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

Subscribe >
Skip to content

An assessment of the welding performance of high-brightness lasers and a comparison with in-vacuum electron beams (October 2008)

   
Geert Verhaeghe and Bruce Dance

Paper presented at ICALEO 2008, 27th International Congress on Applications of Lasers & Electro-Optics, Pechanga Resort & Casino Temecula, CA, USA. 20-23 October 2008. Paper no. 710.

Abstract

A unique experiment carried out at TWI a few years ago compared the welding performance of a number of fibre-delivered lasers with beam qualities ranging between 4 and 23mm.mrad, under identical processing conditions, and demonstrated the effect of laser beam brightness, together with beam quality and spot diameter, on the welding performance on both steel and aluminium. This paper describes a continuation of this earlier study, carrying out welding under identical processing conditions and comparing, against these initial results, the welding performances of laser beams with even higher beam qualities (higher brightness) and of an in-vacuum electron beam. This investigation demonstrates that the welding performance of a high-brightness laser set-up is highly dependent on the conditions of the metal vapour column forming between the processing point and the focusing lens. The effective removal of this metal vapour column, which scatters and/or absorbs some of the incident laser power, is essential in maximising the welding performance that is achievable with high-brightness lasers. By using an argon side-jet shielding and a series of argon cross-jets along the beam path between the focusing lens and the processing point, it was shown that the welding performance of high-brightness lasers could be improved considerably, and matching, or possibly surpassing, that of an equivalent beam quality in-vacuum electron beam used under similar conditions of power, spot size and welding speed.

Introduction

In the past two to three years, considerable advances have been made in high-power, fibre-delivered solid-state laser technology. This has resulted in both disc and fibre lasers, capable of several kilowatts of output power, delivered through optical fibres as small as 50µm in diameter. These new lasers possess a very good beam quality, such that their 'brightness', defined as the power density in the spot per solid angle in the cone of the focused beam, is exceptionally high. From a practical point of view, a better beam quality, i.e. low beam parameter product (BPP), means that, for a given size of processing head, a smaller focal spot can be produced, resulting in a higher power density for a given laser output power. When laser welding using a given laser power, material grade and thickness, it is generally accepted that a smaller focal spot produces better performance in terms of depth of penetration or welding speed, than a large focal spot. Additionally, a better (laser) beam quality allows a larger stand-off distance to be used, for a given focal spot diameter (and power density), which reduces the risk of damage to the processing optics during processing.

To determine the practical impact of beam quality from a processing point of view, TWI initiated a carefully controlled experiment in 2005, measuring the welding performance, in terms of depth of penetration and welding speed, of a range of fibre-delivered, solid-state laser beams with focal spot diameters from 0.6 to 0.14mm, beam quality levels from 23 to 4mm.mrad and beam brightness levels from 7.5 to 239x105W/mm2.sterad, at a fixed power of 4kW.[1] The lasers used included a lamp-pumped Nd:YAG laser (23mm.mrad), a YLR-7000 Yb-fibre laser (18mm.mrad), a Yb:YAG disc laser (7mm.mrad) and a YLR-5000 Yb-fibre laser (4mm.mrad). The results of these experiments showed that, for a given laser power of 4kW and focal spot diameter of 0.4mm, the welding performance on both steel and aluminium could be improved by welding with a higher beam quality laser. The results also suggested that an optimum laser beam brightness existed, producing a maximum welding performance in terms of depth penetration and speed. This concept of optimum brightness was used to explain why the laser with the highest beam quality did not produce the best performance in terms of depth of penetration.

However, it was clear from these pioneering trials, that further work was needed to assess the effect of beam brightness on welding performance. For instance, one observation that remained unexplained at the time occurred when welding steel with the highest beam quality laser (available at that time). When welding at a laser power of 4kW and a 0.4mm diameter focal spot, the depth of penetration was irregular and considerably lower than was expected. In fact, the depth of penetration was even lower than that achieved for a conventional lamp-pumped Nd:YAG laser, used at the same power and focal spot size. However, by applying simple inert side-shielding, normally not required when Nd:YAG laser welding steel, this was remedied. [1]

To further understand the underlying reason(s) of this irregular penetration and the effect of beam brightness on welding performance (depth of penetration and welding speed), additional trials were carried out with even brighter beams, using both (Yb-fibre) laser and electron beam. This paper details the results and findings of these trials.

