**3. The geometry of weld bead and the metallurgical changes in the weld metal of the GMAW process variants**

#### **3.1 CW-GMAW**

First, before dealing directly with metallurgy, there is a need to evaluate the geometry of the weld beads, considering that the first step to estimate whether a given process worked or not is to evaluate the geometric characteristics of the weld, and furthermore, to observe whether there is some apparent discontinuity. Thus, it will be possible to understand whether, in fact, the variant has validity and prospects for future development.

Several studies prove that the weld bead geometry is dependent on several factors such as current, voltage, polarity, welding speed, shielding gas, metal transfer mode, torch position, welding position, etc., that act directly on the weld pool and, consequently, provide a specific profile and dimensions of reinforcement, width, penetration, and dilution. This applies to the most varied types of arc welding, including GMAW, and it is still possible to predict this geometry according to the parameters used. Thus, the GMAW variants mentioned in this chapter also present their reinforcement convexity patterns and their respective characteristic dimensions of the cross sections, since the search for optimized welding parameters is always prioritized to propose a process stability standard used. Therefore, the authors [15] state that cold wires or hot wires must have smaller diameters than the electrode wire, since much larger cold wires can cause lack of fusion in the weld pool. Also, there is the possibility of simulating the bead profile using mathematical modeling of the dimensions in mm of penetration, dilution, height, and width of the bead, proposed by the same authors.

As an application of the CW-GMAW variant in welds for marine steel (ASTM A 131—different grades), the parameters of **Table 1** are generally used, all using wires of the AWS ER70S-6 class. The weld beads produced, most of the time, have a good surface finish and absence of discontinuities. Thus, the consolidation of the CW-GMAW variant provides a geometric pattern of the beads (**Figure 4**) when welded in the flat position in situations of simple deposition and V-bevels. GMAW

*A Brief Study of Unconventional Variants of GMAW Welding: Parameters, Weld Bead… DOI: http://dx.doi.org/10.5772/intechopen.104525*


**Table 1.**

*Standard welding parameter range for CW-GMAW.*

process weld and the CW-GMAW variant for three electrode wire feed values and the percentages of cold wire added. It can be seen that the gradual increase in the melting rate and the deposition rate with the insertion of the cold wire provides an increase

#### **Figure 4.**

*Standard profile of welded beads with the GMAW and CW-GMAW processes with percentages of 20%, 40%, 60%, 80%, and 100% of cold wire. Based on the work of [22].*

in convexity and a decrease in dilution, since the energy supplied to the part was converted to the melting of the wire additionally.

In this sense, the works of [9, 10, 22, 24] showed significant results regarding the CW-GMAW variants, the bead profiles are shown in **Figure 4**. All these authors welded steel sheets naval ASTM A 131 of different grades and dimensions of 9.5 mm X 150 mm X 300 V chamfer, 45° chamfer angle (bisel 22.5°), and the other parameters identical to **Table 1**, modifying the percentages used. Thus, using percentages of 20% and 40% with CW-GMAW, the authors [14] measured residual stress levels through X-ray diffraction and acoustic birefringence methods and using comparative analysis concluded that the variant helps to decrease the level of these stresses with the percentage increase of cold wire incorporated into the weld metal. **Figure 5** shows a specimen welded with CW-GMAW 40%, where it is observed that the sheet was clamped during welding to avoid distortion and measure residual stresses with restrictions. The geometric dimensions of the beads (in mm) were measured by determining the values of the width (w), penetration (p), height of the reinforcement (h), and the angle of wettability (α) of the weld metal in **Figure 6**. When comparing the measures presented, the conclusions can be reached: increase of the width of the bead and height of the reinforcement and the reduction of the penetration and of the HAZ.

Likewise, the results of [10] previously compared the fatigue strength of joints welded with GMAW and CW-GMAW, both in semiautomatic mode, showing the versatility in welding ASTM A131 grade A naval steel sheets when using wires with diameters varying from: 1.2 mm as electrode wire and cold wires of 0.8 mm and 1.0 mm, with cold wire feed rates of 50%. The dimensions of the sample body were the same (9.5 mm X 150 mm X 300 mm), requiring two passes for the total filling of the chamfer, the first pass was the one from scratch applied with GMAW to all parts, the second finishing pass, this being the comparative parameter between the GMAW process and the CW-GMAW. It was concluded that the fatigue behavior of the joints welded by the GMAW-CW process in both conditions is practically the same when compared with the conventional GMAW process. In addition, some metallurgical considerations were observed, such as the decrease of HAZ in the coarse-grained region and the formation of the fraction of primary ferrite and Widmansttäten, which influences the increase in hardness in the region, this occurs proportional to the diameter of the cold wire added, as the more cold wire, the greater the change.

