The coupling of laser and MIG processes to form a hybrid process with a single melting pool offers advantages, compared to each process individually. These advantages are characterized by high welding speed, gap bridging capability and welding depth1,2). At the same time, the low line energy leads to a reduction of thermal stress and thus to low distortion3). First trials have shown that these attributes are perfectly applicable for 3-D joining applications. In particular the joining of only a few millimeters thick sheet metal structures of cold worked aluminum alloys, which should meet high quality requirements. Pre-processing tolerances, which could cause gaps of more than one millimeter, can be compensated in the welding process. This is possible by means of position-specific increased filler material addition.

Despite some parameters, the laser hybrid process is just partially investigated. There is a lack of studies on quality parameters, such as those that influence the root connection4). This decisive parameter for the strength of the joint is adjusted through the laser power only5,6). It is even postulated that an increase in welding depth cannot be influenced by the MIG contribution7).

Studies on process parameters such as welding speed or wire feed rate are available on a large scale3,4,8,9). They have even been carried out using artificial neural networks and genetic algorithms10). These investigations were carried out on butt, lap, and flanged joints. If processes are optimized for welding depths, it is done exclusively at blind seams11,12). Investigations of T-joints are available on a very small extent and require seam preparation13), or are produced by welding through the belt plate14).

Due to these conditions, this paper analyzes the penetration of fillet welds of the T-joint. Specifically, the effect of the MIG process on the root connection is investigated.

The setup is comprised of an articulated robot and a Fronius laser hybrid welding head, with an integrated Trans Puls Synergic welding system. The laser is an YPG fiber laser YLR-5000. The laser runs in front of the MIG torch, piercing with a 5° angle and the MIG welding head follows 3 mm behind the -1 mm focal position, piercing with a 35° angle. The Welding speed is 45 mm/s, a laser power of 4.1 kW and focal diameter of 220 µm. The sheets are out of AlMg3Mn, with a thickness of 3.5 mm and the filler wire is out of AlSi5. The gap between the plates of the T-joint is 0 mm.

The temporal characteristics of voltage and current of the MIG power source were recorded using a HKS measurement system. The system consists of a Weld QAS measuring computer, the WeldAnalyst software and the P1000 process sensor. The sampling frequency was selected with 52 kHz, whereby the measurement error is ±1%.

In order to determine the effect of the MIG arc length on the root connection A_W, the arc length is varied in the following. The arc length is essentially influenced by the voltage U_eff. However, the adjustment of the voltage U_eff mandates additional parameter adaption to ensure a controlled material transfer.

If the voltage in the welding process is changed, the current also adjusts. The filler wire will be partially melted in the base current phase and detached in the pulse current phase. If the base current increases, the pulse frequency also has to be increased to prevent the arc from moving up the wire too high in the pulse current phase. An unstable arc and filler wire drop, too large for a stable process would be the consequences. In addition, if the base current increases, it may also be necessary to increase the impulse current. This raises the detaching pinch force and thus the probability for a stable filler wire drop separation. Key figures of the used parameter set are summarized in Fig. 1. These figures are clustered through the auxiliary variable arc correction L_K.

Fig. 2 illustrates the measured values of the arc length and arc correction L_K. If the arc length is given in relation to the stickout, the resulting lengths are 35%, 40% and 45%. In the considered range of the arc correction L_K, both parameters change relatively by the same amount.

To analyze the effect of the arc correction L_K to weld seam properties, this was varied from L_K = -15% to L_K = -1%. Other process parameters are invariant. Cross sections of the weldings are presented in Fig. 3. These cross sections show with decreasing arc correction L_K an undercut on the weld seam surface evolving and increasing in size. Also the melt spill out at the weld root increases with decreasing arc correction L_K. This indicates a greater root connection A_W. This change represents also a decline in quality. However, this paper aims to provide a basic understanding of the interaction between MIG arc parameters and the root connection. This understanding should be used in comparable applications to optimize the weld seam quality regarding the assurance of the root connection.

For further investigations the seam width b_Seam and root connection A_W were measured on the cross sections of Fig. 3 and on additional samples. With increasing arc correction L_K, respectively the voltage, the seam width b_Seam and MIG power P_MIG increase too, as shown in Fig. 4. For a more profound understanding of the influence of these parameters on the root connection A_W, they will have to be related with the intensity of the MIG arc.

If an even distribution and circular intensity of the MIG arc is presumed, the arc intensity I_Arc can be determined by means of the MIG power P_MIG and the seam width b_Seam. This approximation is summarized in equation (1).

With the thermal efficiency η_th of the MIG process assumed to be 90%, the calculated intensity I_Arc is plotted over the arc correction L_K, as shown in Fig. 5. With decreasing arc correction L_K, the intensity I_Arc and the root connection A_W rise. The rise of both functions is not linear. Even though the parameter adjustment of the arc correction L_K lead to linear changes in seam width b_Seam and MIG power P_MIG. The reasons for the non-linearity is the quadratic consideration of the seam width b_Seam in equation (1).

The causal relationship between the functions is the change in heat flow. The increase of the intensity I_Arc leads to changes in heat transfer and the melting behavior. These changes in heat conduction influence the root connection A_W and the melt emerging at the weld root.

The scattering of the data points in Fig. 5 is high, which can be traced back to the process. For example, the determined measured values of the arc correction L_K = -11% have to be considered (see arrows in Fig. 5). Three measured values with high intensities I_Arc correlate with high root connection A_W. The same applies to two measured values with low intensities I_Arc and low root connection A_W.

Nonetheless, these results show that an increase of the root connection A_W of approximately 1.1 mm (+30%) can be achieved in the considered area of the arc correction L_K. This is done exclusively by adjusting the MIG parameters of the laser hybrid process.