficient heat removal, thanks to the large surface-to-volume ratios of the fibre geometry. Moreover, all these characteristics are complemented by record wall-plug efficiencies (currently 40 per cent), resulting in minimum electric power consumption and cooling requirements.

The fibre laser power scaling success relies heavily on cladding pumping, a scheme where low-cost, high divergence diode pumps – single-emitters, bars or stacks – are combined geometrically and/or spectrally before being launched into the much larger, higher numerical aperture (NA) – and therefore more accommodating – fibre cladding, in which pump power is gradually absorbed as it propagates and crosses the active core repeatedly. Laser light then emitted from the smaller and lower NA core is much more intense and directional, resulting in three to four orders of magnitude brightness enhancement, compared to the pumping diodes [1]. This higher brightness largely accounts for fibre lasers’ superior material processing and other industrial application capabilities.

Figure one shows the evolution of Yb3+-doped fibre laser power at 1μm over time. In the case of single-fibre, single-mode, near-diffraction-limited fibre lasers, output power has reached 3-5kW with diode-pumping master-oscillator power amplifier (MOPA) arrangements (red dots). Further power scaling to 10kW has been achieved with in-band or tandem-pumping (green dots), where fibre lasers are used to pump fibre lasers, which results in minimum thermal load at the final lasing stage [2]. Further power scaling of single-fibre, single-mode output has stagnated over the last decade, primarily because of the onset of nonlinear effects, such as stimulated Raman scattering and stimulated Brillouin scattering (SRS/SBS) and, more recently, by transverse modal instabilities, as discussed below.

Figure 1: The evolution of Yb3+-doped fibre laser power at 1μm over time

However, kilowatt-level single-mode laser modules have been combined incoherently to an astounding 100kW multimode output (blue triangles), suitable for heavy-duty industrial applications [3]. In addition, thanks to their superior spectral characteristics, beam quality and output power stability, tens to hundreds of 0.5-1kW single-mode fibre lasers can be combined coherently or spectrally to achieve near-diffraction-limited outputs, currently reaching 60kW levels (orange squares) [4]. Coherent or spectral combination is emerging as a powerful fibre laser power scaling technique, promising to reach, or even exceed, the 100kW near-diffraction-limited beam power mark with high efficiency in the near future, enabling a number of advanced directed energy applications.

Industrial laser application space

In addition to raw power, industrial laser efficacy depends also on the beam quality, quantified by the beam parameter product (BPP), which determines the beam focusing ability, spot size and depth of focus at the workpiece. Different applications have varying power and BPP requirements, depending on the specific laser-material interaction. In general, for all laser technologies power scaling comes at the expense of higher BPP (worse beam quality). It can be argued that the lower the BPP can be maintained with increasing power, the more applications can potentially be addressed. Fibre lasers, in single-mode or multimode beam format, offer record output powers (10-100kW) with high beam quality (BPPs in the range 0.35-16mm-mrad), significantly outperforming all other competing laser technologies.

It is this superlative parameter combination that has enabled fibre lasers to achieve the fastest growth and highest market penetration in industry. They have replaced existing laser technologies, dominated entire application sectors, such as marking, and enabled new applications, such as micro-drilling, sintering and flexography, as well as new laser cutting variants, such as gas-free remote cutting, which particularly benefit from the high beam quality [5].

However, success rarely comes without hiccups. In the case of single-mode high-power fibre lasers, the extreme intensities associated with multi-kilowatt near-diffraction-limited beams pose a number of challenges in the processing heads and other optics, such as unacceptable focus shifts when working with high-intensity laser beams [6]. The resultant power-dependent focal shifts can result in lower processing quality, or even process failure. New special processing heads and new ways to optimise laser-matter interaction have to be developed to harness the full potential of ultra-bright high-power fibre lasers. It is ironic that single-mode high-power fibre lasers have eliminated thermal lensing from the lasing medium, only to pass it further down to the focusing optics.

