Laser joining method can process thick steel

An emerging technique to join thick metal sheets is being investigated for use with equipment readily available at many industrial companies that employ laser welding or cladding already.

Content Dam Ils Print Articles 2017 3 1703ils Fro F1

JAN FROSTEVARG and JONAS NÄSSTRÖM

Multi-layer welding technique works with standard equipment

An emerging technique to join thick metal sheets is being investigated for use with equipment readily available at many industrial companies that employ laser welding or cladding already. The method works by resistance-heating the filler wire using standard weld equipment to produce welds that have close to net-shape surfaces, which in turn could decrease post-processing.

Numerous industrial fields, including railways, shipbuilding, heavy equipment, and pipelines, rely on joining of thick metal sheets. In some cases, the chemistry and structure of the metal are especially important—particularly in pipelines located in remote arctic conditions, which have corrosive ocean elements and cold temperatures that decrease toughness. Weld seams required to have high toughness can be obtained by modifying composition of filler wire, resulting in adapted microstructure.

Traditional method limitations

Typically, gas metal arc welding (GMAW) is used to join thick steel sheets. Even though the weld source is relatively cheap, GMAW requires a wide joint preparation that involves multiple passes. This is both time- and resource-consuming, and also produces a wide heat-affected zone (HAZ). Submerged arc welding (SAW) is an alternative, but has high heat input and slow cooling cycles—and is thereby limited to mild low-alloy, high-strength steels.

Other techniques can be considered to avoid high production costs by arc welding in thick sheets. For example, laser beam welding (LBW) is a method with high energy density that produces a less-wide HAZ, higher welding speeds, and less distortion to the parts being welded. It also allows for much larger penetration depths in single passes—it yields flawless depths up to 12mm with full penetration and up to 20mm with partial penetration, depending on the laser source and optics.

However, LBW has limited gap bridgeability and can suffer from porosity, solidification cracking, and "too-hard" weld fusion zones (FZ) because of the rapid cooling by surrounding metal/atmosphere, effectively quenching the weld so that hard microstructures are produced. To increase homogeneity and quality throughout the depth of a weld, treatments such as preheating can be applied.

Another technique that utilizes the advantages of laser processing is laser-arc hybrid welding (LAHW). Although it is more complex, it combines laser and arc technology to improve both welding speed and penetration depth when compared to GMAW and LBW alone. In thick sheets, the method still suffers from nonhomogeneous filler-base metal mixing through the depth of the weld, and suffers from LBW imperfections such as solidification cracking and porosity. Starting at ~8mm, LAHW also has increased formation of root humping in full penetration welds, which gets increasingly more difficult to suppress when welding thicker sheets.

Narrow-gap multi-layer laser welding

If welding speed requirements are reduced, an emerging technique called narrow-gap multi-layer laser welding (NGMLW) is available for joining sheets that are 8mm and thicker. NGMLW is a laser cladding technique applied in a gap that is gradually filled by each added layer (FIGURE 1) and being developed in many countries, including Japan, China, Germany, and Finland. The method is similar to 3D printing with laser metal deposition (LMD), but applied as a joining technique. The weld speed of NGMLW is between arc and laser welding, but offers higher control of weld metallurgy.

Content Dam Ils Print Articles 2017 3 1703ils Fro F1
FIGURE 1. The NGMLW process shown incrementally filled (a) and in cross-section (b).

Multiple layers create a more uniform microstructure in each layer while some heat treatment occurs in the previous layer, reducing residual stresses created during solidification. A potential risk of applying multiple layers is that every layer is a weld, which can exhibit imperfections in the form of porosity, lack of fusion, solidification cracking, and humping. However, a potential advantage of the process is the possibility to adapt the chemistry of the filler material, adapted with the heat cycle to produce welds with tailored properties. For example, to obtain high toughness in low temperatures, microstructures such as acicular ferrite are desired. These can be obtained through composition of the base material (added material for nucleation sites) as well as thermal cycles that favor their formation.

Proper interaction between the laser and wire is of high importance to ensure stable melting behavior. Besides heating the melt pool—and possibly preheating the gap edge—the laser should irradiate the full width of the wire, which can be accomplished by scanning or defocusing the beam. The energy needed from the laser can be lowered by preheating the wire—for example, by resistance heating to limit excessive melting of base material. This technique is often applied using specialized equipment that includes nozzles to ensure wire positioning, nozzles for providing shielding gas, and oscillating mirror optics (or a galvanometer) for the laser.

Process mechanics

In this demonstrated case, thick steel sheets are joined using a 3mm-wide gap (2.5:1mm diameter of the filler wire) with added root support. Travel speed for each layer is ~1m/min and wire feed rate is ~7m/min, providing a layer thickness of ~3mm. The welding wire is ~1.2mm wide and resistance-heated by a Fronius GMA welding source using constant current, with a target current set beforehand. The laser power source used is an IPG Photonics Yb:fiber laser with a 6kW defocused beam to form a Gaussian profile with a large-enough spot size to ensure full irradiation of the process regime. Both the arc torch and the optics are moved by a robot, with the wire fed from the front of the welding zone and shielding gas provided by a tube in front of the process zone. FIGURE 2 shows a frame obtained with high-speed imaging (HSI) from the process front.

