Experimental Investigation of Piping Potential in Earthen Structures

Abstract

Current methods for evaluation of piping potential have not been successful in preventing piping failures. The initiation of classic backwards-erosion piping failure is currently assessed from the anticipated hydraulic gradient at the point of seepage exit, using a theoretical method originally developed by Karl Terzaghi in 1922. Recently, some researchers suggested that there may be a relation between effective stress and the critical hydraulic gradient and recommended further investigations. To date, there has been little research performed to evaluate constitutive behavior of soil that relate to the initiation and progression of backwards erosion piping. In order to evaluate piping potential, a series of laboratory experiments were conducted. A true-triaxial load cell was developed and used for the testing. The load cell was designed to provide the flexibility to modify loading conditions along three orthogonal axes, and to permit loading the cell with pressurized water. The parameters investigated with respect to pipe initiation are: (1) pipe initiation behavior under variable stress tensors; (2) effect of exit geometry on piping potential; (3) effect of load path on piping potential, (4) pipe initiation behavior under variable seepage stress rates. Preliminary test results confirm that there is an energy component in pipe initiation that currently is not adequately considered in piping evaluations and that the exit velocity is a better predictor of piping potential than the hydraulic gradient. Hydraulic exit losses were found to play a key role in pipe initiation. The critical hydraulic gradients determined in these horizontal flow tests are lower than standard theory would predict. A weak relationship between confining stresses and critical hydraulic gradients was observed at higher confining stresses.