Ultimate power scaling limits

In addition to well-known optical nonlinear effects, such as SRS/SBS, power scaling in single-mode fibre lasers and amplifiers has been seriously hindered by a new nonlinear effect, namely transverse modal instability (TMI) [7]. TMI amounts to a threshold-like onset of transverse spatial mode competition that severely degrades the output beam quality. TMI has been observed experimentally in a large variety of Yb3+-doped optical fibres under different pumping and seeding conditions (see [1] and references therein). In the case of broad line width operation (∆ν > 25GHz), when plotted against the active fibre core diameter, the high-power fibre amplifier TMI threshold appears to be largely inversely proportional to the core area (i.e. 1/D2), regardless of the type of fibre used (see figure two, top data set, blue dashed line). However, in the case of narrow line width operation (∆ν < 25GHz), the dependence of the TMI threshold on the core diameters appears to be much more severe (see figure two, lower data set, red dashed line), demonstrating that the seed line width has a strong effect on the TMI threshold. It should be mentioned that, on the contrary, SRS/SBS power thresholds increase with fibre core area, placing conflicting requirements on the fibre design.

Figure 2: Core diameter plotted against transverse modal instability for broad linewidth and narrow linewidth lasers

In practical monolithic high-power fibre laser systems the active fibres should be bendable, which restricts the maximum cladding diameter to around 600μm for mechanical reliability issues. Mechanical reliability, along with TMI and SRS and current pump brightness, limit output powers to about 28kW and 52kW with diode- and tandem-pumping, respectively. To achieve these power levels, special low-moded fibres with core diameters in the 45-55μm range have to be developed. Setting a practical limit of current maximum core diameter to approximately 35μm, the power limits reduce to about 15kW and 25kW, respectively [8].

What next?

It has been apparent so far that single-mode high-power fibre lasers in their current stage (stage I) of development already provide enough raw power not only to serve existing but also to enable new industrial applications. This leads inevitably to the question: what next?

Fibre technology – the most controllable, low-loss wave guiding technology – offers a number of different attributes, which so far have been largely unexplored. These include spatial features, in the form of well-defined and stable modes, wide spectral characteristics accessible with different dopants, and nonlinearities and plurality of polarisation states. It is then obvious that fibre technology offers the possibility to not only generate efficiently, but also manipulate and deliver photons remotely to the workpiece. This provides the opportunity to proceed with the next generation – stage II – of high-power fibre laser development (see figure three), where smart photon engines can be combined with smart photon pipes, to build the ultimate manufacturing tools with added functionality and reconfigurability, advanced laser and process monitoring capabilities, meeting the requirements of the emerging digital manufacturing era. The new disruptive features are expected to augment the industrial and other application parameter space, further increasing fibre laser market penetration.

Figure 3: Stage II of high-power fibre laser development involves combining smart photon engines with smart photon pipes

The smart photon engine features include:

• Ultimate efficiency – greater than 55 per cent wall-plug efficiency is possible – by optimum core and splice design, to minimise electricity requirements and running costs, and turn high-power fibre lasers into the greenest of manufacturing tools.
• Maximum power, beam stability and power scalability by mitigating SRS/SBS and TMI effects. Stable and reliable single-mode fibre lasers with greater than 10kW outputs – 28-52kW has been predicted as possible – can expand remote material processing capabilities [5], offering flexibility and high operational speed [9]. They will also minimise the required number of high-power fibre laser modules and, therefore, make it easier and more economical for them to be combined to the 100kW-300kW range with brighter multimode output beams. With this power, they can meet the requirements of demanding applications, such as deep welding, rock-cutting [3], and heavy-plate welding for ship-building and pressure-vessel fabrication [10].
• Wavelength plurality, using other rare-earth dopants – such as thulium and holmium, and optimised silica hosts – highly efficient multi-kilowatt fibre lasers in the 2μm eye-safe spectral region can be developed. Non-oxide glass fibres, on the other hand, could provide access to the 2-3μm mid-infrared spectral region [11], to further increase the material processing range. In addition, novel fibre designs can increase nonlinearities to enable development of efficient optical parametric oscillators (OPOs) or supercontinuum sources. Using nonlinear wavelength conversion, the fibre laser output can be extended into the green or UV spectral region [12], for machining insulators – clear glasses, plastics and ceramics – 2D materials such as graphene, and organics and semiconductors.
• Polarised output: high birefringence active and passive fibres can enable efficient generation and stable delivery of linearly polarised light at multi-kilowatt power level. Laser beams linearly polarised along the cutting direction are already shown to result in cutting speed improvements [13]. In addition, single-mode linearly polarised light can be transformed into more exotic polarisation beams [14] with superior cutting performance [15] – or facilitate coherent combination for power scaling – and nonlinear wavelength conversion for increasing the wavelength coverage.

