CO<sub>2</sub> the future

As fibre and disk lasers achieve ever higher powers at ever lower costs, Stephen Mounsey asks if there's still a place for CO2-based technology

When an industrial player is choosing any tool for an application, the equipment that will do the best job at the lowest cost is always going to win out. At lower power levels, used when cutting fine detail or welding in thin material, or when the use of an optical fibre delivery is appropriate (eg car production lines), the best option has traditionally been a solid state laser based on Nd:YAG crystals or something similar. At higher powers, the cost per watt favours carbon dioxide (CO2) laser.

CO2 lasers consist of a tube containing a mixture of gases (CO2, N2, H2 and/or Xe, and He), with a mirror at one end and a semi-reflective output coupler at the other end. The laser is pumped by an electrical discharge through the gas, and this can either be driven by DC, or by AC in the RF domain. The efficiency of a CO2 laser can be as high as 20 per cent.

Upsetting the status quo

This was the status quo until the mid-1990s, with lamp-pumped solid state lasers used for small-scale materials processing applications, and CO2 lasers used as a cost-effective high-power light source. When fibre lasers were introduced, they promised to revolutionise the market for materials processing lasers, because of their increased overall power efficiency (sometimes called wallplug efficiency), their high brightness (good beam parameter product), and the ease of delivering a laser beam through a fibre. The high brightness of the fibre laser means that it is easy to focus their energy to a sharp point. Fibre lasers are a subset of solid state lasers and, like YAG lasers, produce a beam with a wavelength of approximately 1μm (1,064nm).

Launched at about the same time as fibre lasers, the disk lasers developed by larger industrial integrators such as Trumpf and Rofin feature a similar blend of high-power performance and efficiency. Disk lasers can also be fibre-coupled for ease of beam delivery. Jens Bleher, managing director of Trumpf’s laser technology division, believes that disk lasers are the correct choice for multi kW applications requiring high beam quality, such as remote welding or deep, narrow welding. Disk lasers are less sensitive to reflection than fibre lasers, as the resonating medium is maintained at a power density far below its breakdown threshold. Bleher says that, for this reason, disk lasers can be more effective than fibre lasers for some applications.

The ongoing development of reliable high power laser diodes has also contributed to the erosion of CO2’s market share. Firstly, high-reliability diodes are vital components of disk lasers, fibre lasers, and other DPSS lasers as a pumping source. Additionally, as the brightness and beam quality obtainable from the diode itself continues to improve, direct diode lasers are beginning to be applied to some mid-power applications that were once the domain of CO2, such as plastics welding, braising, or heat treatments.

The niche at the thick end

In materials processing, cutting applications are those requiring a combination of the highest power and the highest beam quality. Fibre and disk lasers are both able to emit a lot of power from a very small aperture, and therefore have high brightness. High brightness may at first seem to make fibre lasers the right choice for cutting processes in general, and in many applications they have indeed taken a large portion of the CO2 market share, but there are still a large number of applications for which fibre lasers are not suitable for some fundamental reasons.

Dr John Powell is a senior member of the Association of Industrial Laser Users, and technical director at Laser Expertise, a laser cutting consultancy and job shop based in Nottingham, UK. He explains: ‘Most of the market for high-power laser processing is in cutting applications and a lot of that money is associated with cutting 3-15mm thick steels.’

A key parameter for materials processing is the absorption characteristics between the laser used and the substrate processed. Metals actually absorb the 1μm wavelength of fibre lasers more efficiently than the 10.6μm wavelength, and so cutting could be thought of as more efficient by using these wavelengths. There are, however, other effects at work besides the basic absorption properties.

The first of these effects is related to the Brewster angle. Also called the polarisation angle, this is the angle between an incident light beam and a given material’s surface at which a particular polarisation of the light is optimally absorbed by the surface, with minimal reflection. When cutting with a laser, we can think of this angle as being equivalent to a particular slope of the cut surface at which the maximum amount of the laser’s energy is sucked into the substrate material. This angle varies with the cutting wavelength and with the materials involved.

When cutting steel using the 10.6μm from a CO2 laser, the Brewster angle is near vertical at approximately 87°, but when using 1μm solid state light, the Brewster angle is further from vertical at 78°. At high cutting speeds, the slope of the cut surface will stabilise at these angles, meaning that eventually, when cutting thicker pieces of metal, the cut front can exceed the size of the incident beam.

The geometry of the cut front, the Brewster angle and the focused beam for 1μm lasers, means that the light bounces down the cut in a random and messy way, and when the thickness of the metal is greater than 2-3mm, this leads to an uneven cut with some dross. In contrast, the same geometrical considerations for the 10.6μm of the CO2 laser work in its favour, producing a smoother cut right up to 25 or even 30mm of steel.

The other consideration in favour of CO2 is related to the brightness of the alternatives. While brightness is useful for a quick cut in a thin sample, very high power intensities in a thicker sample mean that the material coming out of the cut will boil away rather than melt. As Dr Powell explains, this is an inefficient process: ‘If you wanted to move a load of wax from A to B, you could melt it, and this wouldn’t take much energy,’ he says. ‘If you decide to boil it off, however, that’s going to take you all day. You don’t need to produce a vapour to produce a cut; a slot is fine, and a slot can be produced just by pouring liquid out of it. It’s a much more effective method, as you end up with droplets of liquid that didn’t cost you a lot in energy, whereas turning the whole lot to a gas would cost you an awful lot of energy. It’s the same as if you’d produced your cut and then spent an extra load of energy turning the stuff that came out of the bottom of the cut into a vapour.’

