Innovation in Welding in Terms of Processes, Weldability or Design Connections

Abstract

Innovation in welding when analysed from several aspects: In terms of processes, it can be seen on a daily basis, with brands presenting new concepts that have been implemented in new machines and equipment; regarding weldability, the new developments are implemented and depend on the acquisition of techniques. At the project level, innovation is much slower as it depends on checking the consistency of weldability and approval in the form of standards or codes. In this document, simple experimental studies are used, where some of the points mentioned above are put into practice, based on MIG and laser welding processes. The MIG variants have led to development in welding in terms of robustness and productivity. Meanwhile, laser welding has introduced new competitive welding solutions. The experimental component focused on welding steel, stainless steel and aluminium alloy as these are the most used and complement each other in the most varied structural and constructive solutions. Meanwhile, aluminium alloys have many advantages in uses where a good relationship between mechanical strength and weight is desired. The methods are based on simple visual or micrographic analyses and tensile and hardness tests. In the end, it is concluded that the welds meet the quality that would be necessary for the design requirements in butt joints.

1. Introduction

The greatest developments in welding processes have received great attention due their versatility, productivity and good weldability. This chapter focuses on two processes, of completely different nature, that simultaneously satisfy the four aforementioned requirements, one using an electric arc – MIG (Metal Inert Gas) –and the other using a laser – LBW (Laser Beam Welding). Among these processes, MIG is perhaps the most relevant since it’s also the most available, in terms of weldability of the main metal alloys used in the production of construction and stainless-steel parts as well as aluminium. On the other hand, what is most relevant in design is highlighted in simple welds, such as those used in the constitution of the samples from which the study specimens came. With this investigation, the aim is to demonstrate the potential of these processes, with the available equipment; pulsed MIG and LBW are used to test the weldability of steel, stainless steel, aluminium and mixed steel-stainless steel connections in cases of medium (3 and 6 mm) and thin thickness (1,3 and 2 mm).

2. Innovation in welding processes

At the level of industrial widespread use, MIG/MAG (or GMAW) with or without pulsed mode is required. Laser welding (LBW) is undergoing high development, even though with some problems associated that are yet be resolved. Some mixed cases, with dissimilar metals, present good weldability solutions.

2.1 Innovation in the MIG process

Progress in welding processes with filler material has occurred more in exploring the metal transfer mode. In the MIG/MAG process, the objective is to obtain an optimal metal transfer with minimal heat introduced into the parts. The evolution has been carried out through the following variants: CMT (Cold Metal Transfer, from Fronius) that detects a short circuit with digital process control and detaches the droplet by retracting the wire, during welding [1, 2]; SST (Surface Tension Transfer, from Lincoln Electric) that combines high-frequency inverter technology with advanced “Waveform Control Technology” in place of traditional short-arc GMAW welding [3]; RMD (Regulated Metal Deposition, from Miller) [4]; Wise solutions (terms WiseRoot, WiseThin and FastRoot, from Kemppi) [5]; the pulsed arc with developments by most welding machine manufacturers. The pulsed arc is based on the simplified rectangular wave (pulsed MIG, MIG-P or P-GMAW) [6], Figure 1, with peak intensity at a high value and average time as low as possible. The solution for the transfer of a droplet during the impulse is controlled by a synergistic mode [7, 8] and governed by the relationship (1), where T is the period, d is the droplet detachment constant and the constant n = 1.1 to 2.

Figure 1.

Scheme of a rectangular wave, where: - I_p, I_b and I_av – Peak, base and average current intensity values. - t_p, t_b – Peak and base times. - T = t_p + t_b - period.

$$I_p^n{t}_p\mathrm{T}=\mathrm{d}$$E1

The simplified relationship for calculating the average intensity value can be done by the relationship (2). This can be used to estimate the heat introduced into the part by pulsed MIG welding.