1703ils Fro F2
FIGURE 2. A frame from high-speed imaging (HSI) filmed at the front of the weld, with annotations of the NGMLW process (a) and sealing-pass HSI frames with and without arcing behavior (b).

The wire should be kept in the process zone at all times without touching the sheet edges. It is important to mention that the wire, with a long stick-out, is very difficult to extrude perfectly straight—therefore, specialized nozzles are often employed to guide the wire towards the melt zone. However, the wire can be kept reasonably straight with a stick-out (stand-off distance) at 32mm—even if the wire is originally positioned in the middle of the process zone, it will oscillate. But since the laser spot size covers the whole process region, this problem is partially eliminated so that customized wire nozzles are not necessarily required.

Influence of laser positioning

Positioning of the laser is of high importance, and different settings are recommended for the three different layer regimes: root pass, intermediate pass, and sealing pass. Since the wire is preheated and fed into the melt pool, the laser does not have to directly hit the wire for it to melt, as long as the melt pool is sufficiently deep and hot. Any defocusing width in the ~3–6.5mm range at the process zone works to melt the wire and produce a proper melt pool. If the laser hits the sheet edges, they are partly heated and melted, ensuring wetting and fusion of the weld.

1703ils Fro F3 Small
FIGURE 3. Example images for various parameter settings, showing cross-cuts (a) and weld-pass surfaces (b).

It is important to note that using this setup, the welds in the intermediate passes can suffer from centerline cracking (FIGURE 3) when the laser spot width is too wide. Though not directly observed, it is believed that prominent wetting of the gap edges is the cause by actually providing too-high wetting (FIGURE 4).

1703ils Fro F4 Small
FIGURE 4. An explanation of centerline cracking because of laser shape (a) and arc formation (b).

If the laser heats (and partially melts) the gap edges higher than the intended layer height, surface tension will pull the edges of the melt pool to the rims and during solidification cause tensional forces, promoting crack formation. If the laser only hits the wall at the intended layer height, crack formation is less likely to occur. For the sealing pass, the laser width should be slightly larger than the gap width to enhance proper wetting at the surface, achieving near- to flat (net-shape) surfaces.

Influence of voltage for wire preheating

Sufficient preheating is a prerequisite to have proper melting of the wire and proper fusion to the gap edges, whereas inadequate heat input causes lack of melting of the filler wire and, often, lack of fusion. Conversely, an over-powerful heat input causes arcing between the wire and the melt pool, which may prevent formation of intended microstructure in the weld zone. Depending on machine properties for resistance-heating the wire, process instabilities could cause heat inputs that vary during processing.

In the case shown in FIGURES 2 and 3, the current was locked, but voltage can be adjusted by the welding machine to ensure heating of the wire. If voltage is too high, the process gets increased arc formation frequency and more pronounced arc-related problems. An arc can form if the molten metal bridge between the melt pool and wire breaks, either by wire movement oscillation or off-burning by the laser or current. Once the arc has formed, it uses the path with least resistance (contact area/distance), which in this process are the gap edges that are melted.

The arc formed at low voltages has no or little impact—but at higher voltages, arc formation is easier and occurs for longer durations with a larger and stronger arc, which causes melting of the gap edges. This leads to dilution of FZ and possibly weakens the base material in the HAZ. To ensure high weld quality, arcing should be avoided.

Process recommendations

The NGMLW technique has many advantages compared to pure GMAW, SAW, LBW, or LAHW for joining thick sheets, with a productivity rate between pure arc or laser techniques. However, most disadvantages compared to the other techniques are eliminated and every layer can have adapted metallurgy, which is something to be considered for tempered thick sheets.

With NGMLW, the following processes are recommended:

  • For intermediate passes, laser optics should be positioned appropriately to obtain a spot size that is equal to the gap width at the intended layer height;
  • The wire should be fed into the laser spot on the bottom of the gap/surface of the previous layer at ~35°;
  • The wire current settings should be as high as possible (750W or higher) while simultaneously avoiding any form of arcing; and
  • For the sealing pass that forms the weld cap, wire feeding rates should be adjusted to obtain proper filling, and have a laser spot size slightly larger than the gap width.

This NGMLW method will be used in a current EU Horizon 2020 project for welding thick steels in arctic offshore applications, which means that the welds (and base metal) need to have high properties in cold (-60°C) and corrosive (withstand rust) conditions. In the project, submerged arc welding for thick-sheet steel welding will be compared with a "creative" laser-arc hybrid welding setup, as well as NGMLW.

REFERENCES

1. G. Turichin, M. Kuznetsov, M. Sokolov, and A. Salminen, "Hybrid laser arc welding of X80 steel: Influence of welding speed and preheating on the microstructure and mechanical properties," 15th Nordic Laser Materials Processing Conference, Lappeenranta, Finland (2015).

2. T. Ilar, I. Eriksson, J. Powell, and A. Kaplan, Phys. Procedia, 39, 27–32 (2012).

3. X. Zhang et al., J. Laser Appl., 23, 022002 (2011); doi:10.2351/1.3567961.

4. R. Schedewy, D. Dittrich, B. Brenner, and E. Beyer, Opt. Lasers Eng., 50, 1230–1241 (2012).


DR. JAN FROSTEVARG(jan.frostevarg@ltu.se) is an associate professor andJONAS NÄSSTRÖM is a graduate student, both in the Division of Product and Production Development at the Luleå University of Technology, Luleå, Sweden.

More in Welding