Influence of Preheating Temperature on Changes in Properties in the HAZ during Multipass MIG Welding of Alloy AW 6061 and Possibilities of Their Restoration

Abstract

Fusion welding of heat-treatable aluminum alloys is generally accompanied by a significant decrease in mechanical properties in the HAZ caused by the dissolution of the hardening phase. The intensity of this decrease in mechanical properties can be reduced by limiting the heat input value. However, this approach is in direct conflict with the principles for welding aluminum and its alloys. Due to the very high thermal conductivity of aluminum alloys, it is necessary to use preheating for thicknesses larger than 5 mm to eliminate non-penetration and cold joints. This paper aims to show the influence of multiple temperature cycles, performed at different preheating temperatures, on changes in the microstructure and mechanical properties. At the same time, the extent to which the original properties of the material can be restored by natural and artificial aging at 160, 175 and 190 °C is also investigated.

1. Introduction

Heat-treatable aluminum alloys are some of the most progressive engineering materials and are used in many industries. Their main advantages are their low specific weight and good corrosion resistance, combined with their very good mechanical properties. At the same time, these are materials in which the hardening phases are dissolved at high temperatures, meaning the strength properties are significantly reduced. Therefore, most of the papers in this field are focused on finding methods to eliminate these disadvantages or finding methods to restore the strength of welded joints via heat treatment, as evidenced by the considerable number of papers published recently and dealing with specific types of heat-treatable alloys.

The studies carried out on alloy AW 6061 can be divided into several areas. Most studies and papers have been devoted to the fusion welding effects on the changes in mechanical properties. Vargas et al. [1] assessed the effects of the heat input value in GMAW welding for a single pass weld measuring 4.8 mm in thickness. They found that for different heat input values, decreases in mechanical properties varying from 38% to 46% were observed. Nie et al. [2] investigated the effects of pulsed MIG welding on the microstructural and mechanical property changes in a heterogeneous AW 6061/A356 joint. It was found that minimal microhardness appeared in the HAZ of alloy A356 and the specimens fractured under mixed failure during tensile testing, mainly in the partially fused zone of alloy A356. Kumar and Sundarrajan [3] investigated the process parameters affected in multipass pulsed TIG welding and the effects on the mechanical properties. They observed an inverse relationship between the notch tensile strength and impact toughness and also assessed the microstructures in different zones of the joint. Another special case was a study comparing the properties of joints formed by fusion welding and welding under the solidus temperature. Kulekci et al. [4] compared the properties of a joint treated using the MIG method and FSW method. They came to the expected conclusion that better mechanical and fatigue properties were obtained with the FSW method. A similar comparison was also made by Gencer et al. [5] and Jannet et al. [6], with a similar result to that found by Kuleci.

The possibility of using a hybrid fusion welding laser–TIG method and the effects on the mechanical properties have been discussed by Wang et al. [7]. The weld was projected as a single pass butt weld on a 3-mm-thick sheet, which resulted in a 31% decrease in mechanical properties compared to the base material.

Authors such as Guzman et al. [8], Peng et al. [9], Hilty [10], Wang et al. [7] and Gencer et al. [5], in addition to assessing the effects of degradation on the mechanical properties during welding, have also dealt with the possibility of restoring the mechanical properties of welded joints via heat treatment. Guzman [8] assessed the mechanical properties of welded joints made by MIG welding with two types of filler materials (ER 4043 and ER 5356) and with subsequent heat treatment under T4 and T6 conditions. The best results were obtained after welding with the filler material ER 4043, whereby ultimate tensile strength (UTS) values of 153 MPa were achieved identically after both T4 and T6 heat treatment. Peng [9], on a single-sided butt weld made by TIG welding, showed that with an increase in heat input, the UTS values of the welded joints after post-weld aging treatment improved from 142 to 240 MPa. Wang [7] was able to increase the mechanical properties of welded joints from an initial value of 69% up to 84% of the base material after applying the heat treatment. The UTS value of the joint was 284 MPa and Gencer [5] achieved a joint strength of 83% of the base material after applying T6 heat treatment using the FSW method.