Experimental trials

Material, equipment and set-up

The welding trials reported in this paper were performed on S275JR grade (EN ISO 10025) C-Mn steel, selected because of its common use as a structural material for a variety of industrial applications, including railroad cars, marine components, tanks and drilling rigs. The material, supplied in a thickness of 5 and 10mm, was cold-band sawn into samples 300mm long and 75mm wide, and machined to give a tapered profile, such that the thickness along the 300mm length varied continuously from 1 to 5mm, for the 5mm thickness samples, and from 4 to 10mm, for the 10mm thickness material. No shielding was applied during welding, unless stated otherwise.

The trials were carried out using a 4kW Yb-fibre laser, further referred to as YLR-4000, at the laser facilities of IWS in Dresden (Germany). The YLR-4000 had a nominal BPP of 2mm.mrad, as measured by IWS, with the laser power delivered via a 50µm diameter fibre optic into a 120/500 output housing, producing a nominal focal spot diameter of 200µm. In this configuration, the laser produced a beam brightness of 1372x105W/mm2.sterad. A Kuka 6-axis articulated robot arm was used to move the laser beam along the workpiece (Figure 1).

Fig.1. Set-up used for the trials with a YRL-4000 Yb-fibre laser at the IWS facilities (Dresden, Germany)
Fig.1. Set-up used for the trials with a YRL-4000 Yb-fibre laser at the IWS facilities (Dresden, Germany)

 

For the electron beam (EB) trials reported in this paper, the TWI 'HS2' EB machine was used. This EB machine is a modified Hawker-Siddeley 6kW, 150kV-rated, ‘high-vacuum' machine, equipped with an improved TWI-designed triode electron gun. This machine/gun combination was chosen because of its capability to produce a high quality beam with up to double the beam brightness achievable with 'conventional' EB machine/gun combinations. The system was configured to produce a beam diameter of 4mm at the lens and a distance from the lens pole piece gap to workpiece of approximately 400mm, resulting in a beam quality of 4.8mm.mrad and a nominal focal spot size of 0.37mm. The nominal beam diameter at focus contains 95% of the beam current, as measured using TWI's beam probing equipment, shown in Figure 2. This equipment, commercially available from CVE Ltd and capable of measuring EB power beams up to 100kW, allows rapid and accurate determination of the beam's power density distribution, yielding 2-D, 3-D or 4-D data sets, which describe the beam properties. With the electron machine/gun used, the average beam brightness produced within the nominal beam diameter was 2130x105W/mm2.sterad. During welding, the beam was traversed over the stationary workpiece, using electro-magnetic beam deflection, because mechanical manipulation at these speeds and accelerations was not available.

Fig.2. EB probe system developed at TWI, with a 2D data sample and typical compiled data set
Fig.2. EB probe system developed at TWI, with a 2D data sample and typical compiled data set
 

All samples were held stationary, with the machined side facing down, in a jigging arrangement, which was the same for both the laser and the EB welding trials, and the same as the one used for the initial welding trials.[1] The sandwich-type jigging arrangement comprised top clamping bars made of steel and a heavy-section steel backing plate. For the laser welding trials, an airknife was used to reduce the risk of fume and/or spatter damaging the cover slide and focussing optic.

Scope of work

All Yb-fibre laser and electron beam welds produced in this investigation comprised melt runs carried out in the flat (PA) welding position. To avoid problems with back-reflection in case of laser welding, the (centre of the) laser beam was rotated 10° around the welding direction, which is noticeable from the cross-sections shown further in this paper. All Yb-fibre laser and EB melt runs were produced with the focal spot positioned on the surface of the material. Melt runs were produced at travel speeds ranging between 1 and 20m/min by moving the beam from the thick to the thin end of the sample. The points at which full penetration was achieved were recorded and the thickness of the tapered sample measured at those points. These points were used to construct the welding performance graphs described further in the paper.

