The evolution of the electronic control of welding power sources has made it possible to develop several variations or modalities of the gas metal arc welding (GMAW) process in recent decades. The simplest and probably most widespread example is pulsed GMAW. This basically consists of imposing a current signal that alternates between a lower current level sufficient to maintain the arc and keep the end of the electrode in the liquid state and a higher level aimed at the formation and detachment of a metal drop. This approach provides greater versatility of the process, overcoming the limitations imposed by conventional GMAW welding power sources, especially with respect to metal transfer modes (short-circuit, globular and spray). In short, GMAW with current pulsing allows metal transfer without contact between the wire and the workpiece (spray), but with an average current typical of metal transfer with contact between the wire and the workpiece (short-circuit or globular) [1]. The most cited advantages are the possibility of welding out of the flat position, spatter reduction, greater penetration, and a weld bead with a more adequate geometry, or better wettability, especially when welding materials such as stainless steels and aluminum alloys [2].
In recent years, there has been a profusion of GMAW process modalities developed by various manufacturers (e.g., STT, RMD, ColdArc, Cold MIG, CMT and MicroMIG). Kah et al. [3] give a detailed description of these and other modalities, where the authors argue that although the innovations are designed to meet industry demands they are not always used to their full potential due to a lack of deeper understanding. One such process is variable polarity GMAW, for which there is no consensus as to the most appropriate designation and the term AC-GMAW (alternating current - gas metal arc welding) is used in this article. The aim of this process is to combine the desirable characteristics of pulsed GMAW (controlled metal transfer) and of negative polarity GMAW (direct current electrode negative – DCEN). The most widespread advantage of DCEN polarity is the higher melting rate of the electrode wire and lower heat input to the workpiece. Although these characteristics are well documented, they are still subject of discussion and are assessed herein.
The DCEN welding process has been described as a reversal in the balance of energy that is transferred to the electrode and the workpiece [4–5]. More specifically, in DCEN polarity, it is generally reported that 70% of the arc energy is transferred to the electrode and the rest to the workpiece. This feature may be desirable from a productivity standpoint, however, conventional GMAW in the DCEN condition is considered to be very unstable, subject to excessive spatter formation, with droplet repulsion, low penetration and low wettability weld beads, along with being limited to the globular metal transfer mode [6]. Souza et al. [7] demonstrated that with the use of a suitable gas mixture (98% Ar + 2% O2), a stable transfer without droplet repulsion and with low spatter is obtained for steel wires. However, low wettability persists. Furthermore, was employed two average current values (150 A and 250 A) and observed that the metal transfer remains globular at 150 A [7]. Thus, it appears that the stable DCEN condition is limited to high average currents, for example, above 250 A, which in practice makes it impossible to weld thin sheets or perform welds in out-of-flat positions.
However, disregarding the energy balance, the higher melting rate in the DCEN condition is generally attributed to the phenomenon of arc climbing in the solid extension of the wire electrode [8]. This behavior is characterized by the action of the arc over much of the solid extension of the electrode, which causes a kind of preheating and this is responsible for the higher melting rate compared to the same process operating only in the direct current electrode positive (DCEP) mode under the same conditions. This is mainly due to the arc ‘seeking’ the oxide layer present on the surface of the electrode, which is known to facilitate electron emission and arc maintenance. Yarmuch e Patchett [9] cite that the predominant type of emission is cathodic emission, a phenomenon that produces more energy at the cathode, thus increasing the melting rate. Souza et al. [7] noted that the higher melting rate in DCEN is due to the higher thermal efficiency of the arc over the area of the electrode as it ‘seeks’ oxides for cathodic emission, rather than the heat generated in the arc and wire connection.