A separate but very important area involves studies [11,12,13,14] on the weld sensitivity to hot cracking. These generally occur when the welding process involves filler materials made from 6000 and 7000 series aluminum alloys [11,14], and of course also when parts are fabricated using additive methods such as WAAM (wire arc additive manufacturing) [11,15]. For example, Soysal and Kou [14] performed hot crack predictions using Pandat software, based on experiments performed using a lower moveable plate. The experiment was designed by the authors themselves. Then, Zhang et al. [13] studied the sensitivity of alloy AW 6061 to hot cracking when applying a double-sided arc welding process. Huang and Kou [12] then assessed the sensitivity of alloy AW 6061 to hot cracking using the circular patch test.

A special group of studies describes the changes in the mechanical properties of alloy AW 6061 with different specific joining methods or other thermal influences on the material. Wittman and Lipkin [16] investigated the effects of heat treatment on alloy AW 6061 joints made by explosion welding. Alternatively, Demir and Gündüz [17] assessed the effects of aging on the machinability of AW 6061 aluminum alloy.

As is evident from the literature, many different types of studies have been carried out on alloy AW 6061. However, only Kumar and Sundarrajan [3] have assessed the effects of multipass welding on the mechanical properties when welding on thicknesses larger than 5 mm. However, none of the above authors have assessed how the preheating temperature can affect the resulting mechanical properties of the joint, even though in industrial production (especially in transportation) profiles with thicknesses generally larger than 10 mm are welded, at which point it is already necessary to preheat the materials to be welded to avoid cold joints and non-penetration. The present paper, therefore, focuses directly on the above aspects affecting joint strength in multipass welding.

2. Materials and Methods

EN AW 6061 (AlMg1SiCu) aluminum alloy was used in the experiment. It is a heat-treatable aluminum–magnesium–silicon alloy that is capable of age-hardening due to the temperature-dependent change in the solubility of the intermetallic phase of the Mg₂Si in aluminum. This material has very good hot and cold formability (beyond temper condition T6—condition after solution annealing and artificial aging) and very good weldability using almost all methods. Moreover, it also has good resistance to atmospheric corrosion, even in marine climate conditions or in heavy industrial production atmospheres. Alloy EN AW 6061 is mainly used in the production of mechanically loaded welded structures exposed to atmospheric influences in general engineering, in the manufacture of piping and pipe assemblies, in bridge construction and also in the manufacture of road and rail vehicles and ships [18]. As it is suitable for surface treatment by anodic oxidation, it can be also used in the manufacture of load-bearing elements of interior furnishings and furniture.

The material was supplied in the form of 15-mm-thick slabs under T651 conditions (condition after solution annealing, internal stress relief controlled by magnitude of deformation and artificial aging). The chemical composition of the used alloy according to EN 573-3 [19] was determined using a Q4 Tasman spectrometer (Bruker Elemental GmbH, Karlsruhe, Germany) and is shown in

Table 1. Chemical composition of alloy AW 6061 (wt%).

Element	Si	Fe	Cu	Mn	Mg	Cr	Ni	Zn	Ti	Al
ČSN EN 573-3	0.4–0.8	0.7	0.15–0.4	0.15	0.8–1.2	0.04–0.35	-	0.25	0.15	-
Experiment	0.566	0.407	0.21	0.099	0.882	0.211	<0.0020	0.011	0.023	97.51

. The mechanical properties according to EN 485-2 [20] were determined via the static tensile test at room temperature and are given in

Table 2. Mechanical properties of alloy AW 6061.