Results and discussions

Laser beam welding trials

In the initial welding trials on steel[1], it was observed that, when welding with the highest beam quality laser (available at the time), i.e. a 5kW YLR-5000 Yb-fibre laser with a BPP of 4mm.mrad, the depth of penetration was irregular and lower than expected, when using 4kW of laser power and a 400µm (86% point) focal spot diameter. This lower-than-expected performance can be seen in Figure 3. In fact, the performance of this laser was lower than that achieved for a conventional lamp-pumped Nd:YAG laser, at 4kW and a focal spot size of 440µm. The erratic penetration behaviour suggested the presence of a metal vapour plume, or perhaps even a plasma above the weld pool, scattering or absorbing the laser power. For this reason, a simple inert gas side-jet shielding arrangement, as used when welding aluminium, for example, was applied in an effort to suppress the metal vapour interference. This side-jet shielding comprised a 4mm inner diameter copper tube, orientated behind the laser, at an angle of between 20 and 30° from the weld centre-line. Using helium, the welding performance was considerably improved, as shown in Figure 3, confirming the presence of a metal vapour plume. However, the presence of a plasma remained doubtful, as the use of argon (instead of helium) also resulted in a similar performance improvement. Also, what remained unexplained at the time, was why the presence of a metal vapour plume would affect the welding performance for welding speeds up to as high as 8 and 10m/min. After all, past experience on steel has shown the effects of plasma, at the CO2 laser wavelength, and plume, at the Nd:YAG wavelength, but only at welding speeds lower than 1.5 to 2m/min. [2,3]

Fig.3. The welding performance on steel, of a 4mm.mrad Yb-fibre laser, YLR-5000, used at 4kW and a 400µm diameter spot, both without and with (argon) side-jet shielding, and of a 23mm.mrad lamp-pumped Nd:YAG laser used at the same power and spot size of 440µm (without side-jet shielding)
Fig.3. The welding performance on steel, of a 4mm.mrad Yb-fibre laser, YLR-5000, used at 4kW and a 400µm diameter spot, both without and with (argon) side-jet shielding, and of a 23mm.mrad lamp-pumped Nd:YAG laser used at the same power and spot size of 440µm (without side-jet shielding)

The set-up and laser processing conditions used in these initial trials were identical,[1] with all trials carried out on the same batch of materials, clamped using the same jigging arrangement (by moving the laser beam over the samples) and using the same laser power of 4kW (as measured at the workpiece). The only differences between the two laser systems shown in Figure 3, were the beam quality of the laser sources and the processing optics used, resulting in a small difference in focal spot diameter, i.e. 400µm instead of 450µm (86% points). Despite a 27% higher power density for the higher beam quality laser, a lower welding performance was recorded. Moreover, when using the same high beam quality laser, focused in a 140µm spot diameter, no irregular penetration was noticeable without the inert side-jet shielding. Despite a power density more than 8 times higher, the welding performance was similar to the one achieved for the 400µm spot size with side-jet shielding, for welding speeds below 5m/min, as shown in Figure 4.

Fig.4. The welding performance improvement that can be achieved by applying a side-jet shielding (at 4kW, 400µm focal spot), or by choosing a smaller spot size (140µm), with the performance of a 23mm.mrad lamp-pumped Nd:YAG laser (at 4kW, 450µm focal spot) as reference
Fig.4. The welding performance improvement that can be achieved by applying a side-jet shielding (at 4kW, 400µm focal spot), or by choosing a smaller spot size (140µm), with the performance of a 23mm.mrad lamp-pumped Nd:YAG laser (at 4kW, 450µm focal spot) as reference
 

In the initial trials,[1] a conventional lamp-pumped Nd:YAG laser was used as a point of reference. The 200/150 processing head for this laser produced a 450µm focal spot and a beam brightness of 11x105W/mm2.sterad. In contrast, for the YLR-5000 laser, with 4kW of laser power transmitted through a 100µm optical fibre, a 120/500 processing head was used, producing a nominal spot diameter of 400µm, and a beam brightness of around 239x105W/mm2.sterad. Initially, it was thought that this 22-fold increase in beam brightness was the reason of the irregular penetration. However, when using the same YLR-5000 Yb-fibre laser at the same power (4kW) but focused in a 140µm spot size, this irregular penetration was not observed, notwithstanding a similar beam brightness of 230x105W/mm2.sterad. An attempt at explaining this, is given further below.

