Internal Erosion: Laboratory Testing to Identify End States in Internally Unstable Soils
Internal erosion is one of the primary failure modes of concern for embankment dams and canals. Internal instability is a type of internal erosion that refers to any process by which the finer soil particles can be selectively eroded out of the coarser particles. Internal instability can occur in natural soils and is common in soils that are either gap graded or broadly graded due to insufficient fractions of certain particle sizes being present to act as a filter for the finer particles. If erosion of finer particles initiates and is not detected soon enough, it can result in substantial damage to embankments and potential for embankment failure.
In the United States, assessing the probability of failure due to internal erosion has relied exclusively on the application of event trees to combine results of analyses and engineering judgement into a quantitative probability of failure. For failure to occur, four separate events must happen: 1) erosion must initiate, 2) the erosion must be allowed to progress continuously to the reservoir, 3) erosion must be able to continue to develop and enlarge without hindrance after hydraulically connected to the reservoir, and 4) the embankment ultimately needs to collapse resulting in uncontrolled release of the reservoir. With regards to internal instability, all research to date has focused on developing analysis techniques for determining how and when erosion initiates - i.e., informing the first node in the event tree. All analysis methods to date can be summarized as "gradation-shape analysis" techniques (essentially looking at distribution of particle sizes), and while they inform whether unstable soil is likely present, they do not inform nodes two through four; therefore, the current analysis approach relies on significant engineering judgement.
Need and Benefit
Internal erosion is one of the most common reasons embankment dams fail. We need research to develop analysis methods for assessing all nodes in the internal erosion event tree leading to failure. Until the state of the art is advanced in these areas condition assessment and design will continue to depend almost exclusively on engineering judgment.
Contact the Principal Investigator for information about partners.
Please contact firstname.lastname@example.org about research products related to this project.