AW-6061 T651	Rp0.2 [MPa]	Rm [MPa]	Ag [%]	A50mm [%]	HBW	HV5
ČSN EN 485-2	min. 240	min. 290	min. 8	-	min. 88	-
Experiment	280.8 ± 1.0	302.5 ± 1.2	12.62 ± 0.21	23.25 ± 0.38	107 ± 5	115 ± 7

. The static tensile test was carried out in accordance with standard EN ISO 6892-1 [21] on a TIRA Test 2300 testing device. The course of force was recorded using a KAF 100 strain gauge (A.S.T.—Angewandte System Technik GmbH, Dresden, Germany) with a range of 100 kN and the magnitude of elongation was measured with an MFX 500 extensometer (MF Mess- & Feinwerktechnik GmbH, Velbert, Germany). The loading rate was selected in accordance with the above standard so that the loading rate was 1 mm·min⁻¹ up to the yield strength, then after exceeding this value the rate continuously increased up to a value of 15 mm·min⁻¹. The Brinell hardness HBW was measured in accordance with EN ISO 6506-1 [22] and the Vickers hardness HV 5 in accordance with ISO 6507-1 [23].

Figure 1 shows the structure of the delivered material, i.e., in the T651 condition. This involves a heat treatment consisting of solution annealing followed by a combination of stress–relief annealing via controlled deformation and artificial aging. Samples used for structure observation were prepared using common metallographic procedures and observed on a MIRA 3 TESCAN electron microscope (Tescan Orsay holding a.s., Brno, Czech Republic). The structure consisted of a solid solution α in which the Mg₂Si phase was finely precipitated. Additional particles were observed in the structure, most probably α-AlFeMnSi particles based on the EDX analysis.

2.1. Evaluation of Heat Treatment Effects on the Mechanical Properties of Alloy AW 6061

The mechanical properties of AW 6061 depend not only on the condition in which the material is delivered but also on the temperature cycle to which the material is subjected. The main goal of the experiments carried out in this paper was to assess both the influence of the preheating temperature during multipass welding and the possibility of restoring the mechanical properties of the welded joint by heat treatment. Therefore, in the first phase of experiments, attention was focused on finding the optimal heat treatment parameters (temperature and artificial aging time). The most suitable heat treatment parameters were subsequently used to restore the mechanical properties in the HAZ. The experiments also included an assessment of the natural aging effect.

First, testing samples measuring 15 × 15 × 15 mm were fabricated from the delivered material. For AW 6061, it is recommended to perform solution annealing at 530 °C for 1 h followed by cooling in water at 20–40 °C. For subsequent artificial aging, heat treatment at either 160 °C for 16 h or 175 °C for 8 h followed by air cooling is recommended [18]. Therefore, to assess the heat treatment intensity, the experiments were designed and subsequently carried out in accordance with

Table 3. Performed heat treatment experiments of base material.

Heat Treatment Method	Holding Time (h)
	0.5	1	2	4	8	12	24	120	360
Solution annealing 530 °C		X
Artificial aging 160 °C	X	X	X	X	X	X	X
Artificial aging 175 °C	X	X	X	X	X	X	X
Artificial aging 190 °C	X	X	X	X	X	X	X
Natural aging RT							X	X	X

.

The solution annealing was carried out in a 11016S Classic electric resistance furnace (Clasic CZ s.r.o., Praha, Czech Republic) with a temperature tolerance of ±10 °C. It was performed at a heating rate of 5 °C·min⁻¹ to temperature 530 °C, with a holding time of 1 h, followed by cooling in water at 20 °C. The solution annealing was immediately followed by artificial aging at temperatures of 160, 175 and 190 °C, followed by cooling in air. This was carried out in a Venticell Standart 404 furnace (BMT a.s., Brno, Czech Republic) with a temperature tolerance of ±3 °C. An assessment of the heat treatment intensity was done using a Brinell hardness test measurement. This method is commonly recommended to measure hardness for non-ferrous metals. In line with EN ISO 6506-1 [22], a ball with a diameter of 2.5 mm and a load of 62.5 kp (612.9 N) was used for this measurement. An HPO 250 hardness tester (Kögel Werkstoff- und Materialprüfsysteme GmbH, Leipzig, Germany) with LabControl CCD evaluation software (LabControl s.r.o., Opava, Czech Republic) was used for evaluation results. After solution annealing, we measured a hardness value of 50 ± 1 HBW. The results of the artificial aging for each temperature are shown in Figure 2. In addition, the effect of natural aging was also monitored in this paper. For natural aging, a hardness of 55 ± 1 HBW was achieved after 24 h, followed by 60 ± 1 HBW after 5 days. The hardness did not increase further.