For this reason, focusing a beam down to too small a point, even with a CO2 laser, is undesirable, because melting with minimal boiling is the most efficient way of removing metal. Even the beams that come out of CO2 lasers (which have an inferior focusing quality to those of fibre lasers) are sometimes defocused in order to get the correct cutting properties: ‘You wouldn’t want to focus a CO2 beam down to too small a point,’ says Dr Powell, ‘because you need a certain width of beam in order to get enough gas going through the cut to be able to blow the liquid out. Also, you really don’t want to boil anything.’ These boiling considerations are particularly relevant to cutting of thicker metals, where the beam has to travel slowly, and so high power CO2 setups are set to remain the industry standard for thicker metals for some time to come.

Image courtesy of Trumpf

This is perhaps best summed up in the words of a recent statement from Trumpf: ‘CO2 lasers can be used across the board without any loss of edge quality, even when it comes to sheets of 25 or 30mm thickness. The thicker the sheet, the faster and the better the results produced by these lasers in comparison to the solid state alternatives.’

Cutting out a smaller market

Lower power applications from 10-500W remain a high growth area for CO2 lasers, thanks in part to the low cost per Watt of these systems compared to their solid state alternatives and thanks to the versatility of the 10.6μm CO2 wavelength. 10.6μm is readily absorbed by non-metals such as glass, ceramics, polymers, and organics such as wood or paper.

Gary Broadhead is sales director for industrial lasers at Laser Lines, a UK company that distributes sealed CO2 lasers from US manufacturer Synrad. ‘Our involvement in the CO2 laser world is at the low to medium power range of 10-500W,’ he says. ‘Within that range we’ve seen very little impact from fibre lasers, and I think that’s going to continue for a long time. I’m very optimistic about the future of the sealed CO2 laser,’ he says.

Larger kilowatt and multikilowatt CO2 lasers circulate the gas medium rapidly through coolers in order to remove excess heat. In contrast, sealed CO2 lasers remove excess heat by diffusion air cooling, or by water cooling for powers over 100W or where greater thermal stability is required. The technology is well established, but recent advances, such as those made by Synrad to the RF drive electronics, have meant that the lasers are becoming smaller and more efficient. In addition, the technology has been made cheaper by low cost components such as high-power transistors.

Smaller sealed CO2 lasers are sold to OEMs for incorporation into the plotter style cutting and engraving systems. These are low-cost systems, often used by so-called ‘mom and pop’ enterprises, for cutting non-metals such as wood or polymers – materials that cannot be cut by solid state, 1μm radiation. The other major market for sealed CO2 is in highspeed marking systems for packaging, such as burning the best-before-date on to a cereal box; these are cleaner, safer, and cheaper than legacy inkjet systems.

While solid state alternatives to low-power sealed CO2 lasers are available, the costs are currently high. Broadhead believes that a fibre or a disk system would not be suitable for applications under 200W due to the capital cost of the system, but he also believes that fibre may offer a cost-effective alternative at powers above 200W due to the technology’s high efficiency and associated low running costs.

Image courtesy of Trumpf

As well as advantages in terms of compactness, sealed CO2 lasers can have better performance characteristics than their equivalent flowing-gas cousins. Coherent offers its Diamond E-series sealed CO2 laser emitting 1kW of CW radiation at 10.6μm, which the company claims offers the same brightness as its 2kW flowing gas laser, but with the added advantage of lower initial capital costs, higher reliability, and higher efficiency. Stewart Woods, materials processing market manager at Coherent, states that, based on his own experience, ‘in the 2-4mm thickness range, material absorption is no longer the primary factor in cutting speed, and fibre lasers and sealed CO2 at the same power offer essentially the same processing speeds’. He believes that sealed CO2 is, however, more attractive due to recent advances in the technology that mean they can be mounted directly onto a robot arm, eliminating the need for fibre delivery. ‘Sealed CO2 lasers are an increasingly attractive alternative to fibre lasers for more cost sensitive uses,’ he concludes, ‘especially for situations where both metal and organics must be processed.’

So what will it be?

Taking cutting as the chief application specific to high-power lasers, we can classify different types of job by the thickness of the materials involved. For materials under 2mm thick, fibre lasers deliver high cutting speeds and good cut quality for metals, and their high beam quality may make distance cutting a possibility. Where non-metals are also to be processed, however, or where the cost of a fibre system is high, sealed CO2 lasers are emerging as an increasingly attractive option. At the 2-4mm range, fibre lasers generally continue to deliver high speed and cut quality. CO2 is still a more cost-effective alternative, and is particularly well suited to non-metals. Fibre and solid state have a strong hold on the market here, but more recent advances in sealed CO2 have clawed back some ground.

For applications between 4mm and 6mm thickness, fibre (or disk) lasers are up against flowing gas CO2 lasers. Fibre offers lower operating costs and maintenance downtime, but at an increased initial cost. Fibre or other solid state systems win out where fibre delivery is an advantage. CO2 lasers offer better cut quality in both metals and non-metals, and a lower initial purchase cost, but lower efficiency, increased maintenance downtime, and complicated beam delivery. Although expensive, fibre and disk lasers have taken market share away from CO2 at these thicknesses, but the wavelength-related limitations mean that for situations where the thickness or material of the substrate may vary, they are less suitable.

When cutting anything above 6mm in thickness, flowing gas CO2 lasers remain the most attractive option. While both CO2 and fibre/disk lasers are capable of the high raw powers necessary to cut these thicknesses, the wavelength of CO2 creates a cleaner cut. Additionally, the brightness of the solid state alternatives is no longer an advantage, because boiling must be avoided and because the cut width must be wide enough to blow molten metal out with gas.

So the future of CO2 seems certain for the time being, a view echoed by Trumpf’s senior management: ‘We are not currently of the opinion that a “disruptive technology” will appear in the near future,’ says Jens Bleher. ‘A broad range of different laser technologies will continue to be used; The wide variety of applications shows that there is no one optimal laser in material processing.’