dc.description.abstract | Abstract
In order to reduce the interference with the surface traffic, tunnelling methods have been widely used in the construction of metro and sewage systems in Taipei. However, these tunnels were driven through the ground where structures and lifelines already existed. Therefore, it is very important for engineers to evaluate the impacts of tunnelling on the environment, i.e. the effect of surface settlement on adjacent buildings and the influence of subsurface ground movements on underground pipelines. Besides, because the tunnels in cities are usually driven in limited areas, the down track tunnel and up track tunnel are parallel and very close to each other. Hence, it is necessary to evaluate the settlements induced by parallel tunnelling and the interaction between two tunnels.
In this study, both centrifuge model tests and a finite difference program (FLAC) have been taken in plane strain conditions to investigate the tunnel stability and soil movements for both the single tunnel and parallel tunnels. A series of model tests was performed in centrifugal acceleration of 100g, including the single tunnel with various cover-to-diameter ratio (C/D) and the parallel tunnels with various C/D and distance-to-diameter ratio (d/D). The supporting air pressure was applied inside the tunnels to keep the tunnels in equilibrium during the accelerating stage, and then the supporting pressure was reduced gradually to simulate the process of excavation. From centrifuge modelling and numerical modelling, the ground movements and the changes of soil stresses and pore water pressures could be studied. Besides, 15 field cases of measured surface settlements due to shield tunnelling in Taipei were used to verify the predictions from centrifuge modelling and numerical modelling.
From test results, the relationships of both the volume of ground loss versus the maximum settlement and the burial depth of tunnels versus the width parameter of the settlement trough have been proposed. Since the volume of ground loss could be evaluated prior to tunnel construction and the burial depth was determined in the preliminary design stage, the settlement troughs induced by single tunnelling and parallel tunnelling could be predicted. Using empirical methods, the settlement troughs can be reasonably fit by a Normal Distribution Curve suggested by Peck (1969). By superimposing two curves with a factor of 0.85 times the width parameter fit from the single tunnel, the settlement trough due to parallel tunnelling with d/D less than 2 may be estimated. For the parallel tunnels with d/D greater than 2, the simple superposition of two settlement curves from single tunnel is suggested. Besides, the collapse mechanisms of both single tunnel and parallel tunnels derived from upper bound theorem compared well with the observed ground movements. From 15 case studies in field, the predicted settlements show reasonable agreement with the field data. Hence, the prediction from proposed formula is verified.
In addition, the program FLAC was carried out to solve the whole problem in the same material and boundary conditions as those in centrifuge tests. The surface settlement troughs computed from FLAC show reasonable agreement with results measured from model tests. According the pore water pressures measured from model tests and the soil stresses computed from numerical analysis, the extent of soil arching for single tunnel and parallel tunnels could be defined. It may be adopted to explain the transferring mechanism of soil stresses around tunnels and the interaction between parallel tunnels. If the underground facilities existed within the extent of soil arching, the influence of tunnelling to those structures should be taken into account. Based on test results, numerical modelling and theoretical solutions, the lower bound and upper bound of overload factors are determined for evaluation of tunnel stability. From test results verified by numerical modelling and field measurement, this study may provide reliable predictions of ground movements induced by single tunnelling and parallel tunnelling. | en_US |