Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Cutting Stainless Steel with Disc and CO2 Lasers

Paul Hilton 

TWI Ltd, Granta Park Great Abington, Cambridge, CB21 6AL, UK

Paper presented at Proceedings of LAMP2009 - the 5th International Congress on Laser Advanced Materials Processing

There is much current interest in the capabilities of 1μm wavelength laser sources, such as disc and fibre lasers, for cutting metal. Due to the many inherent variables in the laser cutting process, it is very difficult to directly compare results obtained using fibre delivered laser beams, with those obtained using the CO2 laser as the cutting source. This paper will describe a series of experiments which go someway towards comparing the results of cutting stainless steel plate, from 0.6 to 6mm in thickness, using an Yb:YAG disc laser and a fast axial flow CO2 laser, both operating at 5kW of power and using nitrogen assist gas. Using both lasers, cutting speeds were established for each material thickness which produced the best cut quality in terms of surface roughness. In addition, a second set of results were established using the CO2 laser cutting system, at the speeds which produced the ‘best’ cuts using the disc laser. These results were then compared, in terms of squareness of cut and edge roughness, and also by photographing the cut edges. The photography was undertaken all at the same time and under identical lighting conditions, to assist a visual comparison. Analysis of the results showed that the disc laser was capable of cutting thin materials (0.6 and 1.2mm thickness) at higher speeds and with lower edge roughness than the equivalent power CO2 laser. However, over the parameter range investigated, for 3 and 6mm thickness material, the CO2 laser produced better cut quality, particularly in terms of surface roughness.

1. Introduction

Cutting is an important application for materials processing lasers, dominated by the use of CO2 gas lasers for the cutting of steels. Worldwide sales of CO2 lasers were expected to pass $1,000 million in 2008 (Kincade and Anderson, 2008), with the majority of this market being lasers used for cutting flat plate. Of this market, the largest application is in the cutting of low alloy steels. Although CO2 lasers are well suited to producing high speed and high quality cuts in flat plate, the moving mirror beam manipulation systems used with CO2 lasers become more complicated for cutting 3D shapes. CO2 laser light has a wavelength of 10.6μm, which is too long to be transmitted by glass optics or by optical fibres, so gold plated mirrors and lenses made of potassium chloride or zinc selenide, are used for beam manipulation and focusing. Lasers whose output wavelength is in the region of 1μm have the advantage that their beams can be transmitted easily and efficiently down optical fibres. One advantage of this is that it removes the need for the moving mirror manipulation systems necessary with CO2 laser cutting systems. Two recent developments in fibre delivered laser technology, the disc laser and the Yb:fibre laser, have overcome the problems of beam quality and efficiency apparent on the earlier Nd:YAG solid state lasers. Disc lasers use a solid disc-shaped crystal of yttrium aluminium garnet doped with ytterbium as the lasing medium, and this is excited using diode lasers. This produces laser light with a wavelength of 1.03μm, which can be transmitted via an optical fibre. Disc lasers have very good beam quality and higher electrical efficiency (10-20%) than Nd:YAG lasers. The Yb:fibre laser is a solid state laser in which an optical fibre doped with a low level of ytterbium is the lasing medium. By combining a series of low power single-mode lasers, this technology enables scaling of laser power from several hundred watts to 30kW (currently in use) and potentially up to 50kW and above. The wavelength produced is 1.07mm, which again can be transmitted via optical fibre. Electrical efficiency is 20-30%, and beam quality is very good. At the laser powers employed for laser cutting, these disc and fibre lasers have such good beam quality that they can be focused to produce spots as small as those generated with CO2 lasers.

The results of early work carried out with fibre lasers investigating their potential for cutting, showed significant differences in the cut quality obtained, when compared with that available using CO2 lasers. For example, the surface appearance of cut edges made on stainless steel with inert assist gas using the fibre laser was not as regular as that on samples made with a CO2 laser (Hilton and Chong, 2006; Sparkes et al 2006). However, there was some evidence that on thicknesses less than 1mm, a faster speed could be obtained using the fibre laser. Since that time other workers (Himmer et al 2007; Wandera et al 2007), using fibre lasers, have found essentially the same, both for oxygen and inert gas assisted cutting. Since that time other works (Petring et al 2008) have tried to compare laser cutting processing terms of beam brightness.

