A neural network assessment tool for estimating the potential for backward erosion in internal erosion studies

A new technique and approach for characterizing piping problems using artificial neural networks is introduced. Recent studies suggest that most large dam failures are a result of internal erosion, although this particular failure mechanism is often less highlighted in the standard procedures for dam designs. Over toppling, foundation analysis, spillway, structural and slope stability analyses are more commonly conducted. In addition, current internal erosion/piping risk assessments are limited because they often tend to be qualitatively based. This is because there is very little understanding of the mechanics of internal erosion with regard to seepage. The contribution of this work is a new approach/tool based on quantitative parameters and soil mechanics.