In the last few years, high-power fiber lasers have been introduced to the marketplace and have gained a significant share of the industrial laser market. It is of great interest to compare the welding performance of these lasers to the older, more established type of lasers. This manuscript undertakes the comparison of welding performance of different laser types using a format first advanced in a comparative study of welding performance of different types of carbon dioxide lasers.1
This format uses the welding energy per unit length of weld and per unit thickness. This welding data is plotted against the thickness or depth of the weld. This method of plotting arose to investigate whether it took twice the energy to weld a piece of material twice as thick; that is, is the welding energy proportional to the thickness in the keyhole or deep penetration welding process? The results with the CO2laser welding showed this to not be the case. For very thin welds, the welding energy per unit thickness increases rapidly with decreasing thickness; these welds are usually performed at very high welding speeds and the keyhole nature of the welding process is lost. For thicker materials, the welding energy per unit thickness also increases. This was attributed to the multiple reflections that the beam undergoes in propagating down a long keyhole, and the curved nature of deep keyholes. The welding energy per unit length and per unit thickness undergoes a minimum for material thickness of about 4 mm.
In plotting the welding data for many sources, there was much scatter. However, it was possible to define a "minimum energy range" where a number of lasers achieved welding performance for a given thickness of material.* This minimum energy range, expressed in J/mm2 or kJ/cm2, increased for both thin materials (1 or 2 mm thick) and for materials thicker than 6 mm.
Welding performance of Nd:YAG lasers
Many believe that Nd:YAG lasers are more suited than CO2 lasers for laser welding operations because of their shorter wavelength. As a result of the shorter wavelength, the absorption of the laser radiation in the material is greater and the absorption of the laser radiation in a plasma that may form over the material surface is less. These characteristics have found the repetitively pulsed Nd:YAG laser many applications in welding of precision electronic parts.
The welding performance of high-power Nd:YAG lasers is compared to that of the CO2 laser in Figure 1. The laser labeled "Hobart" was designed by a company called Martek under US Government contract to have a very small focus so the laser would be useful for welding aluminum. Martek sold the rights to the laser to Hobart Laser Company, which a few years later sold the marketing rights to GSI Lumonics. The data presented here originated with Hobart Laser Company sales literature and publications in trade magazines.2 The output of this laser is delivered by a fiber optic to the workpiece.
The blue line in Figure 1 is the welding performance of a different design of continuously operating Nd:YAG laser developed by Lumonics Corporation.3 This laser was capable of operating not only continuously, but with an output modulated as sine wave or a square wave.
The shaded region in Figure 1 is the "minimum energy range" that is found with CO2 laser welding. For thicker materials, the energy per unit length per unit thickness becomes greater for the Nd:YAG results than that achieved with the CO2 laser.
Welding performance of fiber lasers
Welding performance of fiber lasers from several different institutions is shown in Figure 2. As before, the welding results are presented as energy per unit length per unit thickness. At Rostock,4 5-mm plates of high alloy steel were welded with a 6.9 kW fiber laser at 4 m/min. At The Welding Institute in the UK, welding trials were undertaken on C-Mn steel with a 7 kW fiber laser.5 Quintino et al. performed welding trials on X100 pipeline steel with an 8 kW fiber laser at a variety of power levels.6
With some of the welding trials, the welding energy per unit length per unit thickness remains approximately constant as the thickness is increased. The energy per unit length per unit thickness inevitably increases as the laser reaches the upper limit of its welding capability in terms of material thickness. For higher power lasers, this occurs at greater thicknesses.
Comparative welding performance
The welding performance of CO2 lasers, Nd:YAG lasers, and fiber lasers are compared in Figure 3. As in Figure 1, the shaded region is the minimum energy range achieved by CO2 lasers with good mode quality, from Reference 1. The data presented for the Nd:YAG and fiber lasers is selected from the lowest energy data from Figures 1 and 2.
Over part of the operating range, the fiber laser's performance is equivalent to that of the CO2 laser's, in terms of welding energy per unit length per unit thickness. For thicker materials however, the fiber laser requires a greater energy per unit length than does the better CO2 lasers.
The lowest energies per unit length per unit thickness were achieved with the Nd:YAG lasers. Specifically, these low specific energy values were achieved with a (now discontinued) laser that had an exceptionally good mode quality.7 The energy per unit thickness per unit length is comparable to that of the slow flow CO2 laser referred to above, which also had an exceptionally good mode. In early days of investigation of high-power laser applications, it was felt that the keyhole size of approximately 2 mm was determined by the thermal properties of the steel, and that, as long as the beam could be focused smaller than the keyhole, further improvements in focusability would have little benefit. Clearly, the focusability has a large effect on energy utilization.
The data for the shorter wavelength lasers presented in Figure 3 appear to have a different slope with respect to thickness than does the data from the CO2 laser. Is it possible that the shorter wavelength which results in a higher absorption at the metal surface also results in a high absorption when the beam hits the liquid metal at the leading edge of the keyhole? In this case, the output from the shorter wavelength radiations would have less tendency to be channeled or guided by the keyhole deep into the material.
1. Merchant, V.E., "A comparative analysis of laser welding performance," pg 53 of The Industrial Laser Annual Handbook, 1989 Edition, Ed. by David Belforte and Morris Levitt (PennWell Publishing Company, Tulsa, OK).
2. Webber, T., "Material processing with a CW Nd:YAG laser," Industrial Laser Review, Nov. 1992, pg 5-9.
3. Bransch, H.N., "Cutting and welding performance of high-power CW and pulsed Nd:YAG lasers," manuscript; "Welding with high-power pulsed and CW Nd:YAG lasers," Photonics Spectra, Sept. 1991, pg 107.
4. Jasnau, U., and Seyffarth, P., "Fiber lasers with high powers - after initial welding tests, a wide variety of advantages for the user are in prospect," Welding and cutting 3, No 5, pg 302 (2004).
5. Verhaeghe, G., and Hilton, P., "Deep penetration welding - the fibre laser way," from The Bulletin, The Technical Journal for Industrial members of TWI, Mar/Apr 2005, pg 12.
6. Quintino, L., Costa, A., Miranda, R., Yapp, D., Kumar, V., and Kong, C.J., "Welding with high power fiber lasers - a preliminary study," Materials and Design, 28, 1231-1237 (2007).
7. Jellison, J., Keicher, D., and Fuerschbach, P., "Performance of a three crystal 1800 Watt CW Nc:YAG laser," Proc. ICALEO 1990, pg 123 (Laser Institute of America, Orlando, 1990).
*The author has since learned that a "slow-flow" Carbon Dioxide laser has significantly better welding performance than that outlined by the minimum energy range discussed here. This laser has a stable resonator with exceptionally good beam quality.
Click here for an article by Stan Ream titled "Why so little laser welding?" Click here for an articel by Craig Marley titled, "Laser welding: Is your application report an effective sales tool?"