Following the findings from this initial investigation,[1] trials were carried out using an even higher beam quality and brightness YLR-4000Yb-fibre laser system, at 4kW, on steel tapered samples. Argon side-jet shielding was applied as a result of earlier observations. The beam quality of the YLR-4000 Yb-fibre laser measured 2mm.mrad. Using a 120/500 processing head (and a 50µm processing fibre), a focal spot diameter of 200µm was produced at a beam brightness of 1583x105W/mm2.sterad, i.e. 6.6 times higher than the YLR-5000 [1]Yb-fibre laser system used in the initial investigation. However, a spot diameter close to 400µm, as used in the previous trials, could not be achieved with the processing heads available. Instead, a 200µm spot size was selected, allowing the performance to be compared with that achieved with a 7mm.mrad Yb:YAG disc laser used in the initial trials.[1] This disc laser was equipped with a 200µm processing fibre and 200/200 processing head, producing a beam brightness of 51x105W/mm2.sterad, 31 times lower than the (new) YLR-4000 set-up used.

The welding performance of both these laser systems is shown in Figure 5, from which can be seen that the application of an argon side-jet shielding was unable to reproduce the performance enhancement observed from the initial trials.[1] However, for the brightest of the laser beams used, the 2mm.mrad YLR-4000 Yb-fibre laser focused in a 200µm spot, a step-change in welding performance was observed, as shown in Figure 5, between 7.5m/min and 10m/min. It was at this speed range also, that a step-change in welding performance (on steel) had been observed during the initial trials, for the YLR-5000 laser focused in a 400µm focal spot with the 120/500 process head. [1]

Fig.5. The welding performance on steel, of a 2mm.mrad Yb-fibre laser, YLR-4000, used at 4kW focused in a 200µm diameter spot, with side-jet shielding, compared with that of a 7mm.mrad Yb:YAG disc laser used at the same power and focal spot diameter (without side-jet shielding)
Fig.5. The welding performance on steel, of a 2mm.mrad Yb-fibre laser, YLR-4000, used at 4kW focused in a 200µm diameter spot, with side-jet shielding, compared with that of a 7mm.mrad Yb:YAG disc laser used at the same power and focal spot diameter (without side-jet shielding)
 

As mentioned above, an (ionised metal vapour) plasma was initially thought to have affected the steel welding performance for speeds below 8m/min. However, this was put in doubt, when argon (with a lower ionisation potential) resulted in a welding performance improvement similar that of helium, when supplied through a side-jet shielding arrangement. Moreover, Greses had shown that, when welding steel at the Nd:YAG wavelength, a plume instead of a plasma forms above the weld pool, comprising non-ionised metal vapour.[3] As the wavelength and power density at focus of the YLR-5000 Yb-fibre was similar to that of the Nd:YAG used in the initial trials,[1] for which this phenomenon had not been observed, no plasma was expected. Notwithstanding these arguments, disagreement remains amongst researchers on whether a plasma is created when welding with high-brightness lasers. Researchers in Germany, for instance, recently confirmed the presence of a 'weak' plasma when welding with high beam brightness lasers, although little information was published as to the conditions under which this plasma was formed. [4]

