Fei Gan

The unloading pile-sheet retaining wall is a new type of support/retaining for embankment slopes and has demonstrated excellent performance in engineering applications. However, its mechanical behavior and operational mechanisms are still not fully understood. This study conducted model tests on unloading piles and cantilever piles. Backfilling was conducted in four stages: 30 cm for the first three layers and 10 cm for the final one. The tests focused on the evolution of earth pressure, internal forces, and deformation in both pile types. The results show that: 1) Upon completion of backfilling, the horizontal displacement at the top of the cantilever pile reaches 81.76 mm, which is 5.45 times that of the unloading pile (14.99 mm). The maximum earth pressure on the unloading pile is 10.08 kPa, accounting for 66.40% of the 15.18 kPa recorded on the cantilever pile. The unloading effect alters the distribution pattern and magnitude of earth pressure. 2) The bending moment distributions differ significantly. The cantilever pile exhibits a “fish-belly” pattern with a maximum moment of 115.8 N· m. In contrast, the unloading pile shows an “S-shaped” profile, featuring a pronounced point of contraflexure at the unloading platform and a maximum negative moment of −60.99 N· m. 3) Incorporating an unloading platform effectively reduces earth pressure and enhances the anti-overturning moment. These effects jointly improve sliding resistance and overall structural stability. These findings offer theoretical insights and technical guidance for the practical implementation of unloading pile-sheet retaining walls.

Understanding fault activation and slip is vital for earthquake mechanics and underground structure stability. Fault stability, beyond in situ stresses, depends on their architecture, including asperities, segments, and gouge materials. This study investigates locked fault segments of varying lengths using mortar specimens with prefabricated faults filled with montmorillonite and quartz sand gouge. Uniaxial loading was applied, with embedded strain gauges and digital image speckle techniques monitoring internal stress, surface strain, and displacement near the fault. Results show increased model strength with longer locked segments. Local stress field deflection angles vary more in the dilatation quadrant than the compression quadrant. Gouge material disperses stress concentrations during loading. Slip occurs along the structural surface during failure, with the relative sliding rate of blocks correlating strongly with stress field deflection angle changes. As the system transitions from elastic to stable fracture stages, this correlation in the dilatation quadrant shifts. These findings offer insights into complex fault mechanisms and stability compared to simple planar saw‐cut discontinuities.

Inclined piles are increasingly adopted in the foundation design of large-scale transmission line towers. Nine inclined pile tests were conducted in red clay soil to investigate the load-bearing and deformation behavior and the load transfer mechanism of inclined piles under vertical loading. The effects of pile inclination and length-to-diameter ratio on axial force, bending moment, shear force, and lateral friction force were analyzed. The test results indicate that: (1) As the inclination angle increases, the settlement of the inclined pile increases under vertical loading. (2) The axial force in the inclined pile is smaller compared to that of the corresponding vertical pile with the same material and dimensions. (3) Bending moment and shear force are observed in the upper section of the inclined pile, with the maximum bending moment influenced by both the pile’s inclination angle and length-to-diameter ratio. (4) In the upper section of the pile, the average soil-side frictional resistance of inclined piles is higher than that of vertical piles. The maximum resistance occurs in the section from the top of the pile to a relative depth (Z/L) of approximately 0.14.

To address the challenges of significant scale variations and difficulty in extracting discriminative features for small targets in airborne remote sensing object detection, while considering the constraints and efficiency requirements of onboard systems, this article proposes a lightweight detection algorithm. First, a structural reparameterization strategy is applied to optimize depthwise separable convolutions, simplifying complex structures during the inference stage, which improves inference speed and memory utilization. Second, the G-Shuffle module is designed to significantly enhance feature extraction efficiency and interchannel information interaction, balancing computational complexity and detection accuracy. Lastly, a dynamic multiscale pyramid network is introduced, employing pyramid convolution to effectively extract and fuse multiscale features. Additionally, a channel attention network is integrated to establish channel interconnections, expanding the receptive field and enhancing the extraction of finegrained features at different scales, thereby improving the detection capability for targets of various sizes, especially small targets. Experimental results demonstrate that the proposed method achieves a mAP@0.5 of 0.552 and mAP@0.5:0.95 of 0.265 on the AI-TOD dataset, with a 2.7 ms improvement in inference speed and a 14% reduction in model parameters. These results indicate superior performance across multiple metrics compared to existing state-of-the-art methods, proving the method's potential and practicality for airborne remote sensing object detection.