Experimental Analysis on Fatigue Life Assessment of Dissimilar Aluminum Alloys Weld Joints under Four-Point Rotating Bending Condition

Abstract

Industrial applications often require welded joints of dissimilar metals to perform reliably under adverse working conditions. Such conditions demand high resistance to failure modes, such as fatigue, corrosion, and creep. In applications involving intense vibrational and cyclic loading conditions, high strength-to-weight ratio metals, including aluminum, are more vulnerable to fatigue failure. This study investigates the fatigue strength of base metals and tungsten inert gas welded joints of the two most widely used aluminum alloys, AA7075 and AA2024, under four-point rotating bending. For the comparative study, these dissimilar alloys were joined together by using two different filler rods, AA4047 and AA4043, under similar welding conditions. SN curves for all the samples were generated through the experimental iterations. A comparison of different results shows that welded samples have better fatigue strength than the base metals, while, as a filler, AA4047 has better resistance to fatigue fracture than AA4043.

1. Introduction

Tungsten inert gas (TIG) welding is widely used for the welding of different types of metals. These joints, especially TIG-welded aluminum alloys, are used in various high-performance structural and loading applications. The TIG welding of aluminum alloy has also gained importance in the modern automobile and aerospace industries because of its light weight, high strength and ductility, good thermal conductivity, and corrosion resistance [1,2,3]. Moreover, the wide applications of TIG welding are due to its low distortion, high quality, and precise control of welding variables [4]. Most of these welded structures are subjected to cyclic loadings, in which mean stress and stress amplitude keep on changing with time [5]. This influences the fatigue properties of the welded structures. Fatigue life prediction is one of the most important problems in design engineering for reliability. New technologies permit far more complex geometries to be fabricated. Such complex geometries need sophisticated welded joints between materials with different grain properties [6].

Assemblies of different aluminum grades and weld joints are the potential areas to be studied under fatigue loading, along with the examination of crack initiation and propagation behavior in different regions, such as the interface region, weld zone, and heat-affected zone (HAZ). There is also a need to develop behavior charts for various microstructures in the base metals. Some researchers investigated the microstructure, fatigue life, heat-affected zone, and weld zone by using different types of welding. Ye et al. [7] investigated the behavior of the butt joint of aluminum alloy and low-alloy steel by introducing the new technique of arc braze-welding on both sides. Mechanical properties and microstructures were compared with MIG-welded butt joints. Singh et al. [8] studied the TIG-welding behavior of austenitic stainless steel and low-alloy steel with narrow-gap welding. The HAZ width and martensite layers were prominent in simple welding. Tensile testing showed that a crack propagated through IN82 near the stress concentration. Li et al. [9] evaluated the behavior of TIG-welded joint aluminum alloy AA2219. That welding process used a special active agent to reduce the porosity and oxide content. The results depicted a decrease in HAZ and operating current. Shah et al. [10] analyzed the mechanical properties and microstructure of the TIG-welded lap joint of Al 6061 and iron (galvanized). Two types of Si filler metals, ER4043 and ER4047, were used. Gungor et al. [11] investigated the fatigue strength of friction stir welding two different aluminum alloys. The tensile tests resulted in high-yield stress. Zhou et al. [12] evaluated the fatigue life of aluminum alloy 5083 welded by MIG welding and friction stir welding. Deng et al. [13] evaluated the butt joints of friction stir welding and the fatigue properties of aluminum alloy 7050-T7451. The results showed that fatigue failure of the friction stir butt-welded joint occurred at 7.0 × 10⁸ cycles. In addition, the fatigue strength of FSW-welded joints was superior to that of the metal base.