The work reported here concentrated on a comparison of the cutting capability of a high beam quality disc laser and that of a modern, commercially available, high quality CO2 laser cutting system, for cutting stainless steel (AISI 304 grade) from 0.6 to 6.0mm in thickness, using high pressure inert gas. Laser cut stainless steel is used in many industry sectors in applications such as the manufacture of food processing equipment, medical products, cryogenic equipment, white goods, architectural structures and works of art. The thickness range was chosen to provide the best comparison data, given that there was already some evidence for different performance on thin and thick materials using the 1μm laser source. Inert assist gas cutting was chosen, as opposed to oxygen assist gas cutting, so as to not involve the additional exothermic energy available from the oxygen and its effects on the cut quality.

2. Experimental Programme

Trials were carried out using two laser systems. The CO2 laser was a 5kW Trumpf TruFlow 5000. The disc laser was a Trumpf TruDisc 5001. A fixed laser power of 5kW and nitrogen assist gas were used throughout the trials. In the case of the disc laser, which was used in a laboratory environment, sections were cut from 50mm wide strips of material. Trials were carried out on each of the four material thicknesses, using a Precitec YK52 cutting head. Cuts were made varying the laser focus position and assist gas pressure, to further refine parameters and examine the process tolerances. For each condition, the travel speed was varied in steps above and below the speed which appeared to be the optimum for the particular thickness being cut, based on a visual examination of the cut edges during the experiments. The cutting nozzle diameter and stand-off were kept constant for each material thickness. The cutting parameters giving the ‘best’ results for each material thickness are shown in Table 1. The CO2 laser used for this work was part of a modern, commercially available integrated 2D flat bed cutting system. This production machine had sets of cutting parameters pre-programmed for several material and thickness combinations, so these were used for all thicknesses of stainless steel cut.

Table 1 Cutting parameters obtained with the disc laser for producing the ‘best’ cuts

Material thickness, mm 0.6 1.2 3 6
Cutting speed, m/min 24 6 3.6 3
Nozzle diameter, mm 1.5 1.5 1.5 2
Focus position, mm -0.3 0 -1.5 -4
N2 assist gas pressure, bar 13 13 17 18
Nozzle stand-off, mm 0.7 0.7 0.7 1
Lens focal length, mm 150 150 150 150

The CO2 laser parameters for each material thickness are shown in Table 2. In addition, trials were carried out, for each material, at the speed previously determined as optimum for the disc laser trials.

Table 2 Cutting parameters pre-programmed on the CO2 laser cutting machine

Material thickness, mm 0.6 1.2 3 6
Cutting speed, m/min 10.6 8 4.2 2.7
Nozzle diameter, mm 1.7 1.4 1.4 2.3
Focus position, mm +4.7 +4.7 +2.2 -2.3
N2 assist gas pressure, bar 12 12 17 16
Beam diameter at lens, mm 24 24 20 16
Nozzle stand-off, mm 0.7 0.7 0.7 0.7
Lens focal length, mm 250 250 250 250

From the available technical data of the two laser beams it has been possible to estimate the minimum spot size and the Rayleigh length (here defined as R). For the disc laser the estimated minimum spot size was 0.1mm diameter with a corresponding R of 6mm. For the CO2 laser an estimated minimum spot size was 0.2mm with a corresponding R of 3mm.

Each set of samples was assessed in terms of edge quality by measurement of Rz and squareness, in line with the European standard EN ISO 9013 (2002). Note that this now requires only a single measurement half way down the cut edge. In addition, a visual assessment of the cut quality was also made. This was done by photographing all the cut edges of interest. In some cases, using the visual method, the ‘best’ cut was not necessarily the cut with the lowest recorded roughness. The ‘best’ cut in this visual analysis was defined as that being the one with the best combination of quality and speed. It should be pointed out that in the commercial laser cutting world, hardly any reference is ever made to this European (or other) standard and most cut quality is judged by visual inspection. In much published work on laser cutting, pictorial edge quality comparisons are difficult for a reader to interpret, as the photography (particularly the lighting) of the samples is never the same and can result in very different apparent cut quality. Because of this, in the work reported here, all the samples produced were photographed at the same time, under identical lighting conditions, so that visual edge quality comparisons can be made with a degree of confidence.