$$I_{\mathit{av}}=\frac{I_p{t}_p+I_b{t}_{\mathit{pb}}}{t_p+t_b}$$E2

In order to study the simple MIG-P, a rectangular wave with t_b greater than t_p and also with I_p significantly greater than I_b was used, to carry out welding with I_av, which generates a lower amount of heat introduced, Q_i, into the base metal under ideal conditions of metal transfer (3), where η, V and v_s are the process efficiency, current voltage and welding speed, respectively. There are several benefits in terms of metallurgical and operational weldability, obtaining better heat input values by (3), thermal cycle and thermal distribution, with good directional control over the weld pool, facilitating arc control electrical and the formation of fewer defects: porosity, spatter, distortion and burn-through.

$$Q_i=\mathrm{\eta }\frac{V{I}_{\mathit{av}}}{v_s}$$E3

The evolution of the pulsed arc has been significant, exploring current values and wave times in a fully controlled manner, focusing on new solutions to reduce I_av. Meanwhile, the introduction of the multiple pulsed cycle (MIG double pulse (or DP-GMAW); Figure 2a) uses combined waves that are repeated, also with their own period. A particular development of this is pulsed MIG with negative I_b intensities (calling it alternating, MIG-AC) (Figure 2b) [11] or many other possible combinations, all towards optimising metal transfer and less heat introduced in the parts.

Figure 2.

Multiple pulsed MIG: (a) double pulse schemes; (b) alternate pulse [9, 10].

The development of the pulsed arc continues to be carried out using other possible combinations, such as the dynamic response of wavelengths with current intensity and voltage, through the shape of the pulse, particularly the initial phase of its decay to improve detachment of gout [12]. There are several forms, such as the annealing pulse and the spike pulse (Figure 3a,b). Other types of waves are presented, such as the one shown in Figure 3c, where the way in which the pulse decay occurs allows the best optimizations of metal transfer [13].

Figure 3.

Schemes of controlled wave types: (a) annealing pulse, (b) spike pulse, (c) others.

2.2 Innovation at LBW

The progress of fusion connections without filler material has been seen more in the development of laser beams for use in welding (Figure 4). However, some use of filler metal has also been achieved. The use of the laser welding process (LBW) depends, in first instance, on the power and nature of the response to each material in terms of absorption of the laser beam, which defines its performance η. The heat introduced into the part is now given by (4), where P is the regulated laser power and v_s the welding speed.

Figure 4.

Relationship between the energy density of the laser that hits the surface of a part and the conduction and keyhole modes.

$$Q_i=\mathrm{\eta }\frac{P}{v_s}$$E4

When the Q_i heat is low or affects a larger area, the energy density is low, giving rise to the conduction mode. In this mode, while the surface is heated, heat is conducted through the body of the part. The conduction mode can normally be obtained with any laser and beam quality, allowing a very stable weld pool, without major defects (porosity, cracking, recess on the bead surface or splashes) [14]. As the power level increases or the incidence diameter decreases, the energy density increases considerably, giving rise to the keyhole, which is deeply penetrating, causing narrow HAZ and low distortion in the parts. This mode allows welding of a wider range of materials and thicknesses, higher speeds and better performance and productivity [15], but larger defects (Figure 4) [16].

When the electromagnetic wave hits an interface, one part is partially reflected, another partially absorbed and the other transmitted, with values R, A and T, respectively, their sum being given in (5), [17].

$$\mathrm{A}+\mathrm{T}+\mathrm{R}=1$$E5

The components of the Eq. (5) vary depending on the group of materials and for each material in particular. For example, transmission in metals is negligible, T ≈ 0, and so A = 1–R. The direct absorption of electromagnetic radiation by the metal surface is called Fresnel absorption, and so the laser beam penetrates in depth [18]. Figure 5 shows the absorption behaviour of some metals as a function of wavelength, where the positions of the CO₂ and Nd: YAG lasers are mentioned.

Figure 5.

Absorption of some metals by Nd:YAG laser beams (λ = 1064 μm) and CO₂ (λ ≈ 10 μm).