Lillemäe et al. [14] investigated the effects of fatigue strength and geometric properties of the butt-welded joint on thin slender specimens. While Liting Shi et al. [15] considered spot welding and found that AA5754 spot welded to a high-strength low-alloy steel sheet had higher fatigue life than AA5754 welded to its own self, they resistance-spot welded a 1.1 mm thick AA5754 sheet to a high-strength low-alloy steel sheet of 2.0 mm thickness. Yufang Chang et al. [16] compared nominal stress and notch stress S-N curve and found the scatter band index of nominal stress higher than notch stress. Crupi et al. [17] devised a new method, known as the thermographic method, which was purely based on thermographic analysis to estimate the fatigue behavior of the butt-welded joints. Pasqualino Corigliano and Vincenzo Crupi [18] considered a Ti6Al4V/Inconel 625 joint welded by laser welding with an intermediate layer of AISI 304 stainless steel and Vanadium. They compared the thermographic method traditional procedure and found good agreement. Fricke et al. [19] studied that the residual stresses present in the structures due to welding had a significant effect on the fatigue behavior. The simulation results revealed that the residual stress (compressive) occurring in the less-strained areas might cause the initiation of cracks. Guilherme Alencar et al. [20] deduced algorithms for fatigue analysis for finding full capabilities for welded structures with constant amplitude loading and multi-axial stress states by using Finite element methods software. Balasubramanian et al. [21] analyzed the mechanical properties of two TIG-welded alloys. The authors analyzed the effect of the pulsating mode over the old conventional technique. Juang et al. [22] analyzed the effect of process variables on the weld geometry by using the Taguchi method and concluded that the use of the Taguchi method greatly optimized the TIG welding of stainless steel by improving its quality characteristics.

Shuwan et al. [23] analyzed TIG-welded duplex stainless steel in a single pass, without using the filler metal and grooves. The weld geometries were studied by changing the process parameters, i.e., current and speed. The geometry was carefully evaluated using an optical and scanning electron microscope. Huei et al. [24] deliberated the influence of sample thickness and mean stress on the fatigue life of an Al-Mg-Si alloy butt joined using TIG welding and vacuum brazing. Fatigue loading tests with constant and variable amplitude were performed. Wang et al. [25] studied the influence of process variables applied in the tungsten inert gas welding of a nickel-base alloy on microstructure, tensile strength, and failure. The results showed that when the current was increased and the welding speed was slowed down, the heat input went up.

Salvati et al. [26] noticed the effect of residual stress and crack closure on crack retardation during overload. They found that the crack closure effect was greater at low baseline load ratios but for a longer time, while the residual stress effect was for a short time but present at all values of baseline load ratios. Espinosa et al. [27] found the effects of different parameters, i.e., single and block overloads, and random spectrum loading, on the crack growth. Then, a single tensile overload effect on fatigue crack growth in dual-phase steel was analyzed by Li et al. [28]. It was also studied by Datta et al. [29], for an aluminum alloy under biaxial fatigue loading. Rege and Lemu [30] discussed the finite element method and the extended finite element method for fatigue crack propagation modelling with recent developments.

Xiaobo Yu et al. [31] carried out experimentation on the AA7075-T651 alloy to analyze the fatigue growth behavior under proportional and non-proportional mixed mode I and mode II loads for cyclic tension and torsion. They found that AA7075-T651 was capable of producing non-planar shear mode fatigue crack growth that was long and stable for non-proportional mixed mode I and mode II loads. Afterwards, Dirik and Yalcinkaya [32] incorporated a mesh independent computational algorithm, using Abaqus for automated fatigue crack growth analysis, with various conditions, including mixed-mode variable amplitude loading. The results were compared with those of the experimental data. Rege and Pavlou [33] computed the stress intensity factor for geometry under tension and torsion using the finite element method and found a solution for a variety of crack lengths. They also devised a method for modeling growth delays in crack propagation due to stop-hole-induced fatigue cracks under mixed-mode conditions [34]. Furthermore, Dekker et al. [35] derived a new approach to finding the fatigue crack in ductile material with both mixed-mode loading and overloading. They compared two different methods to measure cohesive traction, i.e., the cohesive zone model and the interfacial thick level set method. It was deduced that both methods were in accordance with the mode I analytical relation and a mixed-mode experiment.

