Projects

Practical challenge and motivation

The goal is to predict for which dykes macro-instability can be allowed and for which dykes flooding will definitely occur afterwards and of course this prediction should be made before the initial failure actually occurs. This allows dyke assessment to be more efficient and will help designers to take residual dyke resistance into account. The current guidelines for residual dyke resistance after macro-instability are largely based on engineering expertise. However, many questions remain on the applicability of this expertise in more complex situations. A modelling tool predicting the failure process after macro-instability will be used to expand upon the engineering expertise, thereby evaluating and expanding the existing guidelines.

Research challenge

How to determine the effect of residual dyke resistance onto the probability of flooding?

Innovative components

The main component of this research is developing and using the (Random) Material Point Method or (R)MPM to determine residual dyke resistance (the difference between macro-instability and flooding (see pictures). MPM is a new tool (similar to the widely used Finite Element Method) which models besides the start of initial failure also the dyke deformations after initial failure, as shown in the video. This can be used to predict if a dyke floods after initial failure.

The failure process is highly variable and the model is therefore expanded to a fully probabilistic tool (Random MPM), in order to stick to the probabilistic framework designed for dykes. The framework is extended for macro-instability to properly include residual dyke resistance and finally the existing guidelines are assessed and extended if necessary in order to provide outcomes for the user practice.

Relevant for whom and where?

Anyone designing or assessing levee and considering to take residual levee strength into account, which has the biggest impact on levees with a large width or overheight.

Progress and practical application

Residual dyke resistance is definitely present in almost any dyke with a reasonable width. The probability of flooding can reduce significantly (even up to almost 100%) compared to the probability of initial failure, even for dykes constructed from fairly weak materials. However, this residual dyke strength is highly dependent on the geometry, material properties, soil variability and external water level. While the guidelines often assume a ‘safe’ remaining geometry, such a geometry has not been found in the examples tested, since for small initial failures (failures where only 1/5 of the crest was damaged), only a 50% reduction of the probability of flooding was observed. On the other hand a lower external water level compared to the dyke height does increase residual dyke resistance considerably.

Status for day-to-day practice

Currently, this research mainly shows that using residual dyke resistance is promising, but it is uncertain in which cases it can be properly used. Further investigation is necessary to determine these cases. From the current research it can already be determined that, dykes with varying water levels can receive a considerable reduction in the overall probability of flooding when residual dyke resistance is used for the lower water levels.