Enhancements to Power Beam Welding Processes for Land Transport
Gongqi Shi, Paul Hilton, Geoff Booth and Chris Punshon
Paper presented at IIW Annual Assembly, 15, 16 July 2004, Osaka Japan
Abstract
The two principal power beam processes, electron beam and laser beam welding, offer significant competitive advantages in terms of productivity, joint performance and joint quality. The processes however have their drawbacks. For example electron beam welding is traditionally carried out in vacuum and laser welding requires accurate part fit-up.
Recent work has been directed at overcoming these limitations. Techniques have been developed that enable electron beam welding to be carried out in non-vacuum and reduced pressure atmospheres. An example is described where non-vacuum electron beam welding has been used for welding aluminium for the automotive industry.
Hybrid laser-arc welding permits gap tolerances similar to those of conventional arc welding whilst maintaining many of the advantages of laser welding. An example from shipbuilding and off highway vehicle manufacture is described where good quality joints can be made in steel for shipyard and heavy manufacturing applications.
1. Introduction
The two principal power beam processes, electron beam and laser beam welding, have been used in industrial applications for more than thirty years. In both processes, the energy beam is focussed to a very small spot; electromagnetic fields are used for electron beams and lenses or mirrors for laser beams. When the spot is focussed on a metal such as steel, aluminium or titanium, the energy density is sufficiently high to form a vapour keyhole in the metal.
Deep penetration keyhole welding enables welds to be made with very large depth to width ratios. This consequently leads to narrow welds, made with relatively low heat input. Frequently, only one weld pass is made and welding speeds are well in excess of those typical of arc processes. As a result, power beam welding gives rise to relatively low levels of distortion which, when combined with high productivity rates arising from high welding speeds, makes power beam welding economically attractive.
In common with other welding processes, however, power beam welding is limited by the characteristics of the process. In the present paper, the requirement that electron beam welding is performed in vacuum is considered and techniques are described that enable electron beam welding to be carried out in non-vacuum and reduced pressure environments.
In addition, the need for good fit-up when laser welding is addressed. Gaps must generally be small enough to ensure that the focussed beam (typically 0.3mm diameter or less for CO 2 laser welding) does not pass directly through the joint without achieving a weld. Techniques for enhancing the joint gap bridging capabilities of laser welding are investigated.
2. Electron Beam Welding
2.1 Vacuum considerations
Electron Beam (EB) welding is traditionally performed with the component to be welded contained entirely within a vacuum chamber. With a conventional beam generator the chamber pressure has to be sufficient to ensure that a vacuum of better than 5x10 -4 mbar can be achieved in the electron gun vacuum envelope. This is primarily to maintain high voltage insulation in the gun, and to prevent oxidation of the hot refractory metal cathode, but also to minimise scattering of the beam at its primary focus position. In industrial systems, the chamber pressure typically ranges from of the order of 10 -2 mbar, for high production equipment used for welding relatively thin sections, to 10 -4 mbar for high power, or precision welding equipment, where beam quality and prevention of oxidation are paramount. Whilst in many applications a vacuum welding environment is attractive, for large parts and where high productivity is required, the need for a pumped chamber can detract from efficient working.
2.2 Non vacuum electron beam welding
In search of a solution to the vacuum requirement, Schumacher [1] , pioneered the development of high power non-vacuum EB (NVEB) welding in which the electron beam was generated in a high vacuum envelope, as usual, but emerged from the gun column via a series of differentially pumped vacuum stages separated by small diameter orifices. This allowed the system to be used for welding at atmospheric pressure, thereby eliminating the need for a vacuum chamber. Medium power (typically 25kW) machines have been employed in the American automotive industry in the production of transmission parts and engines components for many years encouraged by the production volumes in this industry. Until recently, NVEB was not used extensively in Europe owing to the smaller production volumes in the European motor industry. Currently the increased use of aluminium alloys has generated a resurgence of interest in the NVEB welding process.
Schulze and Powers [2] , have reported that NVEB welding was identified in a comparative study to be the optimum method for manufacture of a structural beam in 2.5mm thick AlMg3 aluminium alloy for the VW/Audi group and has since been adopted for production. The process is used to produce edge welds at a welding speed of 12m/min ( Fig.1).
Fig.1. Macro section of NVEB weld (12 m/min, 19.3 kW) at the flange of 2 x 2.5 mm AlMg3 deep-drawn shells (Courtesy of PTR-Precision Technologies Inc.)