The pulsed laser, LBW-P, can be a good alternative to the non-pulsed one in many applications, but it requires some care in terms of introducing defects in the bead and controlling overlap, although it is not as complex as in MIG-P since it does not involve the control of metal transfer, as in the TIG process. Aiming at quality welding of materials, mainly metals and thermoplastics, new lasers are in full development. There is a wide variety of combinations of different types of sources – solid, gaseous, semiconductor and diode laser – with different wavelengths and energetic ranges available, from low to very high-power values [19]. In short, some are already on the market, others are still being developed in the laboratory and new ones will emerge with new applications associated. The quality of the beam is assessed by its dispersion and its diameter in the focusing region, measured by the BPP parameter (mm × mrad), which has evolved significantly. Versatility is also an objective to reach, such as the introduction into the market of laser welding processes for manual use, without or with addition metal (AM) (Figure 6). The improvement of new equipment and the lowering of its cost associated results in turning other welding processes, currently considered important, obsolete.

Figure 6.

Manual laser beam welding in butt and T joints.

2.3 New tendencies

A component of research for the development of welding takes place at the level of processes, execution and new trends that contribute to the more efficient execution of connections. Some welding machines already allow monitoring of changes that occur during the process based on efficient sensors and control algorithms, through automatic devices with adaptive control systems [20]. As a result, there is feedback that interferes with the welding process parameters [21]. The integration of artificial intelligence (AI) systems into self-learning machines in the welding area is being developed. In the MIG process, there have been experiments that include expert systems, fuzzy systems, artificial neural network systems, genetic algorithms and hybrid systems (e.g., neuro-fuzzy approaches) [9]. Furthermore, in laser welding, genetic algorithms have been experimented with optimising process parameters: power, welding speeds and AM feeding speeds to improve weldability and productivity [22]. Learning based on virtual reality (VR) is beginning to be used to train welders, more efficiently, in a safe and controlled environment and with the potential to reduce costs. This technology helps fill some gaps in welding learning, even without the presence of a trainer [23, 24]. Research is prepared to continue to evolve in the area of welding, driven by advances in materials science, automation and digital technologies.

3. Equipment, materials and methods

3.1 Equipment

To put into practice some of the aforementioned welding skills, the best equipment that was available was used. Its characteristics, in terms of process execution, are described in the following paragraphs. The MIG welding equipment is based on a SAFMIG 480 TRS source, with a maximum welding intensity of I = 450 A (Figure 7a). This machine, despite having a pulsed arc of the type shown in Figure 1, with synergistic control and the possibility of including robotization, does not include the developments previously mentioned. A motor with Arduino control was adapted for the linear movement of the fixed torch to a simple mechanical system controlled with linear movement and definition of the welding speed (Figure 7b).

Figure 7.

(a) SAFMIG 480 TRS machine and (b) mechanical torch movement system.

The LBW welding equipment is based on a disc laser source (brand Trumpf TruDisk 6602 [25]) whose main characteristics are: maximum available power of P = 6.6 kW, frequency λ = 1.063 μm and beam quality BPP = 8 mm × mrad. The beam is guided by ∅0.6 mm optical fibre to the optical focusing device. The focusing distance is 72 mm to the surface, point F = 0 mm, which can be defined as being below or above it. It is assisted by a Kuka KRC2 robot (Figure 8a), with a range of linear advance or welding speeds, v_s = 0,001 to 2 m/s [26], and a sample fixing template (Figure 8b). The equipment used does not yet have a device to introduce filler material.

Figure 8.

Equipment used in laser welding: (a) welding station with robot; (b) parts fixing template.

3.2 Materials used

The welded metals were of three types with the aim of experimenting with the difficulty of obtaining connections between them or in a mixed case of different metals. Without going into more complex experiments, connections of different thicknesses were also tried.

3.2.1 Construction steels

One of the base metals (BM) was hot-rolled and pickled steel DD11, numerical designation 1.0332 by EN 10111 and EN 1090–2 standards. The thickness used was 2 mm, with a thickness tolerance of 0.14 mm, according to EN 10051/EN 10029 and composition presented in Table 1 [27]. The main mechanical properties (yield stress (σ_y) and ultimate stress (σ_u)) are presented in Table 2. The other base metal is hot-rolled steel S235, standard designation in EN 10025–2: S 235 J2, with thickness t = 1.3 mm, thickness tolerance of 0.14 mm, in accordance with EN 10051/EN 10029 and chemical composition given in table DD11. For the thickness range in question, the σ_y and σ_u values are presented in Table 2 [28, 29].