**Figure 5.** *Specimen welded with CW-GMAW 40% with clampers for measure residual stresses.* *A Brief Study of Unconventional Variants of GMAW Welding: Parameters, Weld Bead… DOI: http://dx.doi.org/10.5772/intechopen.104525*

#### **Figure 6.**

*The geometric dimensions of the beads of specimens welded with GMAW and CW-GMAW with clampers, 20% and 40% of cold wire.*

This demonstrates that the addition of cold wire may be affecting the cooling rate due to the lower energy imposed on the weld pool.

However, works [22, 23] comparatively studied the GMAW process with the two variants, CW-GMAW and DCW-GMAW, with 03 wire feed speeds of 10 m/ min, 12 m/min, and 14 m/min varying the percentages of 20%, 40%, 60%, 80%, and 100%. Thus, [22] studied the stability of both processes capturing the oscillograms, melting rates, deposition rates, and the geometry of the weld beads (**Figure 4**), where the following results were highlighted comparing the GMAW and the CW-GMAW:


Continuing, using the IIW (International Institute of Welding) C-Mn metal microstructure classification scheme [30], together with the measurement of the volumetric fraction of phases using images obtained by optical microscopy, both

**Figure 7.** *Weld metal grain size: (a) GMAW, (b) CW-GMAW, and (c)DCW-GMAW. Nital solution (2%), optical microscope [17].*

*A Brief Study of Unconventional Variants of GMAW Welding: Parameters, Weld Bead… DOI: http://dx.doi.org/10.5772/intechopen.104525*

works mentioned above studied the influence of cold wire on the formation of the main microconstituents of the weld, mainly acicular ferrite (AF), since this microstructure has a desirable presence in the weld metal, due to its excellent mechanical properties. However, in addition to this, the presence of allotriomorphic ferrite and Widmansttäten ferrite is also part of the predominant phases in carbon steel welds as deposited [31]. Thus, [22, 24] found that in all parts the weld metal is formed by ferrite, in several different forms: primary ferrite (PF), grain boundary ferrite—PF(G), acicular ferrite (AF), intragranular polygonal ferrite—PF(I), non-aligned secondphase ferrite—FS(NA), and aligned second-phase ferrite—FS(A). PF, PF(G), and FS(NA) ferrites predominate in the composition of the microstructures present with almost 100% of the composition for the highest wire feed speeds.

However, for the three wire feed speeds, with up to 60% cold wire, there is an average acicular ferrite increase of around 24% compared with GMAW. For percentages of cold wire of 80% and 100%, an inverse behavior is observed with a decrease in the amount of AF, on average, of 36%. What can possibly be observed is that the microstructures are benefited or inhibited by the presence of certain chemical elements. As, for example, the presence of low and medium percentages of aluminum (Al) forms a TiO layer around the inclusions, where the circular ferrite is nucleated, thus favoring its growth. However, for high Al content, such formation does not occur [32]. Overall, the gradual increase in the insertion of cold wire improved the mechanical properties of hardness.

#### **3.2 DCW-GMAW**

The DCW-GMAW came from the idea of the other variant CW-GMAW to evaluate the ability of how much the GMAW process was able to increase the insertion of "cold" mass using only one electrode wire. One of the great challenges of this variant is the placement of the cold wires. Therefore, the profiles of the evaluated beads are based considering the entry of the cold wires in the angular position in relation to the welding torch, as shown in **Figure 2a**. Still not having significant results, however, [23] concluded that high percentages of cold wire, from 60%, cause a reduction of approximately 15% in the hardness properties of the weld metal. However, before that, it is necessary to deal with the geometry of the weld beads.

Based only on practical works, the DCW-GMAW variant was first tested and patented by [22] and soon after, also analyzed by [24], where in general it was observed that this process was capable of being applied in the industry in fact using the data obtained for this conclusion. Working with carbon steel, parameters similar to **Table 1**, percentages from 20–100%, with a variation of 20% and with wire feed speeds of 10 m/min, 12 m/min, and 14 m/min. **Figure 9** presents the standard profiles found for the weld beads based on [22], the summary below describes some significant results such as:


In the case of the predominant microstructures, the phases in the forms of PF, PF(G) and FS(NA) ferrites constitute 98% of the composition of the microstructures present in the weld metal. What changes are the amount of each phase in each image analyzed? For low amounts of cold wire (20% and 40%) the percentages of AF and FS(NA) have an average increase of 47% and 28%, respectively. However, for percentages of cold wire from DCW-GMAW-60%, there is a decrease in the amount of these phases, while the FS(A) increases, even tripling its composition in the case of low wire feed with 100% of cold wire. An image of the phases present in a sample of DCW-GMAW-60% for the feed speed of 12 m/min can be seen in **Figure 8c**. The work of [24] still shows that the silicon levels are drastically high, in the CW-GMAW and DCW-GMAW variants, thus also increasing the weld metal hardness levels. From average values of 155 HV to peaks of up to 190 HV.