Researchers, for instance, Mudaserullah et al. [36], have conducted quite similar work but with oxy-acetylene-welded aluminum alloys joints and discussed the effect of welding on the fatigue strength of the base material. Yuki Ono et al. [37] did fatigue testing on longitudinal welded joints. Mingzhe Fan et al. [38] studied stress–strain behavior in creep-fatigue and tensile testing on a 9Cr/CrMoV weld joint and compared them with base metals. They deduced that the actual strain amplitude is four-times, while the hysteresis energy density is ten-times that of the base metal. Similarly, V. Manikandan et al. [39] compared mechanical characteristics of aluminum alloy 5086 and its welded joint. In all these studies, the properties and characteristics of welded joints were compared with their respective base metals on similar testing criteria.

The literature revealed that tungsten inert gas welding is widely used in many conventional and modern industrial applications. These industries have lower downtime to maintain their high production rates, so TIG welding is used to save time instead of more advanced joining processes, such as friction welding, which require highly skilled workers and plenty of preparation. Most of these joints in engineering applications are subject to continuous vibrations due to fluid flow or uneven load on them, so said joints are investigated under fatigue characteristics. Current research tends to investigate the fatigue life of these two uninvestigated but important welded joints of aluminum alloys, AA7075 and AA2024, with filler alloys, AA4043 and AA4047. Four-point rotating bending conditions are applied to investigate fatigue life. Moreover, the fatigue life of base samples of alloys AA7075 and AA2024 is also investigated to give a significant comparison with the welded samples. Additionally, fractographic analysis of failed samples is also carried out.

2. Materials and Methods

2.1. Materials

Aluminum alloys AA7075 and AA2024 were used as base materials for this study and their compositions were verified using SEM-EDX. Initial samples were drawn from rods, 21 mm in diameter. The filler metals used to join the base metals were AA 4043 and AA 4047. Aluminum Alloy 4047 was developed to take advantage of its low melting point and narrow freezing range due to high silicon content and hence high fluidity and reduced shrinkage as compared to AA 4043. The chemical composition of above-mentioned alloys and filler metals is as below.

2.2. Base and Weld Samples Preparations

For comparison purposes, base metal specimens and welded specimens of dissimilar metals were prepared. The total number of samples prepared was 120. The standard specimen length for use in the 4-point rotating bending machine was taken as 226 mm. The dog-bone specimens of base samples for fatigue with a diameter of 12 mm and a gauge length of 96 mm were fabricated using a Numerical Control (NC) lathe machine with a notch of 10 mm in the center of the gauge length as shown in Figure 1a.

The carbide tip tool has been used for smooth turning and facing of the specimens. A raw sample with a length of 230 mm was fed into the CNC for machining at the required tolerances to make a standard specimen length of 226 mm for the 4-point rotating bending test. For the welded samples, 3048 mm of raw rod was cut into small pieces of 116 mm and then machined at one end into a bevel shape of 60° to minimize the welding effect on the material properties and grain structure. To weld the specimen, a custom fixture to facilitate alignment had to be fabricated to match the interface of the lathe chuck. It facilitated the collet grip on the specimen before and during welding. A collet chuck fixture is necessary to maintain the shape and alignment of the specimens. Welding without it leads to distorted, unusable samples. After getting the required sample geometries, the specimens were welded using the tungsten inert gas welding process, and the post-welding sample is as shown in Figure 1b.

Welding speed is matched to the lathe chuck RPM by using a rheostat on a separate drive motor.

Table 1. Chemical composition of Aluminum alloys and filler metals.