	C	Si	Mn	P	S	N	Cu
DD11	0.12	0.4	0.6	0.045	0.045	—	—
S235	0.19	0.04	1.5	0.045	0.045	0.014	0.6

Table 1.

Chemical composition of DD11 and S235 steels (maximum values in %).

	σy (MPa)	σu (MPa)	ε (%)	Hardness
DD11	170 a 360	440	24	267 HB
S235	235	360	22	130 HV

Table 2.

Properties of BM and AM steels.

3.2.2 Stainless steel

Another base metal was AISI 316 (X5CrNi17–12-2 (1.4401): EN 10088–2-2005) and AISI 304 (X2CrNi19–11 (1.4301); EN 10088–2-2005) stainless steel’s compositions are shown in Table 3, in thicknesses of 3 and 6 mm. When selecting the additional metal, OK Autrod 347Si was chosen, whose chemical composition is presented in Table 3, with a wire diameter of ∅1.2 mm (Table 4).

	Cr	Ni	Mn	Mo	Si	C	Nb	P	N	S	Cu
AISI316	18.5	13.0	2.0	2.5	1.0	0.07	0.11	0.045	0.11	0.03	—
AISI304	18.5	9.0	1.5	—	1.0	0.07	—	0.045	—	0.03	—
347Si	19.0	9.8	1.7	0.1	0.7	0.04	0.6	—	—	—	0.1

Table 3.

Chemical composition of stainless steel (SS) and filler material [30], (maximum values in %).

	σy (MPa)	σu (MPa)	ε %	HV
AISI316	205	520	40	≤ 200
AISI304	205	520	40	≤ 200
347Si	440	640	37

Table 4.

Properties of BM and AM stainless steels [30].

The weldability of stainless steels, according to EN 1011–3, recommends the use of Schaeffler and WRC diagrams with an appropriate dilution value. This value is defined by the relationship between the volumes of base metal and the total that melt to form the bead. Maintaining the same length, the dilution value can be determined by Eq. (6), where A_BM1, A_BM2 and A_AM are the values of the base and filler metal areas, as can be seen in Figure 9, where (a) and (b) correspond to the strands with and without addition metal, respectively. If A_BM1 is equal to A_BM2 and A_AM = 0 in (b), in the case of a welding without filler material (LBW without filler material, for example), D = 1, and the points coincide in the Schaeffler diagram.

Figure 9.

Welding bead discrimination by fused areas with (a) and without (b) chamfer in joint preparation.

$$D=\frac{A_{\mathit{BM}1}+A_{\mathit{BM}2}}{A_{\mathit{BM}1}+A_{\mathit{BM}2}+A_{\mathit{AM}}}$$E6

The selection of the filler material for MIG-P welds was carried out using the Schaeffler diagram with overlapping risk zones (Figure 10) and the WRC diagram (Figure 11) taking into account that the dilution value with MIG is in the order of 20 at 30%.

Figure 10.

Schaeffler diagram: A and B results.

Figure 11.

WRC diagram: A result.

The results from the Schaeffler diagram allow us to conclude that the welding of AISI 316 with filler material “OK Autrod 347Si” presents the risk of hot cracking tendency in the bead material. With the WRC diagram, it is concluded that the microstructure of the bead is ideal-type austenitic-ferritic for welding stainless steels. In the study of welding between dissimilar materials, AISI 304 and S235, using OK Autrod 347Si filler material, the result of the final alloy of the bead in the Schaeffler diagram presents a slight risk of hot cracking above 1250°C, but this was not observed.

3.2.3 Aluminium

The principle of designating aluminium alloys is based on four numerical digits, in which the first distinguishes the predominant alloy element, in addition to aluminium, as shown in Figure 12.

Figure 12.

Aluminium alloys and weldability by fusion processes. Green – weldable. Red – not weldable.

