In the present study, 2198 Al-Cu-Li alloys were successfully friction stir welded by using various welding speed ranges of 90~180 mm/min with an invariable rotation speed of 950 r/min. The effect of welding speed on microstructure evolution and mechanical properties of the joints was investigated. The results show that, with the welding speed decreasing, the size of the nugget zone (NZ) first increases and then decreases due to different welding temperatures. At a welding speed of 150 mm/min, the size of the NZ in all joints is the biggest and the “S” curve disappears. The equiaxed grains are finer, attributed to a higher degree of dynamic recrystallization, and a larger number of fine reprecipitated phase (δ’, β’ phases) particles are dispersively distributed in the NZ. Correspondingly, the joints have the highest tensile properties, and the tensile strength, yield strength and elongation are, respectively, 406 MPa, 289 MPa and 7.2%. However, compared to the base material, the tensile properties of all joints are reduced because a greater amount of δ’ and β’ phases particles are dissolved in the NZ. Only the joints produced at 150 mm/min are fractured in the TMAZ with detected deep dimples and tearing ridges, and a significant necking phenomenon is observed, which indicates a complete ductile fracture mode.

Simulating fuel sloshing within aircraft tanks during flight is crucial for aircraft safety research. Traditional methods based on Navier-Stokes equations are computationally expensive. In this paper, we treat fluid motion as point cloud transformation and propose the first neural network method specifically designed for simulating fuel sloshing in aircraft. This model is also the first deep learning model capable of stably modeling fluid particle dynamics in such complex scenarios. Our triangle feature fusion design achieves an optimal balance among fluid dynamics modeling, momentum conservation constraints, and global stability control. Additionally, we constructed the Fueltank dataset, the first dataset for aircraft fuel surface sloshing. It comprises 320,000 frames across four typical tank types and covers a wide range of flight maneuvers, including multi-directional rotations. We conducted comprehensive experiments on both our dataset and the take-off scenario of the aircraft. Compared to existing neural network-based fluid simulation algorithms, we significantly enhanced accuracy while maintaining high computational speed. Compared to traditional SPH methods, our speed improved approximately 10 times. Furthermore, compared to traditional fluid simulation software such as Flow3D, our computation speed increased by more than 300 times.