Concrete gravity dams are usually modeled as 2D structures for analysis. However, these modeling assumptions are not always acceptable, especially when they are situated in V-shaped valleys or on narrow canyons with stepped or tapered monoliths. in such situations, a 3D model has been found to give more reliable results. However, a 3D model increases the complexity of the problem in addition to the increase in computation time. Hence, as an alternative to the conventional 2D model, wherein plane stress elements of uniform thickness are used, a new modified 2D model composed of plane stress elements of different thicknesses across the cross section so as to take into account the tapering or stepped geometry is adopted in the present study. its effectiveness in representing such a stepped or a tapered monolith is studied by comparing the vertical stresses induced at its base with that of 3D and the conventional 2D model. Non-overflow section of a typical concrete gravity dam monolith of varying area of cross section is used for this purpose. The stresses in the 3D model from gravity and response spectra analysis was found to be significantly higher than those in 2D models. The 3D model also showed predominant out of plane and torsional modes which both the 2D models were not capable of. In spite of the use of the modified 2D model, there was only a slight improvement in the results. Hence, in these situations a full 3D model should be adopted in order to obtain reliable results.
Dam, seismic analysis, modal analysis, finite element method, ANSYS