The base metal adopted was the aluminium alloy EN AW6063 (standard EN 573 - Alloy designation system for wrought alloys), in the form of a 6- or 3-mm thick bar, whose chemical composition is presented in Table 5. After treatment, it becomes a weldable alloy, whose main mechanical properties are presented in Table 6. The weldability of a weldable aluminium alloy begins with the appropriate selection of the filler metal. Particularly in series alloys that contain Si or Mg, some care is necessary due to the presence of Mg and Si, which can form oxides that can cause cracking. To prevent these effects, careful selection of the filler material is necessary. The graph and table in Figure 13a,b were used to select the additional metal, which fell into ISO 18273-S Al 5356 (AWS A5.10: ER5356), with chemical composition (maximum values) presented in Table 6.

	Si	Mg	Fe	Mn	Cu	Zn	Ti	Cr	others
AW 6063	0.2 – 0.6	0.45 – 0.9	0.35	0.1	0.1	0.1	0.1	0.1	0.15
Al 5356 Si	0.25	5.5	0.4	0.2	0.1	0.1	0.2	0.2	0.15

Table 5.

Chemical composition of aluminium metals BM and AM (Wt%) [31].

Props.	Tf (°C)	σc (MPa)	σr (MPa)	HV	ε (%)
AW6063	600–655	160–210	195–255	76–74	9.8–10
Al5356	640	170 a 360	270	61.9	24%

Table 6.

Mechanical properties of Al6063 T6, [32, 33, 34, 35].

Figure 13.

Parts of graph (a) and table (b) used in the selection of the filler material (red lines). In a) depends on the % Mg and % Si of AW6063 (BM) and Al5356 (AM). (a) Cracking prevention chart valid for aluminium alloys containing Si or Mg. Adapted from version EAA 2015© with permission from European Aluminium association ([e-mail]) [31]. (b) Generic selection table for filler materials depending on the aluminium alloys to be welded and the characteristics to be obtained [36].

3.3 Methods and experimental work

Preventing the process that obtains total penetration is a fundamental condition of operative weldability to guarantee good connection strength, which depends on the joint preparation geometry (Figure 14) with adequate spacing (r) and flank (f) and with a chamfer of suitable angle (α) depending on the thickness of the parts to be welded.

Figure 14.

Generic diagrams of those that served as a basis for preparing joints for welding (a) – Without chamfers and (b) – With chamfers.

The permanent connections of S235 and DD11 sheets were studied in small thickness specimens, 1.3 and 2 mm, using pulsed MIG and compared with LBW. In MIG-P welding, the recommendations of EN 1090-1-1:2008 and the requirements of EN ISO 18273 were considered, for cleaning and preparing joints, parameters, electrodes, nozzle, and so on, with requirements of EN 1090–3. In this preparation of the joints, copper gaskets were used, and the positioning followed the scheme in Figure 14, with r = 1.5 mm for thickness t = 3 mm; r = 1 mm, f = 1,5 mm and chamfer α = 60° or α = 90° for t = 6 mm.

In the weldability of S235 steel, the preheating temperature estimated by the EN 1011–2 standard, with the welding parameters used (Table 7), gave an initial temperature of around 23°C. As this value was slightly higher than the ambient temperature, no special measures were taken. This estimate was obtained with η = 0.8 and consideration of the introduction of hydrogen HD = 6 ml/100 g. The welding acts to make the stainless steel and aluminium alloy specimens were carried out with the parameters set out in Table 7 and, otherwise, under the same conditions as in the case of S235.

AM	wire (mm)	Iav (A)	V (V)	vs (mm/min)	Ar (l/min)	t (mm)
SG2 550.5	∅1.2	48	10	300	15	1.3
347Si	∅1.2	200	25.4	294	20	6
Al 5356	∅1.0	142	11.6	250	20	3
		200	15.6	200	20	6

Table 7.

Main welding parameters used with MIG-P in the additional metals.