From the experimental results, it is evident that the maximum hardness values were achieved after artificial aging at 160 °C for 12 h, at 175 °C for 8 h, or at 190 °C for 4 h. Based on the hardness measurement results (see Figure 2), the mechanical properties of alloy AW 6061 were determined via static tensile test in accordance with EN ISO 6892-1 using the TIRAtest 2300 device (TIRA GmbH, Schalkau, Germany). The samples were subjected to the same heat treatment, with a holding time at temperature corresponding to the maximum hardness achieved at that temperature. The naturally aged samples were mechanically tested 8 days after the solution annealing. All measured mechanical values are given in

Table 4. Mechanical properties of alloy AW 6061 after heat treatment.

	Rp0.2 [MPa]	Rm [MPa]	Ag [%]	A50mm [%]
Solution annealing 530 °C-1 h	76.2 ± 5.4	168.8 ± 1.4	19.21 ± 1.66	25.26 ± 0.15
530 °C-1 h-natural aging	115.4 ± 2,7	218.7 ± 3.2	17.47 ± 0.52	25.14 ± 0.28
530 °C-1 h-160 °C-12 h	267.8 ± 2.1	304.9 ± 1.6	8.63 ± 0.32	13.48 ± 0.46
530 °C-1 h-175 °C-8 h	263.3 ± 1.3	294.1 ± 0.5	7.3 ± 0.24	12.3 ± 0.47
530 °C-1 h-190 °C-4 h	247.20 ± 1.6	275.27 ± 1.29	6.21 ± 0.28	10.95 ± 0.31

.

2.2. Multipass Welding of Alloy AW 6061

The welding experiment was designed as a three-pass welding experiment concerning the thickness of the delivered material (15 mm). This involved welding with a washer (the washer had a root face of 1 mm in this case), as shown in Figure 3. The total size of the specimen used for welding was 250 × 150 × 15 mm. In order to assess the effect of the partial heat treatment influence of each bead, the staggered bead method was chosen. This consisted of the root bead being process carried out along the entire length of the test sample, due to the use of a bead-in and bead-out pad. The filling bead was welded from the rise pad to two-thirds of the test sample length and the capping bead started the same way but was terminated at one-third of the test sample length.

Welding was performed via the MIG method using a Lorch S Speed Pulse instrument (LORCH Schweißtechnik GmbH, Auenwald, Germany) for source power with OK Autrod 5087 (AlMg5) filler wire with a diameter of 1.2 mm. The chemical composition of the filler material according to the material testimonial is given in

Table 5. Chemical composition of the filler material according to the material testimonial.

Element	Si	Fe	Mn	Mg	Cr	Zn + Ti	Al
Experiment	0.25	0.4	0.15	4.5–5.6	max. 0.8	0.25	rest

. The filler material was chosen as a non-heat-treatable alloy of the 5XXX series, as this combination is used by industrial manufacturers in transportation. The reason for this was to avoid the hot cracking tendency of welded metal. As a shielding gas we used Ar with a purity of 4.9 and flow rate of 15 L·min⁻¹. Welding was carried out in pulse mode with currents of Ip = 191 A and Iw = 157 A and times of tp = 0.198 s and tw = 0.132 s. The pulse welding frequency was 3 Hz. Welding was carried out first without preheating and then with preheating temperatures of 75 °C and 150 °C. The process parameters for each bead and preheating temperatures are shown in

Table 6. Process parameters used for multipass welding at different preheating temperatures.

Process Parameters for Multipass Welding	No. of Bead	Effective Welding Current [A]	Welding Voltage [V]	Welding Travel Speed [m·s−1]	Wire Feed Speed [m·min−1]	Total Heat Input [kJ·cm−1]
Weld 1 with preheating 75 °C	1	174.8	23.9	0.00581	11.4	7.191
	2	173.9		0.00369		11.263
	3	174.2		0.00259		16.075
Weld 2 with preheating 150 °C	1	173.1	23.9	0.00595	11.3	6.953
	2	172.7		0.00386		10.693
	3	173.4		0.00268		15.464

.