	Al	Cu	Fe	Mg	Mn	Zn	Ti	Si	Br	Other (Each)	Other (Total)
AA7075	87.1–91.4	0.18–0.28	Max 0.50	2.1–2.9	Max 0.30	5.1–6.1	Max 0.20	Max 0.40	-	0.05	0.15
AA2024	920.7–94.7	Max 0.10	Max 0.50	1.2–1.8	0.3–0.9	0.25	0.15	0.5	-	0.05	0.15
AA4043	Remaining	0.80	0.30	0.05	0.05	0.10	0.2	4.5–6.0	0.0003	-	-
AA4047	86.70	0.5	0.30	0.10	0.15	0.2	-	12	0.0006	-	-

and

Table 2. Illustration of welding parameters.

Sr. No.	Weld Parameters	Values
1	Current supply	90 A
2	Rotational speed during welding	3.7 RPM
3	Specimen gap during welding	1 mm
5	Welding spots	2
6	Passes	3
7	Length contraction during welding	0.7 mm

show the welding parameters and temporal welding parameters, respectively. The specimen is unloaded from the fixture after complete cooling. Weld quality is tested to ensure compliance with the testing standard. The external and internal defects are tested through two different non-destructive techniques, i.e., liquid penetrant testing (LPT) and radiography.

2.3. Fatigue Testing

The fatigue testing machine, manufactured by Jinan Precision Testing Equipment (Jinan, China), is called the “Pure Bending Fatigue Testing Machine,” Model No. PQ 6, and was used for the fatigue test. Standards GB4337-84 and ZBN71006-87 have been applied. The working principle of this machine is to clamp the specimen at both ends and weight is used to determine test load [40]. Stresses include tension and compression changes on upper and lower side of specimen due to rotation of the specimen by the motor. Therefore, for an instant the upper most layers of the specimen are under maximum compression while lower most are under maximum tension. During testing, the specimen is clamped and rotated along the main shaft and the pulling rod applies the load at both ends. Hence, pure bending is evenly distributed along the entire specimen. The direction of load does not change but the specimen rotates, while on the same speed, the specimen faces alternative bending stress at each point. This bending fatigue causes fracture after many alternative cycles. The location of loading or line of four point forces is shown in a schematic diagram [36]. Schematics and real fatigue testing machines are shown in Figure 2a,b.

Samples with standard dimensions can be tested for fatigue failure at different desired stress levels. The testing is carried out at constant amplitude and fully reversed cycling. For each load, three different tests were performed and averaged. Number of specimens tested in this study are presented in

Table 3. Illustration of temporal welding parameters (time measured in seconds).

1st Spot (Root Pass)	Delay	2nd Spot (Root Pass)	Delay	1st Pass (Hot Pass)	Delay	2nd Pass (Fill Pass)	Delay	3rd Pass (Cover Pass)
08	20	12	18	32	27	27	240	30

.

Seven different loads were applied to the samples, and three samples were tested for each load. The S-N curves of base samples and welded samples with different filler materials were plotted. All the samples, i.e., base or welded, were subjected to reverse cyclic stresses, having a stress ratio of R = −1. Figure 3 is the S-N curve for the base metal AA7075.

Table 4. Depiction of experiment iterations.

Description	Load Steps SN Curve	Each Load Iterations	Number of Specimens Tested
SN-Curve welded	14	3	44
SN-Curve base	14	3	44

reveals the seven different loads that are applied to the samples.

The fundamental beam theory was used to evaluate the degree of repeated stress in the un-notched sample after load application.

$$\frac{\sigma }{C}=\frac{M}{I}$$

(1)

where,

• $\sigma =$ Applied (alternating) stress (MPa)
• Q = Load Applied (N)
• C = Distance from neutral axis at which stresses are being evaluated = D/2 (mm)
• M = Moment of load (N-mm) = $Q\times \frac{\mathrm{l}}{2}$
• I = Moment of Inertia.