In LBW welding, all welds were performed without filler material, in gap-free butt joints, that is, type (a) in Figure 14 with r = 0, with the beam inclined at 15° backwards. Beads were obtained in full penetration keyhole mode, ensuring the expulsion of gases to the other side and preventing the formation of porosities. The study samples were obtained with the parameters shown in Table 8.

BM	t (mm)	P (kW)	F (mm)	vs (mm/s)	Ar (l/min)
S235	1.3	2.4	0	33	5
AISI 316	6	4.5	0	30	5
AISI 304	5	4.2	0	30	5
AW 6063	3	4	- 1	29.2	5
	6	6	- 1	29.2	5

Table 8.

Main welding parameters used with LBW.

The tensile tests were carried out on an Instrom 4206 machine (Figure 15a,b). The connection tests were carried out in order to monitor values of the force evolution until rupture and respective displacements.

Figure 15.

Tensile tests (a) were carried out on an Instrom 4206 machine (b).

Polishing was done with sandpaper to increase grain size. After chemical attack, some samples were screened under a Zeiss Axiotec microscope to check for possible internal defects. Microhardness values were obtained on a Shimadzu HMV microhardness meter with a Vickers indenter.

3.4 Design considerations in structural welds

There have also been developments at the project level, but these are of a different type, in the creation, updating and development of new standards and codes and, consequently, slower. These include normative structural design codes, for example, ex. to EN 1993–8 (EC3), and the steel construction standard, EN 1090–2. The aluminium alloy design following the EN 1999–1 standard [37] is more recent than the EC3 and presents some innovations or adaptations. Making a brief comparison of the essentials related to this experimental work, relating to the welding of the three types of metals under study. The design of welded connections of butt joints:

• in building steel structures, Eurocode 3 [37] does not consider the design relationships of butt joints. It is only necessary to guarantee conditions of good weldability: total penetration with complete fusion, the appropriate selection of the MA and the certified execution that guarantees resistance better than or equal to that of the MB.

• The design of stainless-steel structures generally follows Eurocode 3 [37], with some supplementary rules [38]. As with steel, when welding butt joints, it is only necessary to ensure good weldability. In the case of butt joints in aluminium alloy structures, according to Eurocode 9 [39], the normal stresses obey the relationships (7) and (8). For the normal stress perpendicular to the bead, σ_⊥,Ed, both in the weld bead and in the heat affected zone (HAZ), the normal stress perpendicular to the plane in the respective metal, in the HAZ, is σ_haz,Ed, where it is calculated by the relationship (8). The availability of these relations, (7) and (8), have a more realistic consideration, where f_w and f_u,haz are the characteristic resistance values of the filler metal and base metal in the zone affected by heat, respectively. The value γ_Mw is the partial factor for resistance of cross-sections in tension to fracture.

$${\sigma }_{\perp \mathit{Ed}}\le \frac{f_w}{{\gamma }_{\mathit{Mw}}}$$E7

$${\sigma }_{\mathit{haz},\mathit{Ed}}\le \frac{f_{u,\mathit{haz}}}{{\gamma }_{\mathit{Mw}}}$$E8

4. Results and discussions

In a first analysis of the welding seams, visual inspection made it possible to check the quality of weldability and the existence of surface and shape defects that could affect the results.

4.1 Resistance results

From the results of each tensile test, the evolution of the force and displacement F-ΔL was obtained, which, in first instance, allows the quality of the test to be verified. These results are not significant in a comparison process and are therefore omitted. However, the respective stress – strain results, (σ = F/A - ε = ΔL/L0) – is significant, allowing the analysis of the quality of the welds, by checking their replication and evaluating the resistance without effect of the size, that is, abstracting the dimensions of the specimens. These are presented in Figure 16.

Figure 16.

Comparison of resistance results (engineering stress σ versus engineering strain ε) of tensile specimens from the DD11 base material (black line) with those from MIG (a) and LBW (b).

The weld strength results on thin-thickness specimens of DD11 steel clarify the possibility of low-thickness connections with the MIG-P process (Figure 16a). Using LBW, it was necessary to apply welding with and without pulsed (Figure 16b).

