Search

Temperature field data from the steel truss arch girder and longitudinal displacement data from six groups of rubber bearings were collected by the structural health monitoring system for the Dashengguan Yangtze Bridge. By using long-term monitoring data, two correlations are investigated: the linear correlation between the longitudinal displacement and the temperature field (including uniform temperature and temperature gradient); and the linear correlation of the longitudinal displacements in different locations. A multivariate linear regression equation is used to model the first correlation, and a Lagrange polynomial interpolation is used to model the second correlation. The final mathematical models, representing the healthy state of the bearings, can be applied to simulate the longitudinal displacements of the main girder. Furthermore, the change regularity of longitudinal displacements for the degraded rubber bearings is revealed taking advantage of a hysteretic model and presumed envelope curves of frictions caused by bearing degradation. A method of detection and location for the degraded bearings is proposed in four detailed steps, and the numerical results demonstrate that this method is effective. [ABSTRACT FROM AUTHOR]

To study the collapse characteristics of high-density silt under complex loads, three series of hollow-cylinder tests were conducted on high-density saturated silt samples taken from the sea entrance of the Yangtze River. In the first series of tests, samples with high density were isotropically consolidated and then sheared under undrained triaxial loads. In subsequent tests, samples with low and high densities were isotropically consolidated and then cyclically sheared by principal stress rotation of 0-180°. Collapse and liquefaction were clearly observed in all the samples tested under cyclic principal stress rotation when the pore water pressure was equal to the initial effective confining pressure. Two critical points, the graded phase transformation point and the cataclysmic phase transformation point (collapse point), were observed for the high-density samples before liquefaction; these divided the buildup of the pore water pressure into three or two stages depending the shear stress level. At low shear stresses, the characteristics of the transformation points were hardly influenced by load frequency. The isotropic consolidated high-density silt and low-density silt expressed similar behaviors. However, in the high-density silt, the strain development could be divided into two stages by the collapse point, which corresponds to the cataclysmic phase transformation point of the pore water pressure representing the state of structure collapse. The deviator strains were limited within a narrow range of 0.2-0.4% at the collapse state. The stress state of collapse can be normalized by the quasi-unstable line in the p'-q space. Finally, we introduce the modified Seed model to successfully estimate the development of pore water pressure in high-density silt. [ABSTRACT FROM AUTHOR]

A computational case study is presented of the flooding in urban Jiujiang due to the Yangtze River dike break in August 1998. A shallow water hydrodynamic model was deployed based on the weighted average flux -- total variation diminishing (WAF-TVD) finite volume algorithm. A series of numerical modelling was conducted to identify the influences of the mesh resolution and model parameterisation on the flood propagation. Three mesh resolutions (2, 5 and 10 m) were used, and spatially distributed roughness was specified based on land-use classification. Due to the highly irregular topography and densely distributed buildings, the flood propagation is largely hindered. The numerical results show that an area of merely 6 km² was inundated at 6 h following the dike break with a discharge of 400 m3/s. Mesh resolution and roughness parameterisation have considerable effects on the flood propagation, whereas the impact of turbulent eddy viscosity is negligible. A coarser mesh could result in a larger inundation area (10%), and a lumped roughness may produce an appreciable overestimation of the inundation area in the initial period of the flood propagation. This observation highlights the significance of mesh resolution and also physically meaningful estimation of the roughness for urban flood modelling. [ABSTRACT FROM AUTHOR]