The behaviour of flowthrough rockfill dams.

This thesis deals with the hydraulics and stability of "flowthrough" rockfill embankments. This thesis provides new information in the following seven areas: (1) A multi-disciplinary consideration of most of the available equations describing one-dimensional non-Darcy flow. Such flow is characterized by a non-linear dependence of velocity on the hydraulic gradient. New data is presented to assist in the evaluation of the effect of particle irregularity, as compared to perfect spheres. (2) A new computational method to estimate the quantity of flow moving through a rockfill embankment, as a function of the upstream water level, is presented. (3) If some minimum upstream water level is required, an impermeable element (in the form of a wall-like hydraulic barrier) will be needed. The hydraulics of an upstream impermeable facing (UIF) was studied in this research program. Some guidance is presented as to how to account for the effect of an elevated downstream water level on the stage-discharge rating curve of a flowthrough rockfill embankment. (4) Three distinct analytical methods for computing the position of the phreatic line are presented and evaluated. The validity of the Dupuit assumption, which is inherent in all three of the analytic methods, is also discussed. (5) A detailed computational framework is presented to make possible the prediction of the minimum elevation over which mesh protection is needed to prevent erosion of the downstream face of the dam. (6) The possibility of massive deep-seated failure was investigated in detail. As a first step to permit such evaluations, non-Darcy pore-pressure modelling was performed. An efficient method for executing such modelling on an electronic spreadsheet is presented. In an innovative series of experiments it was found that the measured bursting force at incipient failure was far closer to that theoretically predicted by the wedge method than that predicted by a classical rotational slip-circle method (Bishop's Simplified Method). (7) In another innovative series of experiments, this low-temperature behaviour was investigated using the Low Temperature Laboratory of the National Research Council of Canada. Based on this research a simple method for estimating the rate of growth of ice at the phreatic surface is presented. Ice was not observed to form in the voids within the body of the dam, except in one case in a small mixing zone (the base of Parkin's zone 3) below the lip of the UIF. A detailed theoretical framework for performing heat budget calculations is also presented, together with a short computer program to facilitate these calculations. It is quantitatively demonstrated that the viscous dissipation of hydraulic energy plays an important role in preventing ice formation in the voids. (Abstract shortened by UMI.