The authors note that in both cases where lower-than-expected welding performance was observed, i.e. for the YLR-5000 laser during the initial trials[1] and the YLR-4000 in recent trials, process heads with the same, long focal length focussing lenses of 500mm were used. The authors believe that the very narrow laser beam created by the long focal length lens results in a high enough power density above the processing point to result in partial ionisation of the hot metal vapour ejected from the weld pool. This creates a column of hot metal vapour plume mixed with a 'weak' plasma between the focussing lens and the processing point, which obstructs the laser power from entering the keyhole by either reflection or absorption. In case of a shorter focal length lens, the power density drops more quickly (considering the square relationship with the beam diameter) the further away from the processing point, creating less favourable conditions above the weld pool to form a plasma. Researchers in Germany, recently visualised the extent of the column of hot metal vapour ejected from the weld pool, which in some instances reached all the way up to the focusing lens positioned 500mm above the processing point. [7]

The effect is similar to that of 'thermal blooming', whereby a stagnant column of air in a beam path, or a beam path containing hydrocarbons, can magnify and distort a CO2 laser beam passing through it, causing aberrations when focused.[5,6] A way around this 'thermal blooming' is to 'stir up' the air column in the laser beam path by introducing a gentle flow of dry air down the beam path.[5,6] A similar approach was tried here, blowing inert gas across the beam path to disperse the metal vapour column. The approach, established together with researchers at IWS (Dresden, Germany), comprised the application of a range of cross-jets, through which argon gas was fed, positioned at various locations (up to 300mm) above the weld pool, in addition to an argon side-jet shielding, to disperse the metal vapour medium obstructing the laser power and affecting the penetration. As can be seen from Figure 6, this resulted in a noticeable improvement in the steel welding performance of the 2mm.mrad YLR-4000 system focused in a 200µm focal spot.

Fig.6. The welding performance improvement that can be achieved by applying an argon side-jet and a series of argon cross-jets when welding steel using a 2mm.mrad Yb-fibre laser (YLR-4000) at 4kW and a 200µm spot size
Fig.6. The welding performance improvement that can be achieved by applying an argon side-jet and a series of argon cross-jets when welding steel using a 2mm.mrad Yb-fibre laser (YLR-4000) at 4kW and a 200µm spot size
 

It can be observed from Figure 6 that the welding performance (on steel) is improved considerably for welding speeds of 7.5m/min and below. In contrast, no change in performance was recorded for welding speeds of 10/min and higher, which is believed to result from the inertia of the metal vapour column. It is noteworthy that remote welding is often carried out at speeds above 7.5m/min, which means that, based on the above results, the welding performance would not be affected.

The cross-sections shown in Figure 7 and Figure 8 demonstrate the effect of applying a side-jet in combination with the cross-jets when welding with the 2mm.mrad YLR-4000 Yb-fibre laser at 4kW, a 200µm focal spot and a welding speed of 3 and 5m/min. The angled weld axis results from the 10° angle (around the welding direction) employed during welding to avoid problems with back-reflection.

Fig.7. Cross-sections of welds produced using a 2mm.mrad YLR-4000 Yb-fibre laser at 4kW (at the workpiece), a 200µm focal spot and a welding speed of 3m/min: a) with argon side-jet, with argon cross-jets; b) with argon side-jet, without argon cross-jets
Fig.7. Cross-sections of welds produced using a 2mm.mrad YLR-4000 Yb-fibre laser at 4kW (at the workpiece), a 200µm focal spot and a welding speed of 3m/min: a) with argon side-jet, with argon cross-jets; b) with argon side-jet, without argon cross-jets
Fig.8. Cross-sections of welds produced using a 2mm.mrad YLR-4000 Yb-fibre laser at 4kW (at the workpiece), a 200µm focal spot and a welding speed of 5m/min: a) with argon side-jet, with argon cross-jets; b) with argon side-jet, without argon cross-jets
Fig.8. Cross-sections of welds produced using a 2mm.mrad YLR-4000 Yb-fibre laser at 4kW (at the workpiece), a 200µm focal spot and a welding speed of 5m/min: a) with argon side-jet, with argon cross-jets; b) with argon side-jet, without argon cross-jets

It should be noted that the depths of penetration that can be measured from these cross-sections, differ from those that can be deducted from the welding performance curves shown in Figure 6. The welding performance curves are constructed from the points at which full penetration was achieved in the tapered samples, whereas the cross-sections were taken at an arbitrary thickness (along the tapered edge) of approximately 7mm, to demonstrate the difference in weld profile and depth penetration achieved with the different laser systems. For those weld runs produced without the cross-jets (on the right of Figure 7 and Figure 8), the depth penetration along the weld length was very irregular, which is why the depths of penetration measured from the cross-sections differ from those obtained from the welding performance curves.

When using the same YLR-4000 laser, but focused in a 125µm spot size (by using a 120/300 processing head), the reduced welding performance was less evident. This set-up produced, at 4kW, a beam brightness of 1459x105W/mm2.sterad, i.e. very similar to the 1583x105W/mm2.sterad achieved when focusing the same laser in a 200µm spot (using a 120/500 processing head). When comparing this with the performance of the YLR-5000 Yb-fibre laser used in the initial trials[1] focused into a 140µm spot size (by using a 120/160 processing head), as shown in Figure 4, the performance improvement is considerable. Unfortunately, these initial trials were not carried out with a side-jet shielding, as at the time, no irregular penetration had been observed with this set-up. Whether this performance improvement achieved for the YLR-4000 (125µm spot) set-up was the result of the 10% smaller spot size (25% higher power density), the six-fold increase in beam brightness or the application of side-jet shielding, compared with the YLR-5000 (140µm spot) set-up, requires further investigation. Moreover, the performance of the YLR-4000 (125µm spot) set-up could be improved further when applying the cross-jets in addition to the side-jet, as shown in Figure 9, for welding speeds equal to and below 7.5m/min.

Fig.9. The welding performance improvement that can be achieved, for steel, by applying a series of argon cross-jets in addition to an argon side-jet, for a 2mm.mrad Yb-fibre laser (YLR-4000) used at 4kW focused in a 125µm diameter spot
Fig.9. The welding performance improvement that can be achieved, for steel, by applying a series of argon cross-jets in addition to an argon side-jet, for a 2mm.mrad Yb-fibre laser (YLR-4000) used at 4kW focused in a 125µm diameter spot

 

Based on the results, both from the initial and the more recent trials, it is reasonable to conclude that the welding performance of high-brightness lasers on steel in terms of depth of penetration for speeds equal to and lower than 7.5m/min, depends greatly on the presence and composition of the metal vapour column forming between the focusing lens and the processing point. It is believed that this metal vapour column scatters and/or absorbs some of the incident laser radiation, thereby affecting the welding performance at those speeds, if not removed effectively. The above trials have demonstrated that one or more (inert) gas jets directed across the laser beam at various point between the weld pool and the focusing lens, can be effective in dispersing this metal vapour column.

Electron beam welding comparison

Rather than dispersing the metal vapour column, a more efficient approach would be to prevent it from forming, or at least expanding above the weld pool. It was thought that this could be achieved, to some extent, by carrying out the welding in-vacuum. To avoid the complications of setting up a high-brightness laser beam to weld in-vacuum, the welding trials were repeated using a high-brightness (in-vacuum) electron beam instead. A beam of 30.7mA, 130kV was used, giving a power of 4kW, in a vacuum of approximately 1x10-3mbar. The electron beam conditions were chosen so that the beam quality and spot size were, at 5mm.mrad and 370µm diameter, comparable with those for the YLR-5000 Yb-fibre laser used in the initial trials, i.e. 4mm.mrad and 400µm. The average beam brightness of the electron beam, however, was 2735x105W/mm2.sterad, and considerably higher, i.e. over 11 times, than that produced by the YLR-5000 Yb-fibre laser.

The results are shown in Figure 10, from which can be seen that, for the slower welding speeds (2-3m/min), the weld penetration achieved with the (in-vacuum) electron beam remains higher than those achieved with the YLR-5000 Yb-fibre laser. However, the welding performance advantage (in terms of depth of penetration) becomes less obvious for the faster welding speeds. Moreover, it is noteworthy that the results shown in Figure 10 for the YLR-5000 Yb-fibre laser were achieved by applying (only) argon side-jet shielding. Considering the increases in depth of penetration that were achieved by applying a series of argon cross-jets in addition to argon side-jet shielding, as shown in Figure 6 and Figure 9, it is not unreasonable to predict a weld penetration for the YLR-5000 laser, equipped with cross and side-jets, comparable to that of the electron beam.

Fig.10. The welding performance on steel of a 4kW 5mm.mrad in-vacuum electron beam weld focused in a 125µm diameter spot, compared with that of a 4kW 4mm.mrad Yb-fibre laser in a 400µm diameter spot, with argon-side jet shielding
Fig.10. The welding performance on steel of a 4kW 5mm.mrad in-vacuum electron beam weld focused in a 125µm diameter spot, compared with that of a 4kW 4mm.mrad Yb-fibre laser in a 400µm diameter spot, with argon-side jet shielding

 

In Figure 11, the same EB results are shown as in Figure 10, compared with those achieved with the 2mm.mrad YLR-4000 Yb-fibre laser. It should be noted that although the electron beam is 70% brighter than the laser beam, i.e. 2735x105W/mm2.sterad compared with 1583x105W/mm2.sterad, the power density used is more than 3 times lower than that of the laser, because of the smaller spot size, which contributes to the difference in weld performance shown in Figure 11. However, the comparison does show the capability of high-brightness laser beams to produce a depth of penetration in steel, equal to, if not surpassing, that achieved with good quality electron beams. Cross-sections of melt runs produced with the YLR-4000 laser and the electron beam at a welding speed of 3m/min, are shown in Figure 12.

Fig.11. The welding performance on steel of a 5mm.mrad (in-vacuum) electron beam (4kW, 370µm focal spot), compared with those of a 7mm.mrad disc laser (4kW, 200µm focal spot) and a 2mm.mrad Yb-fibre laser (4kW, 200µm focal spot), with and without argon side-jet shielding and/or cross-jets
Fig.11. The welding performance on steel of a 5mm.mrad (in-vacuum) electron beam (4kW, 370µm focal spot), compared with those of a 7mm.mrad disc laser (4kW, 200µm focal spot) and a 2mm.mrad Yb-fibre laser (4kW, 200µm focal spot), with and without argon side-jet shielding and/or cross-jets
Fig.12. Cross-sections produced in 7mm thickness C-Mn steel at 4kW, a welding speed of 3m/min, using: a) a 5mm.mrad (in-vacuum) electron beam focused in a 370µm focal spot; b) a 2mm.mrad Yb-fibre laser beam focused in a 200µm focal spot.
Fig.12. Cross-sections produced in 7mm thickness C-Mn steel at 4kW, a welding speed of 3m/min, using: a) a 5mm.mrad (in-vacuum) electron beam focused in a 370µm focal spot; b) a 2mm.mrad Yb-fibre laser beam focused in a 200µm focal spot.


Concluding remarks

A very practical advantage of welding with high beam quality lasers is the large stand-off distances that can be used, for a given focal spot diameter (power density), which are ideal, for instance, for remote laser processing. In addition to the larger working distances, a higher beam quality is also capable of producing, in combination with a small focal spot, welds with an exceptionally high aspect ratio (penetration/depth). Such focused beams have a high brightness, defined as the power density in the spot per solid angle in the cone of the focused beam, and are capable of achieving weld depths in steel similar to those achievable with (in-vacuum) electron beams.

However, the investigation described in this paper, has shown that increasing the beam brightness of a laser system does not automatically result in a greater depth of penetration (for a given welding speed) in steel. Although the weld penetration can be increased by choosing a brighter beam, the (performance) improvement that can be achieved for welding speeds equal to and lower than 7.5m/min, is highly dependent on the composition and extent of the metal vapour plume that forms between the focusing lens and the processing point. When welding with a high-brightness laser, conditions were such that a significant proportion of the incident laser power appeared scattered and/or absorbed, which resulted in a decrease in welding performance. The lower-than-expected depth of penetration was particularly noticeable when using focusing systems with a long focal length lens, e.g. 500mm, at welding speeds equal to and below 7.5m/min. It was demonstrated that the welding performance, at these speeds, could be improved significantly by applying a series of inert gas jets directed across the laser beam path, positioned between the processing point and the focussing lens. The use of argon cross-jets was shown to improve the welding performance (in terms of depth of penetration) of a 4mm.mrad Yb-fibre laser, used at 4kW in a 400µm focal spot, to approach that of a 5mm.mrad in-vacuum electron beam, used at 4kW in a 370µm focal spot.