Smart photon pipe features, on the other hand, include:

• Flexible long-distance power delivery: special fibre designs, such as distributed SRS filtering fibres [16], can mitigate nonlinear effects, and enable single-mode delivery over long distances. To deliver kilowatt-level powers over more than 100 metres, limitations set by optical nonlinearities – such as SRS and four-wave mixing – and unwanted spatial-mode mixing effects, should be eliminated by proper fibre design or novel excitation or splicing techniques [17]. New exciting developments in low-loss hollow-core fibres [18], on the other hand, open up the possibility of delivering single-mode multi-kilowatt optical powers over long distances [19].
• Beam shape agility and control has already shown significant edge-quality improvements in mild steel cutting [20]. Novel multi-ring fibres enable fast on-the-fly beam shape reconfigurability at 500kHz, and promise new advanced marking and micromachining applications [21]. Specially designed fibres can provide complex multi-beam shapes tailored to a particular process, such as advanced burr-free cutting, replacing cumbersome and expensive bulk optics [22], or external spatial light modulators [23].
• Integrated laser and process control: special delivery fibres can be enhanced with embedded taps and other diagnostics to control not only the laser but also the manufacturing process. Such smart fibre umbilicals, carrying photons and information, will help laser tools to be fully integrated in the wider manufacturing process. This is an important step towards harmonisation, with forthcoming digital manufacturing and Industry 4.0 requirements. 

Figure 4: Smart photons, in addition to power, will carry bits of information that will be used to control thge manufacturing tool and the process

Looking further into the future, it can be envisaged that high-power fibre laser evolution can be followed by a third stage of development (figure three), offering single-beam near-diffraction-limited outputs in the hundreds of kilowatt range by spectrally or coherently combining a number of smart units with ultimate efficiency and power stability. It is impossible to predict the industrial impact of such ultra-bright high-power fibre lasers, and it is likely that they will struggle to find suitable widespread manufacturing applications. It is, though, certain that they will have a tremendous impact on directed energy applications, as well as ambitious endeavours such as future fibre-based particle accelerators [24] and futuristic projects such as space debris removal [25] and laser-propelled nanocraft interstellar flights (Breakthrough Starshot Initiatives) [26].

Photonics and lasers – fibre lasers in particular – will open up new possibilities in designing future production lines. Fibre delivery over long distances and beam agility is critical for implementing this goal. Smart photon pipes will be able to deliver photons around the factory in the same way copper cables currently deliver electrons (figure four). However, smart photons, in addition to power, will carry bits of information that will be used to control the manufacturing tool and the process alike. This information can be fully exploited using modern AI algorithms and deep learning to completely integrate machines and processes, for ultimate efficiency, speed and control in factories of the future [27].