The parts require two welds of 1359mm in length and in excess of 2,000 parts per day are required making a high productvity welding process essential. The system illustrated below ( Fig.2) has been engineered to permit production of two fully welded parts every 65 seconds. The welding time in the production sequence is 7 seconds per seam. The high welding speed is made possible by the high power output available from the electron beam generator (19.3kW). Current research at TWI is examining the performance of an NVEB welding system with a maximum power of 150kW which, it is hoped will extend the performance capability of this welding method further.
|
|
Fig.2. Equipment schematic and the production NVEB welding machine for structural beam production (Courtesy PTR Precision Technologies Inc.) |
2.3 Reduced pressure electron beam welding
In parallel with the development of high power non-vacuum EB welding at TWI, the NVEB system has permitted the investigation of welding with local sealing and pumping to achieve a pressure in the vicinity of the beam of the order of1mbar. This is known as Reduced Pressure Electron Beam welding (RPEM). Working at this pressure permits the use of local seals which is impractical for better levels of vacuum and offers a considerable improvement in beam quality and thus stand-off distance ( Fig.3).
Fig.3. Comparison of 60 kW electron beam profiles at near atmospheric pressure (900mbar) and 5mbar illustrating improved beam profile and possibility of increased stand-off distance tolerance.
To date, this technology has been used predominantly in nuclear waste encapsulation and pipeline fabrication applications involving copper and nickel alloys for the former and C-Mn steels in the latter case. The indications are, however, that welding at this pressure is equally effective for aluminium and titanium alloys where welds have been produced with a marked absence of porosity and obvious contamination in thicknesses in excess of 16mm. The ability to employ local sealing and the reduction in pump down time to near insignificant levels afforded by Reduced Pressure operation will doubtless lead to the adoption of this method in the automotive industry where high vacuum or partial vacuum EB welding has been employed previously.
3. Enhancing the Joint Gap Bridging Capabilities of CO2 Laser Welding
3.1 Background
Generally speaking, the joint gap bridging capability of laser welding is less than that of arc welding. When using laser alone for welding butt joints in thin gauge metals, gaps of less than 10% of material thickness must typically be maintained. If the gap is too wide then the laser beam may pass straight through the joint without touching the joint faces. This requirement for good fit-up is often seen as a major limitation of laser welding, particularly in sectors such as ship building and off-highway vehicle manufacture where changes in fabrication shop practice would be required.
Laser welding with a filler was developed to increase tolerance to joint fit-up; an additional benefit is that by careful selection of the filler material weld metal properties can be improved. There are several ways in which a filler may be introduced; these include a continuous wire feeding system, the introduction of a thin strip between the joint faces and the addition of powder.
Wire feeding techniques for laser welding are well established. The filler wire nozzle is mounted onto the laser welding head so that the wire is aligned with the centre of the joint. Accurate positioning of the filler wire is essential, but significant improvements in tolerance to fit-up can be achieved. One drawback of the use of a cold filler wire, however, is that energy is used to melt the filler wire. This typically leads to a decrease in travel speed of around 20%, to compensate for the energy used.
This shortcoming can be overcome by using hybrid laser MIG/MAG welding, in which the laser beam and arc interact to produce a single process zone ( Fig.4). The arc supplies additional energy and the consumable assists in filling the gap. The characteristics of the hybrid weld are influenced by the ratio of laser power to arc power; the best results are generally achieved when similar levels of laser and arc power are used. One key feature, however, is the detail of the joint preparation used; V-preparations may be beneficial for thicknesses above about 10mm. However, benefits are also available when using a simple square butt preparation, which is of immediate industrial interest.
Fig.4. Hybrid laser-arc welding
3.2 Experimental work
The objective of the present work was to establish the gap bridging capability of autogenous laser welding and then compare this with results obtained using hybrid CO2 laser/MAG welding.
Welding parameters were first obtained to produce fully penetrating welds on close fitting butt joints in 8mm thick C-Mn steel. These conditions were then used on specimens with constant joint gaps to establish limits.
For the hybrid process the same approach was also used; the hybrid parameters were then re-optimised, at several fixed gaps, by adjusting the wire fed speed or travel speed. Additionally samples were welded with continuously varyingjoint gap, achieved by aligning the two pieces to be joined at a small angle to each other.
All welding trials were carried out on 8mm thick C-Mn steel complying with BS EN 10025 S275. This is a structural steel with specified yield strength of 275N/mm 2 ; compositions of the two plates used are given in Table 1. Plate surfaces and edges were milled and degreased prior to welding.
