Aircraft skin and stringer elements are typically fabricated from 2xxx and 7xxx series high strength aluminum alloys. A single friction stir welding (FSW) pass using a specially designed tool with shoulder/pin diameter ratio (D/d) of 3.20 is used to produce dissimilar T-butt welds between AA2024-T4 and AA7075-T6 aluminum alloys at a constant travel speed of 50 mm/min and different rotational speeds of 400, 600 and 800 rpm. The AA2024-T4 is the skin and the AA7075-T6 is the stringer. Sound joints are produced without macro defects in both the weld top surfaces and the joint corners at all rotational speeds used (400, 600, and 800 rpm). The hardness value of the nugget zone increases by increasing the rotational speed from 150 ± 4 Hv at 400 rpm to 167 ± 3 Hv at 600 rpm, while decreases to reach the as-received AA2024-T4 hardness value (132 ± 3 Hv) at 800 rpm. Joint efficiency along the skin exhibits higher values than that along the stringer. Four morphologies of precipitates were detected in the stir zone (SZ); irregular, almost-spherical, spherical and rod-like. Investigations by electron back scattered diffraction (EBSD) technique showed significant grain refinement in the sir zone of the T-welds compared with the as-received aluminum alloys at 600 rpm due to dynamic recrystallization. The grain size reduction percentages reach 85 and 90 % for AA2024 and AA7075 regions in the mixed zone, respectively. Fracture surfaces along the skin and stringer of T-welds indicate that the joints failed through mixed modes of fracture.

The current work presents a detailed investigation for the effect of a wide range friction stir welding (FSW) parameters on the dissimilar joints’ quality of aluminum alloys. Two groups of dissimilar weldments have been produced between AA5083/AA5754 and A5083/AA7020 using tool rotational rates range from 300 to 600 rpm, and tool traverse speeds range from 20 to 80 mm/min. In addition, the effect of reversing the position of the high strength alloy at the advancing side and at retreating side has been investigated. The produced joints have been investigated using macro examination, hardness testing and tensile testing. The results showed that sound joints are obtained at the low heat input FSW parameters investigated while increasing the heat input results in tunnel defects. The hardness profile obtained in the dissimilar AA5083/AA5754 joints is the typical FSW hardness profile of these alloys in which the hardness reduced in the nugget zone due to the loss of the cold deformation strengthening. However, the profile of the dissimilar AA5083/AA7020 showed increase in the hardness in the nugget due to the intimate mixing the high strength alloy with the low strength alloy. The sound joints in both groups of the dissimilar joints showed very high joint strength with efficiency up to 97 and 98%. Having the high strength alloy at the advancing side gives high joint strength and efficiency. Furthermore, the sound joints showed ductile fracture mechanism with clear dimple features mainly and significant plastic deformation occurred before fracture. Moreover, the fracture in these joints occurred in the base materials. On the other, the joints with tunnel defect showed some features of brittle fracture due to the acceleration of the existing crack propagation upon tensile loading.

No abstract

This work investigated the effect of friction stir welding (FSW) tool rotation rate and welding speed on the grain structure evolution in the nugget zone through the thickness of the 10 mm thick AA5083/AA5754 weldments. Three joints were produced at different combinations of FSW parameters. The grain structure and texture were investigated using electron backscattering diffraction (EBSD). In addition, both the hardness and tensile properties were investigated. It was found that the grain size varied through the thickness in the nugget (NG), which was reduced from the top to the base in all welds. Reducing the rotation rate from 600 rpm to 400 rpm at a constant welding speed of 60 mm/min reduced the average grain size from 33 µm to 25 µm at the top and from 19 µm to 12 µm at the base. On the other hand, the increase of the welding speed from 20 mm/min to 60 mm/min had no obvious effect on the average grain size. This implied that the rotation rate was more effective in grain size reduction than the welding speed. The texture was the mainly simple shear texture that required some rotations to obtain the ideal simple shear texture. The hardness distribution, mapped for the nugget zone, and the parent alloys indicated a diffused softened welding zone. The heating effect of the pressure and rotation of the pin shoulder and the heat input parameter (ω/v) on the hardness value of the nugget zone were dominating. Tensile stress-strain curves of the base alloys and that of the FSWed joints were evaluated and presented. Moreover, the true stress-true strain curves were determined and described by the empirical formula after Ludwik, and then the materials strengthening parameters were determined. The tensile specimens of the welded joint at a revolution speed of 400 rpm and travel speed of 60 mm/min possessed the highest strain hardening parameter (n = 0.494).

Currently, the world is still facing a COVID-19 (coronavirus disease 2019) classified as a highly infectious disease due to its rapid spreading. The shortage of X-ray machines may lead to critical situations and delay the diagnosis results, increasing the number of deaths. Therefore, the exploitation of deep learning (DL) and optimization algorithms can be advantageous in early diagnosis and COVID-19 detection. In this paper, we propose a framework for COVID-19 images classification using hybridization of DL and swarm-based algorithms. The MobileNetV3 is used as a backbone feature extraction to learn and extract relevant image representations as a DL model. As a swarm-based algorithm, the Aquila Optimizer (Aqu) is used as a feature selector to reduce the dimensionality of the image representations and improve the classification accuracy using only the most essential selected features. To validate the proposed framework, two datasets with X-ray and CT COVID-19 images are used. The obtained results from the experiments show a good performance of the proposed framework in terms of classification accuracy and dimensionality reduction during the feature extraction and selection phases. The Aqu feature selection algorithm achieves accuracy better than other methods in terms of performance metrics.