References 

1. M. N. Zervas and C. A.  Codemard, High power fibre lasers: A review, IEEE J. Sel. Top. Quantum Electron. 20, 0901509 (2014).
2. M. O’Connor and B. Shiner, High power fibre lasers for industry and defence, in High-Power Laser Handbook, H. Injeyan, G.D. Goodno (eds), Mc Graw Hill (2011), chapter 18.
3. E. A. Shcherbakov, V. V. Fomin, A. A. Abramov, A. A. Ferin, D. V. Mochalov, V. P. Gapontsev, Industrial grade 100kW power CW fibre laser, Advanced Solid-State Lasers Congress Technical Digest, paper ATh4A.2 (2013).
4. https://news.lockheedmartin.com/2017-03-16-Lockheed-Martin-to-Deliver-Wo...
5. A. Wetzig, L. D. Scintilla, C. Goppold, R. Baumann, P. Herwig, A. Mahrle, A. Fürst, J. Hauptmann and E. Beyer, New Progress in Laser Cutting, Lasers in Eng., 35, 75 (2016).
6. F. Abt, A. Hess and F. Dausinger, Temporal behaviour of focal shift of beam forming optics for high power single mode lasers, ICALEO Congress Proc., 561 (2008).
7. T. Eidam, C.Wirth, C. Jauregui, F. Stutzki, F. Jansen, H.-J. Otto, O. Schmidt, T. Schreiber, J. Limpert, and A. Tunnermann, Experimental observations of the threshold-like onset of mode instabilities in high power fibre amplifiers, Opt. Express 19, 13218 (2011).
8. M. N. Zervas, Power scaling limits in high power fibre amplifiers due to transverse mode instability, thermal lensing and fibre mechanical reliability, Proc. SPIE 10512, 1051205 (2018).
9. P. Kah, J. Lu, J. Martikainen and R. Suoranta, Remote laser welding with high power fibre lasers, Engineering 5, 700 (2013).
10. NADEX Laser R&D Centre: https://www.industrial-lasers.com/articles/print/volume-30/issue-1/depar...
11. S. D. Jackson, Towards high-power mid-infrared emission from a fibre laser, Nature Photon. 6, 423 (2012).
12. J. He,  D. Lin, L. Xu, M. Beresna, S. Alam, M. N. Zervas and G. Brambilla, 5.6kW peak power, nanosecond pulses at 274nm from a frequency quadrupled Yb-doped fibre MOPA, Opt. Express, 26, 6554 (2018).
13. C. Goppold, T. Pinder, and P. Herwig, Experimental investigation of the linear polarization state of high power fusion cutting with 1μm laser radiation, J. of Laser Appl., 28 031501 (2016).
14. R. Weber, A. Michalowski, M. Abdou-Ahmed, V. Onuseit, V. Rominger, M. Kraus, T. Graf, Effects of Radial and Tangential Polarization in Laser Material Processing, Physics Procedia 12, 21 (2011).
15. G. Costa Rodrigues, V. Vorkov and J. R. Duflou, Optimal laser beam configurations for laser cutting of metal sheets, Procedia CIRP 74, 714 (2018).
16. C. A. Codemard, N. T. Vukovic , J. S. Chan, P. J. Almeida , J. R. Hayes, M. N. Petrovich , C. Baskiotis, A. Malinowski, and M. N. Zervas, Resonant SRS filtering fibre for high power fibre laser applications, IEEE J. of Select. Top. in Quantum Electr., 24, 0901509 (2018).
17. C. Röhrer, C. Codemard, G. Kleem, M. Abdou Ahmed, and T. Graf, Preservation of good beam quality over several hundred metres in highly multimode fibres, Advanced Photonics Congress, paper SoW4H.2 (2018).
18. T. D. Bradley, et al, Record low-loss 1.3dB/km data transmitting antiresonant hollow core fibre, ECOC 2018 post-deadline paper.
19. G. Palma-Vega, et al, High average power transmission through hollow-core fibres, ASSL ATh1A.7 (2018).
20. D. A. V. Kliner and B. Victor, A breakthrough for fibre lasers: tuneable beam quality enables optimised cutting of thin and thick metal, Laser Technik Journal 2 (2018).
21. P. Almeida, P. Gorman, J. S. Chan, N. Vukovic, C. A. Codemard, and M. N. Zervas, Pulsed fibre laser with spatial and temporal control, Proc. SPIE 10512, 1051215 (2018).
22. F. O. Olsen, K. S. Hansen, and J. S. Nielsen, Multibeam fibre laser cutting, J. Laser Appl. 21, 133 (2009).
23. J. J. J. Kaakkunen, P. Laakso, and V. Kujanpaa, Adaptive multibeam laser cutting of thin steel sheets with fibre laser using spatial light modulator, J. Laser Appl. 26, 032008 (2014). 
24. G. Mourou, B. Brocklesby, T. Tajima and J. Limpert, The future is fibre accelerators, Nature Photon. 7, 258 (2013).
25. J. Kästel, J. Speiser, Laser-based space debris removal: design guidelines for coherent coupling power transmission, Proc. SPIE 9990, 99900L (2016).
26. P. Lubin, A roadmap to interstellar flight, arXiv:1604.01356 (2016). 
27. Factories of the Future, European Commission, Directorate G - Industrial Technologies (2013).