1.1 AC-GMAW waveforms and EN ratio
The expansion in the number of new modalities of the arc welding process is driven by the possible combinations of different transfer modes, for example, welding programs that impose a certain number of transfer cycles using controlled short-circuit interspersed with cycles of free flight transfers typical of a pulsed waveform. It is also common for each of the major equipment manufacturers to name their waveform using an acronym or a name with good commercial impact. In addition to the distinction between the waveforms, it is possible to adopt different control strategies to change the way the metal drop is transferred from the wire to the molten pool, thus offering different options for each type of material and application. Inserted in this context, the AC-GMAW, process based on the control of the current waveform, was conceived to take advantage of the two polarities, positive (DCEP) and negative (DCEN). Initially, this modality was presented as a solution for thin plate welding. Tong et al. [10] and Rosa et al. [11] showed the ability of the process to fill the root in overlapping joints formed by aluminum sheets, when there is an irregularity or a large gap between them, and they used the term gap bridging ability. Harwig et al. [12] explored this same ability, however, for carbon steel-base materials. Joseph [13] showed that the process can be applied in a dissimilar welding process using galvanized steel with aluminum alloys. There are also studies that consider the application of AC-GMAW to thick plates Arif e Chung [14] and for single-pass butt-joint welding of structural steel AH36 [15], filler welding in butt joints with an aluminum "V" groove for the marine industry [15] and cladding of boiler walls with nickel alloy [8]. Kang et al. [17] used AC in GMAW Tandem welding with two wires, to mitigate the harmful action of the magnetic field generated between the two arcs, stabilizing the metal transfer and increasing productivity. The versatility of this process allows for studies in the area of advanced manufacturing that investigate the benefits of AC-GMAW, focusing on the adjustment of welding parameters based on real-time sensing and feedback control [18]. In addition, the reduced heat input that AC-GMAW offers compared to DC-GMAW, contributes to the control of the microstructure formed and allows for dissimilar welding between aluminum and hot-dip galvanized steel, which promotes evaporation of zinc (Zn) layer and formation of brittle Fe-Al intermetallic compound (IMC) layer can affect the joint quality both thermally and metallurgically [19].
However, most applications use the traditional AC-GMAW waveform, which incorporates only a current threshold in negative polarity, bounded by the parameters of negative current (In) and negative time (tn) (Fig. 1.a). In general, the aim of the negative phase is to increase the melting rate and reduce heat input to the workpiece without compromising the process stability, since droplet detachment occurs in a controlled manner during the positive stage. Another approach to this mode features a more complex waveform. This was firstly proposed by Japanese companies that incorporated an additional negative current pulse (Ip–) into the negative polarity together with the base threshold (Ib–) found in traditional AC-GMAW waveforms. Arif and Chung [20] stated that this approach is primarily aimed at further increasing the wire electrode melting rate. Jaskulski [21], on the other hand, states that the negative pulse helps bringing the arc back to an acceptable spatial stability during the inherently unstable EN phase. In any case, this advanced waveform (Fig. 1.b) is used only for steels, while the traditional AC-GMAW waveform (Fig. 1.a), with only a plateau on the negative side, is still used by the same manufacturers for aluminum alloys [22]. Both shapes in Fig. 1 were used in the study reported herein.
To support the discussions in this paper, a schematic drawing of the complex waveform was prepared and is shown in Fig. 2. A legend has been inserted in the Table 1 to show the meanings of the various parameters and thresholds of the current waveform used in this study.
Table 1
Meaning of the complex waveform thresholds corresponding to Fig. 1.
t +: | Time current remains in positive polarity | tp+: | Duration of the positive current pulse |
t –: | Time current remains in negative polarity | tb+: | Duration of the positive current base |
tc: | Duration of one current cycle | tp–: | Duration of the negative current pulse |
tic: | Initial time of one current cycle | tb–: | Duration of the negative current base |
tfc: | Final time of one current cycle | Ip+ | Current Intensity of the positive pulse |
| | Ib+ | Current Intensity of the positive base |
| | Ip– | Current Intensity of the negative pulse |
| | Ib– | Current Intensity of the negative base |
Following the development of modern synergic programs, despite the considerable number of parameters involved, welding equipment operators do not necessarily need an in-depth understanding of the role of each parameter. The study and development of parameterization is not within the scope of this study, but this is addressed in, for instance, [8–23–24]. However, although there are some similarities, it should be noted that the cited authors consider totally rectangular waveforms. Also, the melting rate needs to be proportional in the two polarities in order to obtain a constant arc length and the determination of the negative electrode percentage (%EN) is based on the negative time (tn). This is useful for setting parameters, but the %EN may be insufficient for evaluating the effects of negative polarity on the weld bead and the piece to be welded in advance. This is due to the fact that the %EN often only takes into account times and does not consider current values (In). However, in some cases, the performance of the negative electrode can be represented under different strategies and this merits further attention to estimate the effects of negative polarity on the welding process. Tong et al.[10] and Rosa et al. [11] also considered the current values and adopt this approach when defining what is called the EN ratio. In a simplified way, and considering a rectangular waveform, Harwig et al. [12] and Park et al. [25] used Eq. 1 to estimate more accurately the EN ratio.
