Evaluating residual dyke resistance using the Random Material Point Method

Due to a lack of large deformation dyke assessment models, primary failure mechanisms, such as inner slope failure, are often used as a proxy to assess the probability of failure of a dyke. However, a dyke continues to fulfil its main function unless, or until, flooding occurs. The Random Material Point Method (RMPM) is used here to investigate residual dyke resistance, which is the resistance against flooding after initial failure. RMPM combines random fields with MPM in a Monte Carlo simulation and has been extended here to include the effects of an external hydrostatic pressure on a dyke’s outer slope. The residual resistance of an idealised dyke (computed using RMPM) is shown to reduce the probability of flooding by 25% with respect to the initial failure. A lower degree of anisotropy of the spatial variability increases the residual dyke resistance. RMPM simulates, as expected, a lower residual dyke resistance for larger initial failures and/or a higher water level. A ‘safe’ remaining geometry has not been found, since even small initial failures can result in an unacceptable probability of flooding, highlighting the importance of modelling the entire failure process.