Historically, breaching of dikes, following erosion of the inner slope after heavy wave overtopping, has been a major coastal flood hazard. Over recent years, improvement in understanding of storm surge levels, and wave run-up, has meant that current design practice for the inner slope still relies on criteria, set largely from experience and judgement, for allowable average overtopping discharge.
With sea level rise and tighter planning constraints, there is an increasing need for engineers to be able to calculate the stability, and hence reliability, of grass cover layers on the inner slope of dikes subjected to wave overtopping.
Failure of the inner slope is seen to start with a large crack opening near the dike crest, closely followed by the turf sliding – exposing bare soil. This study examines possible mechanisms that may lead to "turf set-off", and explores relevant parameter uncertainty.
Recent hydraulic research that describes the flow parameters for individual wave overtopping events is applied to a typical range of storm waves at dikes, and it is shown that overtopping flow velocities greater than 4 ms-1 are readily generated, creating a shear stress on the grass of up to 0.4 kPa.
Source test data for the CIRIA Report 116 steady flow erosion model has been reviewed. This study illustrates that this model is not currently valid for use with short duration, very high velocity flow on steep slopes – as found for wave overtopping on dikes. In this range the model is likely to yield an unsafe answer.
A reliability function for turf stability, based on a superficial sliding mechanism, is developed in this study. Sliding failure is considered at the base of the turf layer (approximately 35 cm), and assumes saturated soil response is dominated by its structure and high permeability. Composite action from root cohesion (approximately 1 kPa at the base of the turf) has been modelled, and relevant turf parameters defined. The flow shear load is significant (25%) at this very shallow depth, and is therefore included in the model. Stability is shown to be marginal below 30 cm.
The concept of progressive collapse, occurring with repetitive overtopping events, is introduced as a mechanism to gradually reduce root cohesion, and weaken the turf, so increasing the probability of sliding with discontinuous flow.