\(\mathbf{E}\mathbf{N} \mathbf{R}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{o}= \frac{({\mathbf{I}}_{\mathbf{n} }.{\mathbf{t}}_{\mathbf{n}})}{{(\mathbf{I}}_{\mathbf{p} }.{\mathbf{t}}_{\mathbf{p}})+{(\mathbf{I}}_{\mathbf{b} }{.\mathbf{t}}_{\mathbf{b}})+{(\mathbf{I}}_{\mathbf{n} }{.\mathbf{t}}_{\mathbf{n}})}\) | (Eq. 1) |
Basically, Eq. 1 represents the ratio between the negative phase area (DCEN) and the total area enclosed by the entire welding current signal in one cycle (DCEN + DCEP), as shown in Fig. 2. It is important to highlight that because the electrical signals of the process with alternating current vary between positive and negative polarity, it is necessary to use absolute values so that the different signals are rectified and can be summed to give an overall average. However, even calculated in absolute values, the results for average voltage and average current may present some limitations in relation to the calculation of the power dissipated in the conductor with the application of non-symmetric waveforms and the average signal value will still be inaccurate. Thus, in this study, we employed (and highlight the importance of) the true root mean square (or true RMS) value, which is used when it is necessary to measure alternating current with waveforms that are not necessarily constant over time. Thus, the measuring device performs a mapping of the wave, point by point, marking these points and calculating the average at that particular instant, since the current and voltage were captured at an infinite number of instants in a given period of time. Eq. 2 shows the calculation performed by the electrical signal acquisition system and this was used to calculate both the current and voltage values. The values for these electrical signals are represented by XRMS, the period in each cycle is T as a function of time (t).
\({\mathbf{X}}_{\mathbf{R}\mathbf{M}\mathbf{S}}=\sqrt{\frac{1}{\mathbf{T}}{\int }_{0}^{\mathbf{T}}{\mathbf{X}}^{2}\left(\mathbf{t}\right).\mathbf{d}\mathbf{t}}\) | (Eq. 2) |
Furthermore, considering the representation in Fig. 2, Equations 3 and 4 show mathematical representations of the calculation of the average current values for the positive and negative polarities, respectively. These calculations can be used to verify, in isolation, the effect of the welding current on only one polarity if needed.
\(\mathbf{D}\mathbf{C}\mathbf{E}\mathbf{P}={\int }_{\mathbf{t}\mathbf{i}\mathbf{c}}^{\mathbf{t}\mathbf{f}\mathbf{c}}\mathbf{I}>0\mathbf{d}\mathbf{t}\) | (Eq. 3) |
\(\mathbf{D}\mathbf{C}\mathbf{E}\mathbf{N}={\int }_{\mathbf{t}\mathbf{i}\mathbf{c}}^{\mathbf{t}\mathbf{f}\mathbf{c}}\mathbf{I}<0\mathbf{d}\mathbf{t}\) | (Eq. 4) |
Thus, if the percentage of negative polarity is calculated based on the DCEP and DCEN values, from Equations 3 and 4, respectively, it is possible to obtain the appropriate EN ratio, through Eq. 5.
\(\mathbf{E}\mathbf{N} \mathbf{R}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{o}=\frac{\mathbf{D}\mathbf{C}\mathbf{E}\mathbf{N}}{\mathbf{D}\mathbf{C}\mathbf{E}\mathbf{N}+\mathbf{D}\mathbf{C}\mathbf{E}\mathbf{P}}\) | (Eq. 5) |
Therefore, within the context of technological advances as well as the scarcity of visual results in the literature, an analysis and discussion of the welding process, focused on the effects of negative polarity in the wire electrode are provided herein and constitute the objective of the present work. The practical aspects related to the use of the EN ratio in AC-GMAW are also described. This discussion is supported by analysis of macrographs, the electrical parameters of welding and high-speed footage, as well as infrared thermal monitoring of the workpiece. The waveform used in modern synergic systems for welding stainless steels (Fig. 2.b) is evaluated, especially with regard to the effect of using an additional pulse in the negative polarity of the process and, consequently, the influence on the resulting weld. In addition, the performance obtained with different percentages of negative electrode in this complex waveform is also explored.