3. Results

Figure 1 shows a comparison between the measured Rz value and the cutting speed for various thicknesses of material at the cutting conditions giving the best visual results as defined above. Figure 1 shows that, in this analysis, Rz remained relatively constant with changing material thickness when cutting with the CO2 laser, whereas a large difference in Rz with thickness was observed with the disc laser, particularly at the thicker sections. In all cases other than for the 0.6mm thick material, the ‘best’ results with the CO2 laser were recorded at cutting speeds slightly lower than the speeds selected automatically by the machine control unit. The results ringed by a circle in Figure 1, indicate that for these samples, both roughness and squareness values were within ‘Range 1’ (the best) of the ISO standard. It should also be pointed out that for the two thinnest materials, the cutting speed with the disc laser could probably have been higher but was limited by the speeds available on the cutting table used.

Figure 1 Rz vs cutting speed for the ‘best’ results for both lasers and all thicknesses
Figure 1 Rz vs cutting speed for the 'best' results for both lasers and all thicknesses

These results of roughness were measured (as per the standard) at only a single position, half way down the cut. It is useful to compare the graphical analysis of cut quality with the visual one. Figures 2-5, show, in each material thickness, the edge quality of the ‘best’ disc laser cut, and compares this to the ‘best’ edge quality obtained using the CO2 laser. In addition, the performance of the CO2 laser, at the same speed as the disc laser best cut is also presented. In these images, the CO2 laser cuts appear better than the disc laser cuts, certainly for all thicknesses above 1.2mm.

Figure 2. Edge sections for 0.6mm material
Figure 2. Edge sections for 0.6mm material

a) Best disc laser cut at 24m/min. Rz = 5μm; 

b) Best CO2 laser cut at 11m/min. Rz = 9μm; 

c) CO2 laser cut at 24m/min. Rz = 49μm.

Figure 3. Edge sections for 1.2mm material
Figure 3. Edge sections for 1.2mm material

a) Best disc laser cut at 6m/min. Rz = 10μm; 

b) Best CO2 laser cut at 4.5m/min. Rz = 9μm; 

c) CO2 laser cut at 6m/min. Rz = 11μm. 

Figure 4 Edge sections for 3mm material
Figure 4 Edge sections for 3mm material

a) Best disc laser cut at 3.6m/min. Rz = 18μm; 

b) Best CO2 laser cut at 2.8m/min. Rz = 11μm; 

c) CO2 laser cut at 3.6m/min. Rz = 11μm. 

Figure 5 Edge sections for 6mm material
Figure 5 Edge sections for 6mm material

a) Best disc laser cut at 3.0m/min. Rz = 35μm; 

b) Best CO2 laser cut at 1.8m/min. Rz = 10μm; 

c) CO2 laser cut at 3.0m/min. Rz = 14μm.

The results for different focal positions taken with the disk laser suggest that the cutting process was very tolerant to changes in focus position.

4. Discussion

Himmer et al (2007) and Wandera et al (2007), using both fibre and disc lasers, have also produced results that agree with the findings produced here, but in these two papers the comparison with CO2 laser cutting suffers from the illumination problem mentioned earlier and the laser powers are different between the laser sources used. However, Himmer et al (2007) also found that stainless steel up to 20mm thickness could be cut using a 4kW fibre laser with beam quality of 2.5mm.mrad. Other than the laser and beam quality, the main differences between that work and the work reported here were the assist gas pressure and nozzle diameter, up to 22bar and 3mm being used. These results are partly in contradiction with the results reported by Sparkes et al (2006), who found that higher gas pressures but smaller nozzle diameters were beneficial for cut quality (using a 2.2kW fibre laser with a beam quality of 2mm.mrad). Theories based on modelling of material and gas in the kerf have been reported (Sparkes et al 2006, and Olsen 2007), which go some way to explaining the behaviour of materials when cut with near infra-red wavelength lasers, but do not fully explain the difference in cut quality observed between disc or fibre and CO2 lasers. Seefeld and O’Neill (2008) suggest that this may be due to the higher absorption of shorter wavelength light by materials, as in thicker sections, short wavelength light is largely absorbed close to the material surface, whereas reflections may distribute longer wavelength light deeper into the material.

