Centrifugal model tests were performed to investigate the effectiveness of settlement control using geogrid-reinforced pile-supported embankments for high-speed railways on collapsible loess. First, the acceptability of using remoulded loess to simulate the behaviour of undisturbed collapsible loess in wetting test was examined in centrifuge tests. Then, the settlement of geogrid-reinforced pile-supported embankments under different conditions was investigated using remoulded loess. Experiments results showed that the collapse deformation process of remoulded and undisturbed collapsible loess can be divided into three stages with respect to time. The settlement of embankment built on collapsible loess foundation without improvement does not meet the post-construction settlement constraint for embankment of ballast track high-speed railways. Use of rigid piles, with or without caps, can significantly reduce embankment settlement and the rate of post-construction settlement. It was also observed that the negative skin friction developed in the piles varies with pile spacing and the pile configurations.

Based on the study at home and abroad, the theoretical calculation, numerical analysis is used to analyze pile-net composite foundation in different parameters, to research and analyze the settlement controlling ability of rigid pile composite foundation on soft clay ground in high speed railway. The research results show that settlement basin is in surface of pile (cap) net composite foundation, and uneven settlement of pile raft composite foundation. In a certain range of subgrade slope, the settlement of pile raft composite foundation is greater than that of pile (cap) net composite foundation; The increasing of pile spacing has obvious influence for the settlement of reinforced area, but the settlement of the area of underlying stratum has little change; The total settlement influence of pile cap size is not obvious. When the thickness of soft soil is more than 20m, pile raft composite foundation has economic and technological advantages.

The post-construction settlement of subgrade is strictly limited in high-speed railway. Therefore, pile-net composite foundation is widely used for high-speed railway. The settlement calculation method with pile-net composite foundation is a common method which is used for calculation of pile-slab composite foundation at present. Based on both theoretical research and testing analysis, the results show that significant difference exists in the composite ground between the flexible and rigid foundation. The proposed settlement calculation method of rigid pile-net composite foundation is discussed for the first time with the current code in this paper. The analysis shows that the current code method cannot meet the actual situation under over-consolidated clay. Using the method of rigid pile composite foundation, the distribution of stress is derived according to German Recommendations for Reinforcements with Geosynthetics EBGEO (DGGT 2003). Using the Mindlin-Boussinesq uniting method can calculate the additional stress in the ground. Finally, settlement of the reinforced area and underlying stratum can be calculated by the method of e-lgp curve. This method can also be used to analyze parameters that influence the settlement, such as the pile spacing, the condition of pile cap, the length of pile, and stress history of soil. The distribution of soil settlement and additional stress with depth correspond to the actual situation, and the calculation results of settlement show that the new method is in accordance with the in-situ data.

The corrosion of reinforced concrete could reduce its mechanical properties and cause concrete cracking. Electrophoretic deposition is a new method for repairing rust-cracked reinforced concrete. In this study, we observed the crack surface morphology, mass growth, epoxy resin film thickness, ultrasonic rate, and epoxy resin filling depth of rust-cracked reinforced mortar specimens during the repair process and investigated internal microstructure and phase composition of the repaired mortar, and clarified the principle behind the electrophoretic deposition of cationic epoxy resin and curing agent molecules in the repair solution. The result of research shows that longer repair times decreased the carbonization depth and water absorption of the specimens. After repair, the porosity of the surface mortar decreased, owing to the filling of epoxy resin, and the newly generated epoxy resin film had a dense structure. Thus, the carbonization resistance and waterproof performance of the repaired specimens improved. The repair process of electrophoretic deposition can be divided into the immersion, repair, and curing stages, and the physical and chemical principles of each stage were explained.