Search

The mechanical properties of loess-steel interface are of great significance for understanding the residual strength and deformation of loess. However, the undisturbed loess has significant structural properties, while the remolded loess has weak structural properties. There are few reports on the mechanical properties of loess-steel interface from the structural point of view. This paper focused on the ring shear test between undisturbed loess as well as its remolded loess and steel interface under the same physical mechanics and test conditions (water content, shear rate and vertical pressure), and explored the influence mechanism of structure on the mechanical deformation characteristics of steel-loess interface. The results show that the shear rate has little effect on the residual strength of the undisturbed and remolded loess-steel interface. However, the water content has a significant influence on the residual strength of the loess-steel interface, moreover, the residual internal friction angle is the dominant factor supporting the residual strength of the loess-steel interface. In general, the residual strength of the undisturbed loess-steel interface is greater than that of the remolded loess specimen (for example, the maximum percentage of residual strength difference between undisturbed and remolded loess specimens under the same moisture content is 6.8%), which is because that compared with the mosaic arrangement structure of the remolded loess, the overhead arrangement structure of the undisturbed loess skeleton particles makes the loess particles on the loess-steel interface re-adjust the arrangement direction earlier and reach a stable speed relatively faster. The loess particles with angular angles in the undisturbed loess make the residual internal friction between the particles greater than the smoother particles of the remolded loess (for example, the maximum percentage of residual cohesion difference between undisturbed and remolded loess specimens under the same vertical pressure is 4.29%), and the intact cement between undisturbed loess particles brings stronger cohesion than the remolded loess particles with destroyed cement (for example, the maximum difference percentage of residual cohesion between undisturbed and remolded soil specimens under the same vertical pressure is 33.80%). The test results provide experimental basis for further revealing the influence mechanism of structure, and parameter basis for similar engineering construction.

China is shifting from the stage of large-scale bridge construction to the stage of equal emphasis on the construction and maintenance of bridges. The problems of low efficiency and high cost in bridge inspection are becoming more and more prominent, which threaten people’s life safety. In this paper, the deterioration state prediction model of concrete beam bridge under Boosting DT C5.0 was established as the basis, and optimization suggestions were put forward in terms of bridge inspection standards and processes, which aims to perfect the bridge inspection mechanism, realize the fine management of the bridge and prolong the service life of the bridge. Research shows that: first, the bridge inspection standard with a single index should be improved into the bridge inspection standard with multiple indexes, so as to scientifically determine the bridges that need to be inspected and optimize the allocation of maintenance resources. Second, the bridge deterioration state prediction model is used to add a screening mechanism for the bridge in the inspection plan, which can effectively reduce the workload of bridge inspection and enhance the inspection efficiency. Third, the deterioration phenomenon of coexistence between adjacent traffic assets should be fully considered and the linkage inspection mechanism of adjacent traffic assets should be established to improve the effect of bridge inspection.

Based on the engineering practice of large cross-section highway tunnel, this paper reveals the space-time coordinated evolution law of the construction mechanical characteristics and deformation distribution of the support structure in the construction by half bench CD method through field test. At the same time, the mechanical response calculation model of the supporting structure in the partial excavation is constructed, and the mechanical characteristics of the support structure in the partial excavation process are analyzed by above mechanical calculation model. Then, the mechanical and deformation distribution of the feet-reinforcement bolt in the steel frame—foot-reinforcement bolt combined support system is analyzed under different levels of surrounding rock load and different structural parameters of the feet-reinforcement bolt. The research results show that: (1) The internal force of the supporting structure changes most obviously during the excavation of Part Ⅰ, Part Ⅱ and Part Ⅲ, and the internal force of the support structure gradually tends to be stable after a slight increase in the excavation of Part Ⅳ and Part Ⅴ; (2) The horizontal deformation and vertical deformation of the support structure mainly occur in the excavation process of Part Ⅰ, Part Ⅱ and Part Ⅲ, and the excavation of Part Ⅳ and Ⅴ has little effect on the deformation response of the structure. The vertical displacement of the supporting structure is larger than the horizontal displacement, and the dynamic response of the temporary diaphragm structure during tunnel excavation is shrinkage-expansion-shrinkage-expansion; (3) The bending strain of each measuring point decreases with the increase of the distance from the loading point, and the bending strain of section 1 and section 2 is much larger than that of the other three sections; (4) With the increase of the angle, the section position with strain close to 0 gradually moves to the deeper position of the bolt, and the axial strain of each section on the bolt gradually changes from positive strain to negative strain.

Given multiple source datasets with labels, how can we train a target model with no labeled data?Multi-source domain adaptation (MSDA) aims to train a model using multiple source datasets different from a target dataset in the absence of target data labels. MSDA is a crucial problem applicable to many practical cases where labels for the target data are unavailable due to privacy issues. Existing MSDA frameworks are limited since they align data without considering labels of the features of each domain. They also do not fully utilize the target data without labels and rely on limited feature extraction with a single extractor. In this paper, we proposeMulti-EPL, a novel method for MSDA.Multi-EPLexploits label-wise moment matching to align the conditional distributions of the features for the labels, uses pseudolabels for the unavailable target labels, and introduces an ensemble of multiple feature extractors for accurate domain adaptation. Extensive experiments show thatMulti-EPLprovides the state-of-the-art performance for MSDA tasks in both image domains and text domains, improving the accuracy by up to 13.20%.