Table 1 Chemical analysis of steels used in the hybrid CO2 laser-MAG trials (wt%).
| Grade | Thickness/diameter | C | Si | Mn | P | S | Cr | Mo | Ni | Al | Cu |
Plate |
S275 |
8mm |
0.17 |
0.02 |
0.92 |
0.009 |
0.014 |
0.019 |
0.004 |
0.02 |
0.053 |
0.074 |
Plate |
S275 |
8mm |
0.21 |
0.04 |
0.87 |
0.018 |
0.009 |
0.023 |
0.004 |
0.022 |
0.064 |
0.041 |
C-Mn-Si steel wire |
BSEN 440 |
1.2mm |
0.07 |
0.15 |
1.45 |
0.021 |
0.015 |
0.033 |
0.003 |
0.024 |
0.003 |
0.14 |
The laser used was a continuous wave fast axial flow industrial CO 2 laser manufactured by Laser Ecosse. Laser power at the workpiece was approximately 4kW and the beam was perpendicular to the plate surface. A 150mm focal length lens was used to focus the beam to a spot size of approximately 0.3mm.
The MAG welding equipment was a Fronius TPS 450 inverter power source and TIME 30 wire feed unit. A standard 16mm gas shroud was used for the MAG shielding.
For autogenous laser welding, coaxial gas shielding with helium and a plasma suppression side jet of helium was used. For hybrid welding, helium shielding gas was provided coaxially with the laser beam; plasma jet suppression was not used. The shielding gas through the MAG torch was 55% helium, 43% argon and 2% carbon dioxide.
The hybrid welding trials used a 1.2mm diameter A18 C-Mn steel filler wire; Table 1 provides its composition.
All welds were assessed visually and by sectioning. Welds were characterised in accordance with the requirements of BS EN ISO 13919-1:1997. Table 2 summarises the geometrical quality requirements for 8mm thick steel.
Table 2 Criteria for classification a of quality in laser welded steel, as specified in BS EN ISO 13919:1-1997.
Quality acceptance levels | Moderate D | Intermediate C | Stringent B |
Relative b mm | Absolute b mm | Max. value b mm | Relative b mm | Absolute b mm | Max. value b mm | Relative b mm | Absolute b mm | Max. value b mm |
Undercut |
0.15t |
1.2 |
1.0 |
0.1t |
0.8 |
0.5 |
0.05t |
0.4 |
0.5 |
Excess weld metal |
0.2 + 0.3t |
2.6 |
5.0 |
0.2 + 0.2t |
1.8 |
5.0 |
0.2 + 0.15t |
1.4 |
5.0 |
Incompletely filled groove |
0.3t |
2.4 |
1.0 |
0.2t |
1.6 |
0.5 |
0.1t |
0.8 |
0.5 |
Excessive penetration |
0.2 + 0.3t |
2.6 |
5.0 |
0.2 + 0.2t |
1.8 |
5.0 |
0.2 + 0.15t |
1.4 |
5.0 |
Incomplete penetration |
0.15t |
1.2 |
1.0 |
none |
0.0 |
0.0 |
none |
0.0 |
0.0 |
Linear misalignment |
0.25t |
2.0 |
3.0 |
0.15t |
1.2 |
2.0 |
0.1t |
0.8 |
2.0 |
a The quality levels B, C and D refer to stringent, intermediate and moderate requirements respectively.
b 'Relative' is the value calculated using the plate thickness t, 'absolute' is that same value for the 8.0mm thickness used in this investigation. Maximum is the maximum permissible value. The smaller of the two values(absolute v. maximum) is used to establish the quality level.
3.3 Results
A typical cross section of an autogenous weld is shown in Fig.5. For these joints, the critical imperfection was 'incompletely filled groove'. The experimental points in Fig.6 depict the effect of root gap on the depth of the incompletely filled groove. This style of presentation can be used to predict the gap below which an acceptable weld can be produced. For example, an incompletely filled groove of 0.5mm would be the limit for quality level B. This should be achievable for gaps up to 0.25mm.
Fig.5. Cross section of autogenous CO2 laser weld in 8.0mm C-Mn steel plate produced with 4.0kW laser power and 1.2m/min travel speed, 0.2mm joint gap (mm scale shown).