Next, weld resistance results on specimens of AISI316 stainless steel, 6 mm thick, are presented, which provided good connection results, both with MIG-P and LBW. These are shown in Figure 17. As can be seen, the maximum resistance values are relatively close, but the behaviour, including plastic deformation, is different. From this observation, the specimens obtained by MIG-P fail through BM, in necking mode, while those welded by LBW broke abruptly through the bead but withstood the loading action well. It was an existence of AM that made the difference between behaviours.

Figure 17.

Comparison of resistance results (engineering stress “σ” versus engineering strain “ε”) of tensile specimens of the AISI 316 base material (black line) with those of MIG (a) and LBW (b).

As can be seen in Figure 17, the maximum resistance values are relatively close, but the behaviour, including plastic deformation, is different. From this observation, the specimens obtained by MIG-P broke through the BM in necking mode (Figure 17a), while those welded by LBW broke abruptly through the bead but withstood the loading action well (Figure 17b). Thus, we can conclude that the existence of AM resulted in the difference shown between behaviours.

In the visual comparative analysis and results of all connections between identical and dissimilar steel specimens, pairs of stainless steel and dissimilar structural steel were used, under the same conditions. The results, both with MIG-P and LBW, are very interesting and reflect the good quality of the connections, as is clear in Figure 18 using the MIG-P process and in Figure 19 with LBW.

Figure 18.

Results of the tensile study of dissimilar connections of BMs AISI 304 and S235 by MIG-P: (a) some specimens after tests; (b) resistance (σ -ε).

Figure 19.

Results of the tensile study of dissimilar bonds of AISI 316 and S235 by LBW: (a) some specimens after tests; (b) resistance (σ -ε).

The results portrayed in Figure 18a and Figure 19a show that all specimens of dissimilar metals gave way due to necking by the less resistant MB, which was assumed as long as the welds resisted, which they did. There was also the failure mode by necking in practically all cases of S235 steel connections. In the specimens made of stainless steel, failure began by necking in the HAZ, followed by rupture; the shape of the curves at the end of the tests demonstrates this in Figure 18b and Figure 19b. Thus, a dissimilar joint presents good tensile behaviour, as can be seen in Figure 20a, which is equal to or better than the sample of similar, less resistant metals. However, there is a significant change in the deformation ε_Av (Figure 20b). This effect can be confirmed in studies that present curves σ-ε of specimens of each metal only compared with the respective dissimilar, all with necking in the BM or HAZ region, like [38, 40].

Figure 20.

Average results of tensile study of dissimilar joints (Dissim) and stainless steel (SS) and steel S235 of: (a) resultant stresses (σ) (b) deformations (ε).

The results of the tensile tests relative to AW 6063 (black line) and specimens with connections produced in this alloy with welding by the MIG-P process are presented in Figure 21a) and by LBW in Figure 21b), with a thickness of 3 and 6 mm. The differences can then be compared.

Figure 21.

Comparison of resistance results (engineering stress σ versus engineering strain ε) of the tensile specimens of the base material AW 6063 (black line) with those of MIG (a) and LBW (b).

Although the resistance of all specimens is greater than that of the base material (black line in the graphs in Figure 21) and they yield due to stricture (Figure 21b), the maximum values of σ are lower when compared to MIG due to the behaviour of the MA allowing a greater deformation.

4.2 Hardness analysis across the joint

The hardness distributions are more or less in accordance with [32, 33, 34, 35]. By analysing the graphs in Figure 22a,b, in fusion welding, it is the LBW that presents less extensive HAZs, which was expected due to the less energetic process.

Figure 22.

Comparison of hardness distribution results in welds in AW 6063 by MIG-P (a) and by LBW (b).

The comparison of hardness results for welds at t = 6 mm: MIG and LBW allow conclusions regarding the shape and size of the weld bead, molten zone and the heat-affected zone, HAZ. The more pronounced difference seen at the bottom of the MIG-P bead is justified by its shape influenced by the joint preparation chamfer. In the case of LBW, the bead is only slightly narrower at the bottom, which is due to the dynamics in keyhole formation. The way in which hardness varies was partly due to the existence of filler material, in the case of MIG-P, and the width of the HAZ, in both.