Interpretation of laser cutting results, in terms of measured surface roughness and photographic image quality, both present some difficulties. It must be remembered that the Euronorm standard in this area, calls for only a single measurement of roughness in the centre of the cut. Photographic images of the same cut can differ a lot if illuminated from different angles. In this work the latter problem was reduced by photographing all the samples at the same time under identical lighting. However, Figure 3 illustrates the problem. When looking at these cut edges it is clear that both the CO2 laser cuts appear to the eye, less rough than the disc laser cut, but according to the measurements, the roughness on all these samples is 10 +/-1 μm.

Notwithstanding this difficulty in interpretation of cut quality, the work has shown that the disc laser is capable of cutting thin materials at significantly higher speeds than the CO2 laser and with better (or at least as good) surface roughness. A transition point arises, based on these results, at a thickness of about 3mm. At a thickness of 6mm the CO2 laser produces better cut quality, in terms of both edge roughness and squareness. Despite some variance in the squareness in the thicker sections, the cuts produced by both the disc and CO2 lasers generally produced sharp corners with very little rounding. With the higher absorption of 1μm laser light by metals it has been suggested that this effect could cause rounding of the top of a laser cut edge. However, there was little evidence for this phenomenon in these results, apart from on the 6mm edges.

One topic of great interest in laser cutting, but which has never been fully explained, is the striation patterns found on the cut surfaces and their methods of formation. In Figure 5, for the 6mm thickness material and the CO2 laser, the striations are fairly vertical, with only slight rounding becoming evident at the base of the cut, for the fastest cutting speed. In addition, although there is some evidence of a layering phenomenon, it is not particularly evident. These points are considered consistent with a good cutting quality. The edge shown for the disc laser in this figure, however, presents a different picture. There is more evidence for the striations becoming less vertical and more rounded at the base of the cut but the major difference is the formation of three quite distinct layers, which are evident in the edge shown (and indeed in many of the disc laser cuts at 3 and 6mm thickness). In the case shown, there is a section about 1mm deep at the top of the cut, where at the magnification shown, there appears to be no striations at all. Below this is a section about 2mm deep, where the striations abruptly become visible. Below this is a third layer, this time about 3mm thickness, again well defined, but with a different striation pattern. What is interesting is the fact that the top section of this third layer appears smoother than the second layer. This structure might be due to factors associated with the laser wavelength or the beam/gas delivery. Clearly some very interesting process is in play here that requires more work to fully understand.

In discussing the results it should be remembered that the focus position of the beam on the sample surface changed, depending on the material thickness. For the disc laser with its fibre beam delivery system, the minimum focused spot will be an image of the fibre end and will therefore have a top hat distribution in its energy profile, in contrast to the more Gaussian type of profile found for the focused CO2 laser beam. However, it is known that beam profile for the disc laser moves to something much more Gaussian in shape only a few mm away from the smallest waist position. It is interesting to note that when using the disc laser, the optimum focus position, in this work, was very close to 0, ie on the material surface, for the two thinnest materials, whereas using the CO2 laser beam, the (machine) chosen focus position was some 5mm above the material surface. As the material thickness increased, for both systems, the optimum focus position moved downwards and by 6mm thickness, the optimum position was below the material surface for both lasers. As a general comment, the Rayleigh lengths of the beams used were of the order 6mm, ie close to the maximum thickness of the material used. It is also worth remembering that the CO2 laser was operated with a minimum spot size at least twice that of the disc laser.

CO2 laser cuts in stainless steels are characterised by a bright surface appearance and parallel and regular striations. The laser cuts made with the disc laser appear to show a more uneven surface, (but not necessarily less rough when measured in the way reported here), which gives them a duller appearance. Although these cut edges may be fit-for-purpose, there is now an expectation regarding the quality of a stainless steel laser cut edge, based not only on the squareness and roughness but also the cosmetic appearance that derives from the experience of existing CO2 laser cut quality.

5. Conclusions and Recommendations

CO2 laser cutting is a well accepted industrial process. It must be remembered that in the cutting process there are a multitude of process variables, only some of which have been used in this comparison. However, this work has tried to produce a meaningful comparison of the two sources, by providing both measurements of surface roughness which can be directly compared and visual comparisons of the cut surfaces at ‘best’ as well as the same speeds, for the two sources. The visual comparison is made better by the fact that all the samples were photographed at the same time under the same lighting. It might be expected that similar results to those reported here for the disc laser would be obtained if a similar beam quality fibre laser had been used.