Substituting these values in Equation (2):

$$\frac{\sigma }{\frac{D}{2}}=\frac{Q\times \frac{l}{2}}{\frac{\pi }{64}\times D^4}$$

(2)

Solving Equation (2) for $\sigma$ and putting $l$ = 100 mm,

$$\sigma =\frac{Q \left(\mathrm{N}\right)\times 50 \left(\mathrm{mm}\right)\times 32}{\pi D^3 \left({\mathrm{mm}}^3\right)}$$

$$\sigma =\frac{509.55 Q}{D^3} \left({\mathrm{N}\mathrm{mm}}^{-2}\right)=\frac{32M}{\pi D^3}$$

From the notch specimen, the alternating stress is:

$${\sigma }_{notched}=K_t\times {\sigma }_{un-notched}$$

(3)

After fatigue testing, the fractured surfaces are evaluated through the fractography technique. The specimens are put under the lens of SEM to investigate the surface morphology throughout the fracture zone and the role of different phases of materials in the fracture of the part. A study has also been carried out on the fracture specimen using the optimal metallurgical microscope. After the fatigue failure, the specimens were stored in a low-humidity environment, to avoid any corrosion due to water content and protect the desired study surfaces. The specimens were wrapped in high tight packing and marked according to their respected specifications and marked with a permanent ink marker. In order to find the notch concentration factor, Equation (4) was used. The calculated value of the stress intensity factor K_t is 1.28.

$$K_t=\frac{{\sigma }_{notched}}{\frac{32M}{\pi D^3}}$$

(4)

3. Results and Discussion

3.1. Fatigue Testing

Fatigue testing was performed and S-N curves were formed according to ASTM standard E 739. The SN curve is basically the plot drawn between cyclic stress and the number of cycles until the failure of the material. At a high stress level, a lower number of cycles can cause failure. At a low stress value, a large number of cycles is required for failure and after a certain high number of cycles, unexpected failures can be seen. Base samples and welded samples are both tested and compared.

Table 5. Various loads applied to the samples.

1	2	3	4	5	6	7
170 N	195 N	245 N	295 N	335 N	395 N	470 N

presents the data after testing 16 base samples of 7075, with a notch to concentrate the stress for fractures at point loading.

Table 6. Fatigue test results of base sample AA 7075.

Sr. No.	Load (N)	$\mathit{\sigma }=\frac{509.55 \mathit{Q}}{{\mathit{D}}^3} \left(\mathbf{N} \mathbf{m}{\mathbf{m}}^{-2}\right)$	Alternating Stress $\mathbf{S}\mathbf{a}=\mathit{\sigma }\times {\mathit{K}}_{\mathit{t}}$ = 1.28 × σ	Fatigue Life (No. of Cycles)
				Trial 1	Trial 2	Trial 3	Average Cycles
1	470	239	306	59,897	67,433	70,268	65,866
2	390	199	254	838,546	947,437	855,535	880,506
3	335	171	218	2,572,867	2,192,966	2,347,428	2,371,087
4	295	150	192	3,698,042	3,163,914	2,963,263	3,275,073
5	245	125	160	8,743,953	8,246,498	7,632,763	8,207,738
6	195	99	127	9,355,500	x	x	9,355,500
7	170	87	111	x	x	x

presents the data after testing 16 base samples of 2024, with a notch to concentrate the stress for fractures at point loading. Figure 4 is the S-N curve for AA2024. Under the same loading conditions, AA 7075 showed better fatigue strength as compared to AA 2024. The difference in fatigue life between the two materials is more prominent under high-intensity stresses as compared to lower stresses. The same trend of AA 2024 and AA 7075 fatigue life was observed in previous studies [41,42]. AA7075 had a better fatigue life than AA2024 because it was stronger and had a better microstructure.

Figure 5 shows the comparison of S-N curves of both base samples, i.e., AA 7075 and AA 2024, while in Figure 6, welded samples with different filler material are compared.

Table 7. Experimental results of base sample AA 2024.