Table 6 does not show the process parameters for the specimen welded without preheating. The intensity of heat removal through the specimen was so great that cold joints occurred along the entire length of the specimen, as can be seen in the metallographic scratch pattern in Figure 4. Therefore, the effect of HAZ during welding without preheating will not be shown in this paper because the results cannot be practically used.

During welding with preheating, the entire specimen was first uniformly preheated in a Venticell Standart 404 furnace (BMT a.s., Brno, Czech Republic) to a temperature of 25 °C higher than the required preheating temperature. The test plate was then removed from the furnace and temperature was measured at the welding surfaces using a contact thermometer. When the required preheating temperature was reached, the weld beading was performed. The interpass temperature, which was identical to the preheating temperature, was then measured with a contact thermometer during welding of the filling and capping bead. Immediately after welding, the welded specimens were cut as shown in Figure 5. The sides of each specimen were machined so that the intensity of the specimen’s influence could be measured using the Vickers hardness test.

2.3. Evaluation of the Welding Influence on the Changes in HAZ

When AW 6061 is welded in the hardened state, the hardening phases dissolve in the HAZ, resulting in decreased hardness and strength properties. The intensity of the hardness decrease depends on the heat input during welding, but also on the intensity of heat removal from the weld zone. The higher the preheating temperature, the lower the heat removal intensity, meaning the dissolution of the hardening phases occurs at greater distances from the fusion line. The number of individual beads also plays a role in multipass welding. Areas in the HAZ are, thus, affected by multiple temperature cycles.

The changes in mechanical properties of the welded joints were monitored by measuring the hardness HV 5 according to standard ISO 6507-1 [23]. Testing specimens used for the hardness measurements were prepared as described in Section 2.2. Hardness measurements were carried out in accordance with Figure 6.

For the single-bead specimens, the hardness values were measured in line 1 (3 mm from the bottom edge of the material), whereas the fusion line was taken as position 0. For the samples with two beads, hardness values were again measured in line 1 (3 mm from the bottom edge of the material and positioned as for the samples with one bead). This allowed the effect of the second bead on the changes in HAZ to be assessed. Line 2 was then created at a distance of 8.5 mm from the bottom edge of the sample and position 0 was taken as the intersection of line 2 and the second bead. Line 3 was then made at a distance of 12.5 mm from the bottom edge of the sample and position 0 was taken as the intersection of line 3 and the third bead. A minimum of 50 impressions was always made in each row, with the centers 1 mm apart from each other.

Figure 7 graphically shows the results of HV 5 hardness measurements in the weld metal, HAZ and base material in Line 1 for a weld with a preheating temperature 75 °C. The hardness values are shown after the first, second and third beads. Figure 8 graphically shows identical results, although for a weld with a preheating temperature of 150 °C.

To evaluate the structural changes in the HAZ, metallographic processing of the sample was performed, followed by microscopic evaluation on a MIRA 3 TESCAN electron microscope (Tescan Orsay holding a.s., Brno, Czech Republic). Figure 9a shows the microstructure of the HAZ after welding three weld beads with preheating at 75 °C. This area was located at a distance of 3 mm from the bottom edge of the sample (line 1) and at a distance of 4 mm from the fusion line. Therefore, this was the area most affected by multiple temperature cycles. Figure 9b shows the microstructure of the HAZ after welding three weld beads with preheating at 150 °C in the same location.

In Figure 10, the results of HV 5 hardness measurements in the weld metal, HAZ and base material in lines 2 and 3 are graphically shown with preheating at 75 °C. The hardness values courses in line 2 are shown after welding both the second and third weld beads. The hardness value courses in line 3 (dashed line) are shown only after welding the third weld bead. Figure 11 graphically shows identical results, although for a weld with a preheating temperature of 150 °C.

2.4. Heat Treatment of Multipass Welds on Alloy AW 6061

As can be seen from the results presented in the previous chapter, welding has a significant effect on the changes in joint mechanical properties. To restore the properties in the HAZ, heat treatment must be applied. Based on the results from experiments performed on the base material (Section 2.1), heat treatment at which the highest hardness values were achieved was applied to specimens welded with both preheating temperatures. This involved solution annealing at 530 °C for 1 h and water cooling followed by artificial hardening at 160 °C (holding time 12 h), 175 °C (holding time 8 h) and 190 °C (holding time 4 h). The HV5 hardness was again measured on all samples at the same locations as shown in Figure 6.

To assess the heat treatment effect (160 °C with holding time of 12 h) at different locations of the welded specimen, the hardness value courses at the locations of lines 1, 2 and 3 after welding all three weld beads with preheating at 75 °C are shown in Figure 12. Similarly, Figure 13 shows the hardness value courses at the locations of lines 1, 2 and 3 for welding with preheating at 150 °C. The courses of hardness values after welding the individual weld beads are shown as dashed lines and hardness values measured under different conditions of artificial aging are shown as solid lines.

The cumulative effect of different heat treatment types on a three-pass weld performed at a preheating temperature of 75 °C is shown in Figure 14. This is the hardness values course in the zone of line 1 after welding all three beads. The hardness values measured after welding are shown in black. The orange color indicates the hardness values measured in the zone of line 1 after solution annealing at 530 °C for 1 h, while the red dashed line shows the hardness values obtained at the same location after solution annealing and natural aging for 14 days. The hardness values measured after artificial aging at 160 °C (12 h) are shown in blue, the hardness values measured after artificial aging at 175 °C (8 h) are shown in red and the hardness values obtained after artificial aging at 190 °C (4 h) are shown in green. In the same way, Figure 15 shows the cumulative effect of different heat treatment types on a three-pass weld performed at a preheating temperature of 150 °C.

After application of the heat treatment to the welded joint, the structural changes in the HAZ were again assessed. Figure 16a shows the microstructure of the HAZ after welding three beads with a preheating temperature of 75 °C, solution annealing at 530 °C for 1 h, water cooling and subsequent artificial aging at 160 °C for 12 h. As in Figure 9a, the zone is located 3 mm from the bottom edge of the sample (line 1) and 4 mm from the fusion line. Figure 16b shows the microstructure of the HAZ after welding three beads with a preheating temperature of 150 °C in the same location.

Figure 17a shows the microstructure of the HAZ after welding three runs with a preheating temperature of 150 °C and after subsequent artificial aging at 175 °C for 8 h. Figure 17b shows the microstructure of the HAZ after welding three runs with a preheating temperature of 150 °C and after subsequent artificial aging at 190 °C for 4 h.

3. Discussion

Welding of heat-treatable Al alloys has many specifics. In the HAZ, the temperature field causes dissolution of the hardening phases and a decrease in mechanical properties, while the simultaneous intense heat removal increases the probability of creating cold joints and non-penetrations. Moreover, welded metal can be sensitive to hot cracks. Welds that have a wide solidification interval are more sensitive to cracking. The creation of cracks is also influenced by the crystallization method, segregation and possibly also by liquation [12,14]. If the volume of the liquid eutectic phase in the final stage of crystallization is sufficient (15–25%, depending on the type of alloy) to fill the space between the growing dendrites, the conditions for the creation of crystallization cracks are not fulfilled. In industrial production and especially in transportation, materials of higher thicknesses are usually joined, meaning the utilization of preheating is necessary and individual weld zones are affected by multiple temperature cycles from individual beads.

Based on experiments involving MIG welding on alloy AW 6061, it was found that when welding 15-mm-thick plates without preheating, defects-free welds cannot be made, even at high heat input values. Cold-joint-type defects (see Figure 4) occur throughout the weld length and in all weld beads. However, similar defects also occur when welding lower thicknesses without preheating, as proven by Guzman et al. [8] on 7-mm-thick plates. Materials of lower thicknesses can then be welded using various methods without preheating, without causing cold joint defects. This was demonstrated by Vargas et al. [1] for welds performed by MIG welding on a 4.8-mm-thick plate and by Nie et al. [2] on a 4-mm-thick plate, and for TIG welding by Peng et al. [9] on a 4-mm-thick plate and by Kumar and Sundarrajan [3] on a 3-mm-thick plate.

The preheating temperature has significant effects both in terms of decreasing the hardness in the HAZ and on the width of the HAZ. When the first bead was welded with a preheating temperature of 75 °C, there was 41% decrease in hardness at line 1 to 66 HV (Figure 7) and a 47% decrease in hardness at 150 °C to 61 HV (Figure 8). The largest hardness decrease was detected between 4 and 5 mm from the fusion line, regardless of the preheating temperature. The width of the HAZ after welding the first bead with preheating at 75 °C was 19 mm, while for preheating at 150 °C it was 20 mm.

Welding without preheating at smaller thicknesses resulted in decreased hardness varying from 65 to 73 HV with an HAZ width of 6 mm [8], 70 HV for a width of 6 mm [9], 61–68 HV for an HAZ width of 7 mm [7] and 66 HV for an HAZ width of 14 mm [3]. From these results, it is clear that the effect of welding with preheating has a similar effect on the hardness decrease in HAZ as welding without preheating. The effect of preheating was much more significant in terms of HAZ width. The effect of multiple temperature cycles from individual beads was also quite significant. In the evaluation of the hardness courses in the HAZ at the same sample location (row 1), the welding with preheating at 75 °C resulted in overall hardness decreases of 48% after the second bead and 52% after the third bead. However, there was no further expansion of the HAZ after three beads. Welding with preheating at 150 °C resulted in a 51% decrease in overall hardness after the second bead and a 55% decrease after the third bead. The width of the HAZ expanded in row 1 from the original 20 mm up to 32 mm after three beads.

Based upon these findings, it can be concluded that the preheating temperature affects both the mechanical properties in the HAZ and its width as well. However, it does not result in a maximum hardness decrease in the HAZ. The influence of multipass welding is, therefore, more specific. The temperature effect on a specific area in the HAZ is given by the superposition of the temperature fields generated by the individual weld beads. The influence of the partial weld beads can then be defined as a function of a given weld bead heat input value and its distance from a specific area. A higher preheating temperature then allows more intense heat transfer through the conduction from each weld bead and the intensity of the thermal effect on the given area. This combined effect of multiple temperature cycles is inevitably reflected in the structure in a given area of the sample. The higher the temperature reached and the longer the time when this area is exposed to this temperature, the more intense the dissolution of the hardening particles.

By means of natural and artificial aging, it is possible to improve the properties of the welded joint. While natural aging results in a hardness increase in the HAZ of about 5 HV after 24 h, the precipitation of hardening particles slows down with increasing time and stops completely after 120 h. The hardness in HAZ stabilized at an average value of 65 HV for both preheating temperatures. Peng et al. [9] were able to increase the hardness in the HAZ by up to 15 HV via natural aging, depending on the heat input value. Wang et al. [7] were able to achieve 16 HV increase in HAZ hardness via artificial aging after hybrid laser–TIG welding.

Significantly better results were achieved with artificial aging. Artificial aging at 160 °C for 12 h in the HAZ resulted in HV5 hardness increases of up to 97%, while at 175 °C with a holding time of 8 h, the HV5 hardness was increased by up to 92% of the original hardness value under T651 conditions. The magnitude of preheating did not affect the age-hardening in this case. At 190 °C and with a holding time of 4 h, the HV5 hardness values increased to 88% (preheating at 75 °C) and 83% (preheating at 150 °C), respectively. There was also a partial hardness increase in the weld metal, even though the non-heat-treatable alloy 5083 was used as the filler material. This was due to partial mixing with the base material. This was also confirmed by the fact that slightly higher hardness values in the weld metal were achieved in welds with a preheating temperature of 150 °C, where more intense mixing occurs.

From a microstructural point of view, the delivered material consists of a solid solution α with distributed hardening particles of Mg₂Si and also with non-heat-treatable particles od Al(FeMnSi) (see Figure 1). Furthermore, the microstructure of the HAZ after welding of all three beads was evaluated at a distance of 4 mm from the fusion line (line 1), i.e., in the area of the greatest hardness decrease. After welding with preheating at 75 °C, most of the hardening particles of Mg₂Si were dissolved in this location. Non-heat-treatable particles Al(FeMnSi) were not dissolved (see Figure 9a). After welding with preheating at 150 °C, no more hardening particles of Mg₂Si were detected in the structure and the solid solution α contained only non-heat-treatable particles of Al(FeMnSi) (see Figure 9b). The artificial aging of welds at 160 °C for 12 h resulted in new precipitation of the hardening phase for Mg₂Si and in the same amount as in the base material. The size and distribution of particles were similar for both types of welds, regardless of the used preheating temperature (Figure 16a,b). Artificial aging at 175 °C for 8 h (Figure 17a) and 190 °C for 4 h (Figure 17b) also resulted in the precipitation of the hardening phase for Mg₂Si. However, the total number and size of hardening particles were lower than for hardening at 160 °C for 12 h.

4. Conclusions

This paper aimed to show the effects of multiple temperature cycles, performed at different preheating temperatures, on the changes in the microstructure and properties of a butt-welded joint performed on alloy EN AW 6061. At the same time, we also investigated to what extent the original material properties can be restored in the HAZ by artificial aging at 160, 175 and 190 °C and by natural aging. The experimentally obtained knowledge achieved in this work can be summarized in the following points:

• By means of artificial aging applied after solution annealing, we achieved hardness values HBW of 85–90% concerning the base material supplied under T651 conditions. In terms of the mechanical properties, we achieved 91–100% of the ultimate strength R_m of the base material (condition T651) via artificial aging. The best results were achieved at an artificial aging temperature of 160 °C for 12 h;
• Only 56% of the original HBW hardness was restored by natural aging. After 120 h of natural aging, there was no further increase in HBW hardness;
• With the same adjustment of the process parameters during welding, the effect of preheating temperature on the heat input value was very small. At a preheating temperature of 150 °C, the heat input was about 5% lower for all weld beads than for welding with a preheating temperature of 75 °C;
• When single and multiple temperature cycles were used in welding, the greatest hardness decrease occurred between 4 and 5 mm from the fusion line, regardless of the preheating temperature. In terms of hardness decreases in the HAZ, there were decrease in HV5 hardness of 41% after the first bead, 48% after the second bead and 52% after the third bead when using preheating at 75 °C. With preheating at 150 °C, there were decreases in HV5 hardness of 47% after the first bead, 51% after the second bead and 55% after the third bead;
• The effect of the preheating temperature was particularly significant for the width of HAZ. When welding with a preheating temperature of 75 °C, the width of the HAZ remained the same after welding the individual beads, i.e., 20 mm. When welding at a preheating temperature of 150 °C, the width of the HAZ achieved after the first bead (20 mm) increased to a final width of 32 mm after three beads;
• Restoration of the properties in the HAZ is possible using heat treatment, although its effectiveness depends on the used aging temperature and holding time. The best results were achieved for artificial aging at 160 °C for 12 h, with HV5 hardness values in the HAZ increased to 97% of the original hardness values of the base material under T651 conditions. The amount of preheating did not affect the magnitude of hardening in this case. Natural aging in the HAZ was only able to increase the HV5 hardness to 57% of the original hardness value;
• A partial increase in HV5 hardness also occurred in the weld metal, although the non-heat-treatable alloy 5083 was used. This was due to partial mixing with the base material. This was also confirmed by the fact that slightly higher hardness values in the weld metal were achieved in welds with a preheating temperature of 150 °C, as more intense mixing with the base material occurred;
• The magnitude of the preheating temperature has an effect on the microstructure in the HAZ. In the area affected the most by the multiple temperature cycle, most of the hardening particles were dissolved by preheating at 75 °C and non-heat-treatable particles were found after welding with preheating at 150 °C.