Laser cutting trials have been carried out on four thicknesses of stainless steel, using a disc and a CO2 laser at the same laser power. The following conclusions can be drawn from this work:

  • Analysis of the cuts demonstrated that the disc laser was capable of cutting thin materials (0.6 and 1.2mm thickness) at higher speeds and with lower edge roughness than the equivalent power CO2 laser.
  • For the thin materials, the roughness measurements for the disc laser were almost constant over the full range of speeds investigated.
  • The CO2 laser produced better cut quality in the thicker material (3 and 6mm thickness).
  • Cuts were produced successfully in all material thicknesses, on both the lasers investigated, which conform to quality ranges 1 (highest quality) and 2 according to EN ISO 9013.
  • In the 0.6 and 1.2mm thickness material, where the disc laser bettered the CO2 laser, there was evidence that the cutting speed with the disc laser might be even higher than reported here, for the same roughness.

The potential benefits offered by fibre and disc lasers over CO2 lasers, in terms of size, reliability, running cost and lack of consumables, mean that these laser sources should be looked at seriously as alternatives to the well developed CO2 laser sources that have now been used for laser cutting for over 40 years. Many of the CO2 laser machine tool cutting suppliers are investigating these new technologies. The results reported here, for cutting thin sections, are already being applied in industry. More caution should be used when considering the capability of disc and fibre lasers for the cutting of thicker sections, however, and clearly more work is needed in this area to fully understand the capabilities of the new technology. It must be remembered that these technologies are very recent and were only applied to laser cutting for the first time in 2005, whereas CO2 laser cutting has been in continuous development since its invention at TWI, in 1967 (Sullivan and Houldcroft, 1967).

6. Acknowledgements

This work was funded by Industrial Members of TWI, as part of its Core Research Programme. The authors would like to thank Trumpf GmbH and Cambridge University Centre for Industrial Photonics, for use of the lasers in this work.

7. References

  1. BS EN ISO 9013, 2002: ‘Thermal cutting - classification of thermal cuts - geometrical product specification and quality tolerances’.
  2. Hilton P and Chong P, 2006: ‘A cutting edge solution using fibre lasers’, TWI Bulletin Nov/Dec.
  3. li>
  4. Himmer T, Pinder T, Morgenthal L and Beyer E, 2007: ‘High brightness lasers in cutting applications’. In ICALEO, Applications of Lasers and Electro-optics. Proc of the 26th Int Congress, Orlando, Florida, 29 October-1 November 2007. Publ. Laser Institute of America.
  5. Kincade K and Anderson S G, 2008: ‘Review and forecast, part 1: nondiode lasers’. Optoelectronics Report, Vol.15, No.1.
  6. Olsen F, 2007: ‘Will the new laser sources outperform the CO2 laser in metal cutting?’ In NOLAMP 11, Laser Materials Processing. Proc. of the 11th Nordic Conf, Lappeenranta, Finland 20-22, August. Publ. Lappeenranta University of Technology.
  7. Petring D, Schneider F, Wolf N and Nazery V, 2008: ‘The relevance of brightness for high power laser cutting and welding’. Proc. ICALEO (Paper 206) p95.
  8. Seefeld T and O’Neill B, 2008: ‘Cutting and welding with the new high brightness lasers’. The Laser User, Issue 50.
  9. Sparkes M, Gross M, Celotto S, Zhang T and O’Neill W, 2006: ‘Inert cutting of medium section stainless steel using a 2.2kW high brightness fibre laser’. In ICALEO 2006, Applications of Lasers and Electro-optics. Proc of the 25th Int Congress, Scottsdale, Arizona, 30 October-2 November. Publ. Laser Institute of America.
  10. Sullivan A B J and Houldcroft P T, 1967: ‘Gas-jet laser cutting’, British Welding Journal, August.
  11. Wandera C, Salminen A, Olsen F O and Kujanpää V, 2007: ‘Cutting of stainless steel with fiber and disc laser’. In NOLAMP 11, Laser Materials Processing. Proc of the 11th Nordic Conf, Lappeenranta, Finland 20-22, August. Publ. Lappeenranta University of Technology.

For more information please email: