Fibre lasers defy economy

Share this on social media:

Stephen Mounsey finds the fibre laser industry holding steady in the face of the financial climate

Fibre lasers were introduced recently in comparison to legacy technologies such as CO2, but the technology has been adopted rapidly within key markets, particularly industrial materials processing and medical lasers. According to a report recently commissioned by the European Photonics Industry Consortium (EPIC), the market for fibre lasers has grown from $105m in 2005 to $300m in 2008, equating to a compound annual growth rate of 42 per cent per annum. Analysis shows that materials processing applications continue to be the most important market sector, accounting for 70 per cent of the 2008 market ($210m), which equates to seven per cent of the worldwide market for lasers in materials processing, estimated in the report to be worth $3bn in total.

Tom Pearsall, general secretary of EPIC, believes that there is room for improvement if growth is to continue: ‘If we look at the major application, which is materials processing, fibre lasers offer superb beam properties, but the continuous wave power coming out of a [single-mode] fibre has been stuck at 2kW for a couple of years. Increasing the power output of fibre lasers is necessary in order to make more serious in-roads into the market of CO2 lasers.’ CO2 remains the dominant laser within industry for marking and cutting of steel.

The advantages of fibre lasers are cited by proponents primarily in terms of two key points: cost-effectiveness and beam quality. The cost-effectiveness of fibre-based systems stems from the high wall-plug efficiency (approximately 32 per cent in some systems), lower watt-for-watt purchase cost, low maintenance requirements (the diode life is greater than 100,000 hours, whereas lamps in lamp-pumped systems may need replacing after 200-500 hours), zero warm-up time (meaning the laser does not need to be left powered in order to be ready to work), and ease of installation. The efficiency of fibre systems means that their cooling requirements are reduced, meaning that they can also be more compact and portable. Due to the laser being produced in an optical fibre, coupling to a delivery fibre can be done more efficiently than with other technologies. Additionally, fibre delivery of the 10.6μm radiation emitted by CO2 lasers is not possible, necessitating mirrors and optics.

Beam quality is important for precision in medical applications, and high-powered lasers with high-quality beams are being developed for use as directed energy weapons. Fibre lasers are able to produce a beam that is almost diffraction-limited in quality, meaning that they are the most ‘laser-like’ of all lasers, with very narrow divergence, able to remain coherent and focused even when projected over several kilometres. Furthermore, this beam quality is retained across the entire power range over which the laser is operated; this is not the case when scaling the power of other kinds of laser. Single mode lasers produce the highest beam quality but their power is limited; IPG has just demonstrated a 10kW single-mode fibre laser, but 2kW is more commonly the upper limit in industry. Power levels beyond this are achieved by combining beams, either in series or in parallel.

Mark Richards, product manager for fibre lasers at GSI Group, explains: ‘At lower powers, the beam quality from fibre lasers is very good – almost diffraction-limited. As we move to higher powers, industrial customers no longer have any need for such high beam qualities; the beam quality needs to be better than that achieved with lamp-pumped lasers, but not perfect. We can use techniques to combine the beams from several fibre laser modules together to create a very usable beam of higher power.’ Directed energy applications require beams to be combined coherently in order to produce a very high power, single-mode beam. Current targets are at around 100kW, maintaining beam size over two or three kilometres. In contrast, industrial applications require a beam source to deliver a suitably tightly controlled energy to a work piece, which may be remotely welded, remotely cut, sintered, melted, etc. In this case one can combine a number of beams together spatially, in parallel, and still have sufficiently high beam quality. The typical approach used to achieve powers up to 50kW is to take the output from a number of lasers and feed it into a combiner, which is typically a fused-fibre device. On the input side of the combiner there are a number of single-mode input beams, leading into a multi-modal output fibre, which catches the agglomerated power from the single-mode beams. The beam quality is no longer as high, however, resulting in a higher divergence, and usually requiring a larger core-size within the fibre. The fibre that delivers the energy to the work piece, called the process fibre, is typically 100-200μm in core diameter, compared to a single-mode input fibre with a core diameter of 10-12μm.

Increasing powers – diodes

Aside from combining several beams together, developers also try to push the power output of single-mode fibres. Richards describes the process as an ‘evolution of the original platform in an incremental sort of way.’ Of key importance are the diodes used to pump the fibres, and increasing the brightness of these components increases the power of the laser. Most manufacturers of fibre lasers have a say over the design of their diodes, with some carrying out in-house production.

GSI has received a grant from the UK Technology Strategy Board as part of a programme called Helpsys. The company is working alongside Herriot Watt and Cranfield Universities and Power Photonic, a designer of micro-optic components, in order to characterise and improve the brightness of various fibre-coupled laser diodes. It is hoped that the experience generated will be useful for direct diodes and for source diodes for use in fibre lasers.

Diode performance can be important in terms of application performance: Fibre lasers are modulated by means of modulating the pump semiconductor diodes. Single emitter diodes can be modulated quickly, and so the lasers can be modulated at rates of up to 100kHz. High-speed modulation allows pulse shaping techniques to be used in which, rather than simply turning a laser on and off, beam power is set to trail off in a pre-programmed way. Taking the example of laser welding, pulse shaping can be used to precisely control the cooling profile of the part, allowing a crack-free, low-porosity weld to be produced. In many applications, the way in which energy is applied to the work piece, in terms of both space and time, is important; fibre lasers allow a particularly fine control of this energy. While solid-state lasers may also be modulate in this way, slower response times limit the effectiveness of such techniques.

Ultrafast fibre lasers are set to have an impact on the ultrafast market, mainly on account of their having a fraction of the ownership and running costs of their solid-state counterparts. Ultrafast fibre lasers achieve peak powers of up to 10MW, with a mean power of only 1-10W (50W maximum). Anatoly Grudinin, chief executive of Fianium, calls this approach ‘a more intelligent use of power.’ Ultrafast fibre lasers propagated through a highly non-linear medium, such as a photonic crystal fibre, can produce wide-spectrum supercontinuums, useful for a range of novel applications. As well as being more affordable, fibre lasers have an advantage over solid-state lasers because of the efficient coupling between the input fibre and the photonic crystal fibre.

Increasing powers – overcoming limitations

Fibre lasers of the past were susceptible to the photodarkening phenomena, in which the opacity of the glass fibre increases at high laser powers, reducing the output power of the laser as a whole. Although several models have been proposed, photodarkening is thought to come about when photons interact within the fibre to produce short wavelength photons. Defects within the structure of the glass are created, and these defects, known as colour centres, absorb certain wavelengths of light strongly. The colour centres are usually created on the UV band, but they have a ‘tail’ extending in to the visible and IR wavelengths. Because the colour centres originate in the UV, extending into longer wavelengths from there, the visible red alignment lasers will show evidence of photodarkening before the infrared process beam will. In designing the glass composition, these effects need to be negated, so that not only the power output, but also the alignment functionality is retained over the lifespan of the equipment.

Steve Norman, chief technical officer at SPI lasers, discusses the effect: ‘The glass which is used in the fibres at higher powers today is very different in composition to that which we were using in the early days of our business. Photodarkening was present in lasers operating at lower powers, but it wasn’t an issue because we could put in additional pump power to compensate for photodarkening. The problem emerges as you go up in power level; the photodarkening increases in a non-linear way, meaning that [adding additional pump power] would be un-economic and inefficient. We had to take steps to have a non-photodarkening fibre.’

Each company producing fibre lasers will have its own composition of glass, in terms of dopants and host matrix, and through paying attention to these factors, photodarkening is no longer a limiting factor. Nonetheless, the effect is still monitored in order to ensure reliability of fibre lasers.

Norman described a stability and aging test, carried out by SPI, in which a laser is run at rated power, in ‘open loop’ mode (the pump diodes are driven at constant current, with no feedback adjustment) for 1,000 hours. The output of the laser is monitored, as well as the temperature of the water which cools the diodes. Any photodarkening will be seen early on, as fibre degradation reaches an asymptote. SPI demonstrated stable output power over the 1000-hour test to +/- 0.5 per cent maximum variation, and even this instability could be attributed to fluctuations in diode temperature. Norman describes the solution to the problem as ‘elegant and careful glass development.’

Why aren’t all lasers fibre lasers?

Given the long and wide-reaching list of advantages laid out here, it may seem that all lasers will one day be fibre lasers. SPI’s Norman believes: ‘It’s a case of “horses for courses”; if you view it from the users’ perspective, what they’re buying is a beam source. They don’t worry about the technology that’s inside it; it’s what they can do with that beam source, and how much they have to pay for it, that matters to them... there are some applications for which lamp-pumped lasers will continue to provide the best solution for many years, mainly due to the cost benefit.’ The pulse energies involved in applications such as drilling holes in turbine blades are generally very high – in the region of several joules rather than the millijoules produced by fibre lasers.

Bill Shiner from IPG, a company which has reported 22 per cent growth this year-to-date, is more enthusiastic about the remaining growth potential for fibre lasers: ‘It really, honestly is a revolution – and that’s why our company’s growing so fast!’