Fig.6. Effect of root gap on the depth of the incompletely filled groove for autogenous laser welds in 8mm C-Mn steel, produced with 4kW laser power at 1.2m/min travel speed.
Figure 7 shows typical sections through hybrid welded joints, with gaps ranging from 0mm to 1.7mm. Figure 8 summarises the effect of joint gap on quality, but for hybrid welding there is also the possibility of excess weld metal when welding with small gaps. It should be noted that these results were obtained using joints with a varying gap - 0mm at the start of the weld and 2mm at the end. To achieve quality level B (stringent) the joint gap must be restricted to less than 1.2mm, as can be seen from Fig.8.
Fig.7. Sections of hybrid laser/MAG welds in 8mm thick C-Mn steel produced with 4kW laser power, 4kW MAG power, 5.8m/min wire feed speed, MAG pushing, 0 separation and 1.0m/min travel speed on square edge butt joint with variable gap up to 2mm (mm scale shown).
Fig.8. Effect of joint gap on depth of incompletely filled groove for laser arc hybrid welding conditions as in Fig.4
As discussed earlier, trials were carried out to optimise the welding conditions for predetermined joint gap. During this optimisation, only wire feed speed and travel speed were changed.
Figure 9 presents typical results, in this case a single welding condition was capable of accommodating gaps ranging between 0.6mm and 1.6mm.
3.4 Discussion
Sections 3.1 and 3.2 summarise some results of an extensive experimental programme carried out at TWI. Of necessity, it has not been possible to include all the data obtained in the study, but nevertheless it is possible to make some general points.
For autogenous laser welding, the width of gap that can be tolerated is similar to the diameter of the focussed beam; at greater gaps the beam simply passes through the joint. In the present work, as the joint gap approached 0.3mm(the beam diameter) the weld quality deteriorated and became unacceptable, primarily due to incompletely filled groove. This can simply be related to there being insufficient molten metal to fill the gap, remembering that no material is added.
As expected, the hybrid laser-arc process had much greater tolerance to joint gap. Gaps up to 1.2mm could be accommodated, whilst still maintaining the stringent quality level B, without altering process parameters. Amending the process parameters allowed a gap of 1.6mm to be tolerated.
The increased tolerance is believed to be largely the result of two factors. Firstly, the laser and arc processes interact to develop a process zone that prevents the laser beam simply passing through the joint when the gap is larger than the beam diameter. Secondly, the addition of filler from the consumable means that material is available to fill the joint gap, reducing joint underfill.
Fig.9. Sections of hybrid laser/MAG welds in 8mm thickness C-Mn steel produced with 4kW laser power, 4kW MAG power, 9.0m/min wire feed speed, MAG pulling, 1.5mm separation and 1.0m/min travel speed, on a square edge butt joints (mm scale shown)
It is clear that the greater gap bridging capability of hybrid laser-arc welding broadens the process tolerance and relaxes production constraints. This is of particular interest in industries where relatively heavy sections are used, including ship building and off-highway vehicle manufacture.
The present work has presented data for only a very limited set of conditions. The quantitative conclusions therefore, are relevant to a relatively narrow application, but clearly the general principles can be used to develop tolerances to particular combinations of material, material thickness, laser type and power etc.
In the present study, the capabilities of the hybrid process were established using joints with fixed and varying gaps. Larger gaps could be tolerated by increasing wire feed speed or reducing travel speed, while all other process parameters remained the same. This strongly suggests that the capabilities of the hybrid process could be further enhanced by incorporating an adaptive control approach. In this way, the joint gaps could be sensed and the process parameters changed to ensure that the optimum parameters are used at all times. This area, involving a single parameter, either wire feed speed or travel speed, as the variable for the adaptive control feedback loop, is currently underactive development at TWI. It is believed that large productivity gains may readily be accessible using this approach.
4. Concluding Remarks
All welding processes are continuously being improved in the search for competitive advantage and attention is generally directed to the removal or easement of specific process limitations.
The present paper has indicated that some historically accepted restrictions of power beam can be relaxed, giving rise to a broader range of industrial application. It is believed that there are frequently benefits to be gained by challenging the boundaries of process applicability.
References
- B.W Schumacher: 'High Power Electron Beams in the Atmosphere' Electron and Ion Beam Science and Technology, 3rd International conference (1968).
- K.R Schulze and D.E Powers: 'Applying the EBW Process Directly in Atmosphere to do High Productivity Aluminium Welding Tasks' IIW Doc. ref IV-840-03.