| The objective of this dissertation is to present a systematic procedure, referred to as "information feedback analysis", that may be used to predict induced deformation by collecting useful field information during the excavation. Information feedback analysis is especially useful for staged construction such as deep excavation. Having good field instrumentation, the displacements and stresses can be accurately measured throughout the entire construction process. These measured values, which serves as input information, are then entered into a predefined numerical model of the deep excavation. With the help of appropriate optimization algorithms, this analysis will result in a set of soil parameters that makes the calculated deformation closest to the measured values. These back-calculated soil parameters reflect the overall effects of the field conditions. The next step is to use these back-calculated parameters to make the prediction of possible deformation in subsequent stages until the completion of the excavation. Similar procedures are performed as the excavation proceeds stage by stage, which results in additional updated information continuously entered into the system and hence, the prediction becomes progressively more accurate.; The study has shown that with the information feedback analysis implemented in deep excavation systems there are at least two advantages over the conventional analysis. First of all, the use of field instrumentation results to estimate geotechnical parameters allows the engineer to account for the global response of a soil-structure system. Second since the information is collected throughout the construction, any departure from the original design should be reflected by the updated information. While conventional analyses are conducted in the design stage and always assume that construction proceeds in a "designated" fashion. |
