3D character of backward erosion piping: Small-scale experiments

Abstract

Backward erosion piping is an important failure mechanism for cohesive water retaining structures which are founded on a sandy aquifer. At present, the prediction models for safety assessment are often based on 2D assumptions. In this work, the 3D character of the phenomenon is demonstrated on the basis of small-scale experiments. Our approach reveals the correlation between the occurrence of piping initiation and progression and the width of the physical model, which is a measure for the inclusion of the third dimension (to be regarded in a real dike situation as the influence zone of a crater). In addition, it was found that the model width has an impact on pipe characteristics and pipe development. Therefore, our results enable a better understanding of the complex physical mechanism related to backward erosion piping and thus can lead to a significant improvement in the safety assessment of water retaining structures.

Use of incipient motion data for backward erosion piping models

“…The conventional test equipment to assess backward erosion piping is a rectangular box filled with sand with either a slope, ditch, hole or plane type exit and a transparent cover (among others De Wit, 1984;Hanses, 1985;Robbins et al, 2016;Silvis, 1991;Townsend et al, 1988;Van Beek et al, 2011;Vandenboer et al, 2014), in which the sand bed is subjected to a horizontal hydraulic gradient until piping occurs. Although this set-up has the advantage of allowing the pipe to migrate laterally to find the weakest path, the drawback is that hydraulic conditions cannot be assessed easily as it is unknown where, precisely, the pipe will develop.…”

Analysis of the pipe depth development in small-scale backward erosion piping experiments

“…A calibration allowed for empirical adjustment to take account of various influential parameters [17]. Nevertheless, the Sellmeijer model fails to predict the critical gradient for different constraints (such as larger grain sizes and 3D flow [27]).…”

“…The large number of available models at present with many differences in the basic hypotheses indicates that the true incentive of backward erosion piping is not yet captured. Up to now, experimental studies have led to indispensable knowledge on key aspects of backward erosion piping either by analyzing the critical gradient or by studying the pipe formation in the sand bed: [7,9,20] identified the different phases involved and described the (meandering) character of the pipes; [3] investigated the influence of the different downstream exit configurations, sand layer dimensions, and sand types on the critical gradient; [13] studied the erosion and fluidization in an outflow opening; [17,19] considered a large number of sand types to identify the influence of relative density, uniformity, roundness, permeability, and grain size on the susceptibility to backward erosion piping; [26,27] demonstrated that backward erosion piping should be treated as a three dimensional phenomenon rather than a two dimensional problem both in terms of groundwater flow and pipe development; [23] studied the variation of pipe widths in relation with the grain size; [7,24] indicated that progression is more likely to be driven by local detachment of the particles at the pipe tip, rather than erosion of particles at the pipe bottom.…”