Sr. No.	Load (N)	$\mathit{\sigma }=\frac{509.55 \mathit{Q}}{{\mathit{D}}^3} \left(\mathbf{N} \mathbf{m}{\mathbf{m}}^{-2}\right)$	Alternating Stress $\mathbf{S}\mathbf{a}=\mathit{\sigma }\times {\mathit{K}}_{\mathit{t}}$ = 1.28 × σ	Fatigue Life (No. of Cycles)
				Trial 1	Trial 2	Trial 3	Average Cycles
1	470	239	306	27,989	19,845	27,562	25,132
2	390	199	254	163,382	111,389	141,479	138,750
3	335	171	218	112,836	426,289	492,191	343,772
4	295	150	192	791,337	667,862	819,439	759,546
5	245	125	160	8,963,248	X	X	9,654,416
6	195	99	127	X	X	X	x

and Figure 7 present the results of welded samples of two dissimilar metals using AA 4043 as a filler material. The fatigue test results of welded samples of dissimilar aluminum alloys with AA4043 as a filler material can be seen in

Table 8. Experimental results of welded sample with filler AA 4043.

Sr. No.	Load (N)	$\mathit{\sigma }=\frac{509.55 \mathit{Q}}{{\mathit{D}}^3} \left(\mathbf{N} \mathbf{m}{\mathbf{m}}^{-2}\right)$	Alternating Stress $\mathbf{S}\mathbf{a}=\mathit{\sigma }\times {\mathit{K}}_{\mathit{t}}$ = 2.19 × σ	Fatigue Life (No. of Cycles)
				Trial 1	Trial 2	Trial 3	Average Cycles
1	470	239	524	174,548	203,765	193,658	190,657
2	390	199	435	678,799	648,976	645,685	657,820
3	335	171	374	1,310,876	1,376,758	1,322,658	1,336,764
4	295	150	329	1,409,597	1,419,027	1,395,121	1,407,915
5	245	125	273	1,782,464	1,493,565	1,497,056	1,457,695
6	195	99	217	1,946,526	1,516,786	1,495,766	1,653,026
7	170	87	190	2,197,645	1,864,655	2,197,689	2,086,663

and Figure 8. After carefully analyzing the results, it was clear that the fatigue life of welded samples of two dissimilar metals, i.e., AA7075 and AA2024 using AA4047 as a filler material, is better than welded samples using AA4043 as a filler material when loaded under similar cyclic stress. The difference was due to the low melting point and low freezing range of AA4047 as compared to AA4043. AA4047 provided welded joints with more crack resistance, which is why joints with this filler metal have better fatigue life.

Table 9. Experimental results of welded sample with filler AA 4047.

Sr. No.	Load (N)	$\mathit{\sigma }=\frac{509.55 \mathit{Q}}{{\mathit{D}}^3} \left(\mathbf{N} \mathbf{m}{\mathbf{m}}^{-2}\right)$	Alternating Stress $\mathbf{S}\mathbf{a}=\mathit{\sigma }\times {\mathit{K}}_{\mathit{t}}$ = 2.19 × σ	Fatigue Life (No. of Cycles)
				Trial 1	Trial 2	Trial 3	Average Cycles
1	470	239	524	215,872	274,034	244,647	244,851
2	390	199	435	944,343	1,024,538	961,987	976,956
3	335	171	374	1,123,849	1,225,455	1,169,063	1,172,789
4	295	150	329	1,439,968	1,347,876	1,076,507	1,288,117
5	245	125	273	1,725,789	1,489,768	1,238,879	1,484,812
6	195	99	217	1,953,775	2,019,875	1,578,768	1,850,806
7	170	87	190	2,676,879	2,164,675	1,896,437	2,245,997

is the illustration of the welded sample results using AA4047 as a filler material.