The results of hardness distribution in the connection of dissimilar metals by LBW welding are presented in LBW (Figure 23), which has behaviour equivalent to that seen in [41]. The transition of HV values between the metal on the left (AISI316) and that of the bead increases by around 47% and from that to the right (S235) decreases by around 55%. The more pronounced variation on the left side is more worrying because it is due to a large microstructural variation.

Figure 23.

Distribution of hardness on dissimilar joint with AISI 316 (left) - S235 (right) by LBW welds.

4.3 Micrographic analysis

Micrographic analysis allows us to complement the study of the shape and dimensions of the welding beads and their respective HAZs. In a more exhaustive analysis, it also allows us to find defects, although there are other more suitable methods. In Figure 24a–c, montages of photographs taken under the microscope with 50 times magnification of the cross-section of equivalent beads from the study of weldability between dissimilar metals with the LBW process are presented, which presents less extensive HAZs, which was expected due to less energetic processes, as seen in [42]. In a first analysis of welding quality, visual inspection and the distribution of resistance and hardness results allow important conclusions to be drawn. Obviously, if there are significant internal defects, they will affect the results mentioned.

Figure 24.

Cross-sectional micrographs of dissimilar study LBW welding beads: - (a) S235-S235; − (b) S235- AISI304; − (c) AISI304- AISI304. Samples were etching in Nital (5 ml HNO₃ and 95 ml C₂H_6O) and marble (25 ml CuSO₄, 50 ml HCl and 25 ml H₂O).

4.4 Structural considerations

For the Al6063 alloy, with the precepts described in 3.4 and the values of f_w = 160 MPa and${\gamma }_{\mathit{Mw}}$ =1,25, we obtain from (7)${\mathrm{\sigma }}_{\perp \mathit{Ed}}\le 128\mathit{MPa}.$ With the value of f_u,haz = 110, MPa gives the value ${\mathrm{\sigma }}_{\mathit{haz},\mathit{Ed}}\le 88\mathit{MPa}$.

The representation of these values in Figure 21a shows that all the maximum values of the curves of the welded specimens are located above ${\mathrm{\sigma }}_{\mathit{haz},\mathit{Ed}}=88\mathit{MPa}$ and that only the welding specimen is completely below ${\mathrm{\sigma }}_{\perp \mathit{Ed}}\le 128\mathit{MPa}.$ This is in accordance with the fact that all the welded specimens yielded after being clamped by the base metals. In addition, in the case of welds without filler material, the value of ${\mathrm{\sigma }}_{\perp \mathit{Ed}}$ does not exist. In LBW welds without AM and with the necessary care, which includes the cases in Figure 21b, there is no rupture of the weld bead.

5. Conclusions

As expected, there are significant differences between the results of connections made by different processes, electric arc or laser, due to the respective energetic power, and the use or not of additional metal.

Quality MIG-P welds on stainless steel or aluminium alloy require suitable filler metal and parameters to obtain good weldability. With LBW, good welds were achieved, with narrower beads and HAZ and few defects, allowing good resistance in low or medium thickness parts, which are a consequence of the juxtaposition of the joint and the heat input to form the keyhole. The connection of dissimilar metals, between carbon and stainless steel, showed good results in terms of sample resistance and hardness. The σ-ε curve of specimens with dissimilar BMs shows behaviour at least equal to the corresponding weaker similar BMs, when they are compared under equal circumstances and failure mode beginning by necking. Furthermore, the perception of its role in the project is improved by comparing the resistance limits of aluminium alloy welding seams calculated by Eurocode 9 with the respective σ-ε curve.

New developments bring new opportunities. However, it is also necessary to disseminate good welding techniques and assimilate them to design standards. Furthermore, innovation also occurs in dissemination, on which, however insignificant it may seem, development heavily depends.

Acknowledgments

To the company IPROM for supporting the execution of laser welding.

To Eng. Nelson Santos for supporting the execution of MIG and workshop welding.

To the Research Centre in Digital Services (CISeD) and the Polytechnic University of Viseu for their support.