2.1

Metal vapor plume behavior at mild and stainless steel

Based on the investigation shown in [14] we have demonstrated a clear influence of the metal vapor plume on penetration depth and spatter ejection. The spatter behavior can be traced back to melt pool oscillations: the out flowing metal vapor irregularly tilts along the laser axis in all directions with high dynamics. On the other hand, since there is a strong interaction between the metal vapor and the incident laser beam [8, 9, 10, 11], this leads to a spatially and temporally inhomogeneous coupling of the laser into the work piece.

When the feed rate is increased the capillary front becomes progressively more flat and the weld pool dynamics increases significantly at 6 m/min. As a result, the weld pool starts to bulge behind the capillary aperture and becomes more turbulent. This produces a slightly inhomogeneous upper bead. At this feed rate a transition zone between two different process regimes at low and high welding speeds is attained.

When the feed rate increases further to 8 m/min, larger amounts of molten spatters fly off mainly from the rear of the capillary opposite to the feed direction (see fig. 3). Apart from contamination, this causes a considerable loss of material in the seam whereby no weld reinforcement occurs and therefore the strength of the seam is reduced, especially for dynamic loading. This behavior can be explained by a stronger mean inclination angle of the keyhole, whereby the absorption of laser power occurs increasingly on the capillary front [12]. From this follows a stronger overheating of the melt and a resulting stronger vaporization rate, which is the driving force behind all dynamic effects in laser welding [4].

Figure 1:

Lateral recordings at different times at 3 m/min feed

Figure 2:

Coaxial nozzle connected to a cross jet

Figure 3:

Lateral observation with and without using a coaxial nozzle at different feed rates

Since the laser power can be coupled into the work piece spatially and temporarily more homogeneously, the vaporization inside the capillary is more homogeneous as well. Due to this, the capillary becomes stabilized and the loss of material due to spattering is reduced for feed rates of v < 6 m/min compared to welds without nozzle, whereby a higher upper bead quality can be obtained (see fig. 1). Besides this, using the coaxial nozzle reduces the upper seam width and increases the penetration depth (see fig. 4), as well as the penetration depth stability.

Figure 4:

Penetration depth and seam width with and without the coaxial nozzle

To reduce the influence of the metal vapor a coaxial nozzle was designed (figure 2). Compressed air is blown in the direction of the workpiece to cool down the metal vapor and blow it out of the process area. The interaction length of the laser beam and the metal vapor is reduced to a minimum. Below 8 m/min the effect is clearly visible in figure 1 and 3.

The described effects have been investigated on mild steel when welding without any shielding gas. The next logical step is to transfer the results to stainless steel and hence to oxide-free welds where in general shielding gas is used. The experiments have been done on 3 mm stainless steel 1.4301 with a TruDisk 4001, a laser light cable diameter of 200 μm and spot sizes of 300 μm (1.5 magnification). Further parameters have been set to: feed rate 3.4 m/min, laser power 2.4 kW, defocusing of - 2 mm, flow rate of Argon of 40 l/min. Clear improvements can be seen in figure 5. The metal vapor plume was suppressed and more laser power could couple into the specimen more homogenously and hence smooth welds have been generated.

Figure 5:

Welding of stainless steel 1.4301 with shielding gas lateral nozzle

Figure 6:

Experimental setup

2.2

Welding of different batches of mild steel

Partial and full penetration welds in mild steel S235 are carried out by a Yb:YAG solid-state disk laser (TruDisk 16002) with a wavelength of λ = 1.03 μm and a beam parameter product of 8 mm*mrad from TRUMPF Laser- und Systemtechnik GmbH. Due to a higher achievable penetration depth and a higher weld seam quality the focal position FP was set -2 up to -3 mm into the workpiece [1] and the described coaxial nozzle was used. A high speed camera on the root side in combination with an EOS 7D and another high speed camera on the top side allows a very detailed visualization of the process.

For full penetration welds the weld seam quality can be strongly influenced by weld parameters. The laser power has been varied from 7.75 kW to 13 kW. The trials have been performed on 12 mm thick mild steel S235 and welding speed was kept constant at 1 m/min. Typical behaviors of former analysis [14] are shown in figure 7. Three main regimes can be seen: too low power leads to either just a partial penetration or a dropping effect on the root side. When Increasing the laser power to 8.5 up to 10.2 kW a very good weld seam can be achieved on the bottom side as well as on the root side. Laser powers greater than 10.2 kW lead to melt ejections and hence to a loss of material.

Figure 7:

The influence of laser power on full penetration welds on 12 mm mild steel sheets at v = 1 m/min.

Left side: The quality is rated in school grades, where 1 equals very good quality

Right side: Weld seam quality on top side, cross-sections, and weld seam quality on root side

Based on the shown experiments and results we looked into different batches and different suppliers of mild steel S235. On the left hand side of figure 8 welding results of batch 2 and supplier 1 is shown. We see the typical regimes of poor quality (partial penetration, dropping) below 7.7 kW laser power. Increased laser power of 8 to 9 kW lead to good weld seam qualities. Hence a process window of 1 to 1.5 kW was determined. Batch 3 was performed under the same boundary conditions but the results look quite different (figure 8, right hand side). Even at low power levels of 7 kW a full penetration and good weld seams can be obtained which means that less power is needed for a full penetration weld. With power levels up to 9 kW still good weld qualities are visible and hence a process window of around 2.5 kW was possible. Such big process windows are in general known from welding with 10 μm CO₂ lasers.

Figure 8:

The influence of laser power on full penetration welds on 12 mm sheets at v = 1 m/min.

Left side: Batch 2, Supplier 1; Right side: Batch 3, Supplier 2

The quality is rated in school grades, where 1 equals very good quality

Visual inspection during the weld process showed slight spatters to the bottom side which lead to the assumption that the key hole is open throughout the entire welding process. At other batches a different behavior is seen. The melt flow is redirected to the workpiece and almost no spatters are ejected to the bottom side [14].

figure 9 shows the weld seam qualities of batch 3, supplier 2 at different power levels. One can see a full penetration weld at 6.8 kW with very good weld seam qualities.

Figure 9:

Weld seam qualities of Batch 3, Supplier 2