Figure 9 shows that at log (no. of cycles) 5.0 to 6.5, i.e., the area which carries both the welded samples results and the base sample results, correlate with each other, with the welded samples showing better fatigue strength than the base samples. However, the superiority of the welded samples is only scattered in the middle of the horizontal axis, while at the lowest and the highest limits of the number of cycles, base samples have considerable fatigue life, though still less than those of the welded samples, as the welded samples were prepared with homogeneous environment conditions. Porosity and inclusion of slag and holes were avoided by evaluating the weld quality. Hence, welded samples were less affected by the stress concentration factor than those of base metals.

3.2. Fractographic Analysis

Figure 10a depicts the clear areas of ductile and brittle fracture surfaces. The central part of the fracture shows the brittle fracture. The crescent, as indicated by the arrow, is the crack initiation site at the circumference of the fracture. The final rupture is slightly eccentric due to the relatively low bending force. The circumference of the fracture shows slow crack propagation.

Figure 10b shows high loading, and this is indicated by the fact that more brittle fractures are much more prominent than ductile fractures. The valley of fatigue crack propagation is indicated on one side of the circle, indicating the greater bending force. Ratchet marks at several locations around the circumference indicate multiple fatigue origins and a high stress concentration. It is a cup-and-cone fracture where fractures usually occur at the lowest stress level of the notch. Figure 10c reveals the multiple crack initiation sites, which progress slowly towards the center from the circumference. The crack propagation is off center due to a larger bending force then the torsional force. The maximum fibrous zone can be seen due to a fast fracture.

As shown in Figure 10d, the fracture occurred at the center of AA7075 and in the weld region. The fracture line passes through the fusion region and the bevel part of AA 7075. The crack was initiated from the circumference of the fracture at multiple spots. A compressive stress field was developed at the crack tips, which lessen the crack growth. Ratchet marks or shear ridges can be seen at the center due to multiple crack initiation sites and a high stress concentration condition. Steps are created when multiple crack initiation points are joined [43]. As evident from Figure 10e, three cracks initiated from the lower-right corner and continued towards the center at crescent patches (quench crack). The more threaded portion shows the slow progress of the crack towards the center. The fracture portion indicates low porosity, present at the center of the fracture. The fatigue crack grows slowly at the bottom of the pore. The crack initiation sites are present at the circumference. Hence, a less fibrous zone implies more time will be required for rapturing at low loads.

4. Conclusions

The main conclusions that can be drawn from this study are as below:

• Both pure base metal samples and welded samples have a smaller variation in the fatigue life across trials in the low-cycle regime.
• At high loading conditions, Al Alloy 7075 has greater fatigue strength than Al Alloy 2024, but there is no noticeable difference at low loads.
• At log (no. of cycles) 5.0 to 6.5, i.e., the area which carries both the welded sample results and the base sample results, correlate with each other, and the welded samples show better fatigue strength than the base samples. However, the superiority of the welded samples is only scattered in the middle of the horizontal axis, while at the lowest and the highest limits of the number of cycles, base samples have considerable fatigue life, but still less than those of the welded samples.
• The filler material AA4047 shows higher joint strength in comparison to AA4043 due to high silicon concentration. High silicon concentration leads to low melting temperature range, hence, giving high resistance to solidification cracking and undercut. It results in low porosity and inclusion, increasing the fatigue strength of AA4047. The melting range of AA4043 is 60 degrees, while that of AA 4047 is 6 degrees. Therefore, 4047 is more crack resistant. Magnesium silicide Mg₂Si is also present in AA4047 filler material due to the higher concentration of magnesium in it and, hence, increases its tensile strength and bonding forces in the material. Some deviation from the suggested results can be seen in Figure 8 at point 3 and 4, either due to the formation of any welding defect or welding pool instability. One of the reasons could be the absence of magnesium silicide, which reduces crack sensitivity. Although the trend line of AA4047 is higher than AA4043 in all cases, referring to AA4047 as better filler material, it is important to note here that the real-time fatigue testing is carried out in the current study. Generally, fatigue results show high variability in the results, even when identical test pieces are used, so some abnormality in the results is typical in fatigue testing [44,45,46].