Tensile strength is one of the important mechanical parameters for controlling the development of cracks in the Loess Plateau. However, the tensile strength of soil is difficult to be precisely measured due to the lack of satisfying laboratory techniques. Unconfined penetration test (UPT) is commonly used as an indicator to predict soil tensile strength characteristics, while there has been less research study about test parameter values range of UPT and micromechanical characteristics at failure. This study finished 50 groups of UPTs of remolded loess with different disk diameters (12.36–27.81 mm) at various sample heights (37–87 mm) based on Q₂ loess in Lintong areas, and the crack development and failure essence in the different failure modes process were investigated by the discrete element modeling. The results showed that disk diameters and sample height were closely correlated with the failure mode. The failure of remolded loess in the UPT is not only tensile failure, which also can occur compression failure and tensile–compression coupling failure. Among them, when the ratio of disk diameters to sample height is about in the range of 0.189–0.309, the sample appears tensile failure. In addition, the ratio of compressive and tensile crack numbers in the sample is 0.16–0.22 in the tensile failure mode. The numerical simulation method provides a way for the observation of failure process and crack development, thereby improving the understanding of failure essence in UPT. The findings will provide a reference way to improve test theory and method of loess tensile strength.

Pile reuse is a common technique in bridge renovation projects. However, the interaction mechanisms between new and existing piles under a shared cap remain unclear, restricting the size setting and further optimization of new piles in existing pile foundation environments. This study analyzes the effects of key parameters for new piles on the settlement behavior of existing piles under a shared pile cap using field measurement data and simulation results. The findings indicate that, within one pile cap, the settlement of both new and existing piles exhibits a negative correlation with the increasing new pile length. With the different load distribution patterns, the settlement differences between new and existing piles tend to be more stable in a lateral arrangement compared to a symmetrical distribution. Additionally, the pile cap size has a boundary effect on the combined pile system, specifically, as the pile cap length/width ratio is 4:2, the settlement disparity between new and existing piles tends to stabilize. Settlement behavior is also significantly affected by soil properties, with stiffer soils (higher elastic modulus) showing smaller settlements. Introduce the existing pile efficiency parameter, the main factors influencing settlement behavior rank as follows: soil properties, load distribution, pile distribution, pile length, pile diameter, and pile cap size. Based on these findings, it is recommended that the length of new piles be controlled to 1.0–1.1 times the length of existing piles, and the diameter of new piles be 1.0–1.2 times the diameter of existing piles. The study explores the interaction effects between new and existing piles, aiming to optimize the performance of pile reuse.