This confirms the potential of high-brightness lasers, subject to efficient removal of the metal vapour column, to be considered for welding applications up-to-now only suitable for welding with an in-vacuum electron beam. Lasers do not require a vacuum, so they have an obvious practical and economic advantage, provided, of course, that the weld quality is acceptable for any given application. The EB used for comparison purposes here, represents a relatively high-intensity, good-quality welding beam. Electron beams of even higher intensity have been used for welding purposes before, but can suffer from reduced weld quality and weld width, to the point that both weld integrity and consistent beam-joint alignment becomes problematic for particular applications. Based on the weld profiles produced in this investigation, similar concerns should be addressed for high-brightness laser beams.

A further study should be carried out to map the spatial distribution and composition of the metal vapour column to allow the maximum performance to be achieved when using high-brightness laser systems for welding steel, particularly for welding speeds lower than 7.5m/min. The mechanism of power loss, as described in the paper, should also be investigated further, and a systematic approach developed for the efficient dispersal of the metal vapour column through practical means.

Acknowledgments

The authors would like to thank Paul Hilton of TWI and the colleagues at IWS (Dresden, Germany) for their assistance in the trials with the YLR-4000 laser. The authors are also grateful for the assistance of colleagues at Trumpf and IPG in realising some of initial results described in this paper.

Meet the Authors

Dr Ing Geert Verhaeghe (EngD) - Geert is a Mechanical Engineer (MSc) and European Welding Engineer (EWE), who joined TWI in 1996. He is a Principal Project Leader in the Laser and Sheet Processes Group, with particular experience in the welding of aluminium, robotics and automation for welding, weld distortion, arc and laser process monitoring, hybrid laser-arc processing and welding with high-brightness laser sources.

Bruce Dance (MA) - Bruce is a Principal Project Leader in the Electron Beam Department of TWI, with over 20 years of experience in materials processing using electron beams.

References

  1. Verhaeghe, G., Hilton, P.A. (2005) The Effect of spot size and Laser Beam Quality on welding performance When using High-Power continuous wave Solid-State Lasers, in Proceedings of the 24th International Congresson Applications of Lasers & Electro-Optics (ICALEO), Miami, FL, US, Oct 31-Nov 3 2005, 264-271.
  2. Olivier, C.A., Gerritsen, C.H.J. (2000) Study into the use of process gases in high power Nd:YAG laser welding of C-Mn and stainless steels. TWI Industrial Members Report 723/2000, Nov 2000.
  3. Greses, J. (2003) Plasma/plume effects in CO2 and Nd:YAG laser welding, PhD dissertation, University of Cambridge, England, March 2003.
  4. Göbel, G., Brenner, B., Beyer, E. (2007) New application possibilities for fiber laser welding, in Proceedings of the 26th International Congress on Applications of Lasers & Electro-Optics (ICALEO),Orlando, FL, US, Oct 29-Nov 1 2007, 102-108.
  5. Schellhorn, M., von Bülow, H. (1995) CO laser deep penetration welding - a comparative study to CO2 laser welding, in SPIE, V2502, 1995, 664.
  6. Gebhardt, F.G. (1976) High power laser propagation, in Appl Optics, Vol.15, No.6, 1479.
  7. Michalowski, A., He, A., Ru, A., Dausinger, F. (2007) Plume attenuation under high power Yb:YAG laser material processing, in Proceedings of the 4th International WLT Conference on Lasers in Manufacturing (LIM),Munich, June 2007, 357-361.

For more information please email:


contactus@twi.co.uk