In the case of the predominant microstructures, the phases in the forms of PF, PF(G), and FS(NA) ferrites constitute 98% of the composition of the microstructures present in the weld metal. What changes are the amount of each phase in each image analyzed? For low amounts of cold wire (20% and 40%), the percentages of AF and FS(NA) have an average increase of 47% and 28%, respectively. However, for percentages of cold wire from DCW-GMAW-60%, there is a decrease in the amount of these phases, while the FS(A) increases, even tripling its composition in the case of low wire feed with 100% of cold wire. An image of the phases present in a sample of DCW-GMAW-60% for the feed speed of 12 m/min can be seen in **Figure 8c**. The work of [24] still shows that the silicon levels are drastically high, in the CW-GMAW and DCW-GMAW variants, thus also increasing the weld metal hardness levels. From average values of 155 HV to peaks of up to 190 HV.

*A Brief Study of Unconventional Variants of GMAW Welding: Parameters, Weld Bead… DOI: http://dx.doi.org/10.5772/intechopen.104525*

#### **3.3 HW-GMAW**

Thinking about increasing the melting rate and the deposition rate of the GMAW process, making an adaptation in the CW-GMAW variant, the design of the HW-GMAW was arrived at. That is, the additional wire, which was previously free of energy, now has a low direct current to assist in the fusion of the filler metal, using the Joule effect as a basic principle, which, through the resistance of the metal, converts electrical energy into thermal energy. Bearing in mind that when introducing hot wire into the process, it is not intended to significantly increase the heat imposed on the part, but only to increase productivity with the design of this variant of GMAW.

The proposition of the HW-GMAW variant is relatively new, despite the similarity with more consolidated and widely used processes in the industry in general. Currently, his studies focus on the application of hard coatings on surfaces to increase wear resistance [27, 28, 33].

Firstly, in terms of welding itself, the works by [11, 34] studied the influence of generic parameters such as: welding direction, hot wire feed rates. However, remembering that the extra wire feed rates obey the ratio given according to Eq. (1), both of which are related having as a reference point a percentage of the electrode wire feed speed (m/min). Thus, [27] using 5 m/min electrode wire speed with a percentage of 140% hot wire, casing welds were performed on flat bars (9.5 mm X 56 mm X

#### **Figure 9.**

*Standard profile of welded beads with the GMAW and DCW-GMAW processes with percentages of 20%, 40%, 60%, 80%, and 100% of cold wire. Based on the work of [22].*

225 mm) of AISI/SAE 1020 carbon steel, both wires used were of the AWS ER70S-6 class, the electrode wire having a diameter of 1.2 mm and the energized wire having a diameter of 1.0 mm. The welding parameters were a voltage of 23.6 V, current of 180 A and a contact tip-to-work distance of 15 mm, in the pulling welding technique. The variable parameter used was the direct polarity current of the hot wire at levels of 40 A, 80 A, 120 A, and 150 A. The solder used as a comparison was the CW-GMAW with 50% cold wire and parameters almost identical to those mentioned previously. The results obtained suggest that the w/h ratio has values above the previously established limit, greater than 0.3, with an average of 0.35. They may not be ideal for chamfering, but excellent for application as a coating. Penetration is slightly higher with values of up to 30% higher and the HAZ practically remains very similar. In general, the bead profiles are similar to those in **Figure 10**, in which they are based on the work of [11], in which the interference of other parameters such as the stability of the process through cyclograms, the polarity of the hot wire (on both poles: positive and negative), the welding direction (pull or push) varying wire feed rates at 20% and 100% hot wire. Emphasizing that the same material of low carbon steel and electrode wire of the same AWS class were used.

*A Brief Study of Unconventional Variants of GMAW Welding: Parameters, Weld Bead… DOI: http://dx.doi.org/10.5772/intechopen.104525*

#### **Figure 10.**

*Standard profile of welded beads with the HW-GMAW processes with percentages of 20% and 100% of hot wire in the torch movements: (a) pull and (b) push. Based on the work of [11].*

Still in the work of [11], the authors concluded that, for high wire feed rates, the penetration can drop by up to 45% and the dilution by up to 25%, when compared with the original GMAW. It has also been shown that hot wire polarity can attract or repel the arc and, together with the HW feed rate, can change bead geometry through changes in penetration depth and bead height. On the other hand, welding directions and wire feed rates are the parameters that most affect arc stability. And finally, in most weld beads the penetration is lower than the same weld in the conventional process.

However, regarding the metallurgical issues of grain size and the formation of microstructures from HW-GMAW welding in carbon steel materials, it will be necessary to continue the research, since it has not yet been published. Noting that there are already many works that show the structures of alloys based on Ni and FeCrC, which will not be addressed in this chapter.
