THEME 4: Prediction of wave impact (overtopping and loads) on the dike during storms | CREST project

THEME 4: Prediction of wave impact (overtopping and loads) on the dike during storms


  1. Hidden in plain sight: imperceptible to the naked eye, very long waves are unexpected adversaries for our coastal defence protection against flooding during violent storms.
  2. Shallow beaches are essential in our hybrid soft-hard coastal defence system – Size matters in the protection against violent storms: wider beaches reduce the wave impact on the dike.
  3. Complexity begets complexity: the complex shape of our coastal defence system leads to complex hydrodynamics – Accurate prediction of the wave impact on the dike and buildings requires state of the art numerical modelling to avoid over-conservative design.
  4. Using the more detailed insight in physical processes and the validated numerical tools acquired in the CREST project, the complete calculation methodology for safety assessments and risk calculations can be improved.
  5. Large-scale experiments on overtopping wave loads (WALOWA project) suggest that the impact force acting on dike mounted vertical walls with shallow foreshores can be estimated using a hydrostatic pressure assumption.
  6. A new experimental dataset is available of 2D wave flume physical modelling of (individual) wave overtopping and impacts on dikes with very shallow foreshores (very relevant to the Belgian coast). The dataset also includes high spatial resolution measurements of surface elevations along the foreshore slope, allowing a more detailed study of long waves.
  7. A new experimental dataset is available of 3D wave basin physical modelling of (individual) wave overtopping and impacts on dikes with very shallow foreshores (very relevant to the Belgian coast). This dataset includes long-crested, obliquely incident long-crested and short-crested wave tests, allowing the study of 3D effects at the dike and directional spreading of the waves.
  8. The spectral wave period at the toe of the dike, Tm-1,0,t is used in all overtopping and wave impact prediction formulas. An existing semi-empirical formula for Tm-1,0,t was validated using data from mildly sloping shallow foreshores, but returns an overestimated value for the case of steep shallow foreshore slopes. A modification of the formula has been carried out, making it applicable and more accurate for use in cases with steeper shallow foreshore slopes.
  9. A significant effect of the foreshore slope angle and the dike geometry (promenade length, inclusion of storm wall,…) on the wave overtopping and wave impact force is discovered. A modification of the existing prediction formulas is ongoing.
  10. Long waves (or infragravity waves) significantly affect the wave-induced structural response (overtopping, wave impact) of dikes for the case of very shallow foreshores. However, very little is actually known about these long waves in the nearshore region during storm conditions, especially along the Belgian coast. Dedicated field measurements are strongly recommended.
  11. Long waves feature strong reflection from a dike with shallow foreshore, while they might break on mildly sloping beaches in the surf zone and reflect much less from the shoreline in case no dike is present. The presence of the dike therefore affects long wave reflection on mildly sloping beaches. Further research into the role of the dike in this process, might lead to further insight into changes in the hydrodynamics and their influence on the surf zone morphodynamics during storm conditions.
  12. Active wave absorption in physical models should be tuned to include both reflected long waves and seiches (if the wave paddle stroke length allows it) when testing coastal structures with a very shallow foreshore. Otherwise, build-up of long wave energy will significantly affect the measurements of wave-induced structural response.
  13. Measured experimental wave impact forces have a low repeatability, because of a high dependence on small changes in environmental conditions. On the other hand, repeatability is important to reduce uncertainty in prediction formulas derived from experiments and for validation of deterministic numerical models. Low-pass filtering of the measured signal of the impact forces in the post-processing step, effectively removing mostly the stochastic part of the dynamic impact types, improves repeatability.
  14. Smaller elements of buildings, such as windows and doors, usually have a higher natural frequency than the recommended low-pass filter for experimentally measured impact forces and are affected by the stochastic part of the dynamic impact types. Therefore, a dynamic impact force safety factor should be applied to a calculated maximum force (determined from low-pass filtered force measurements) for the design of such elements.
  15. Directional spreading, expressing the degree of short crestedness of real sea waves, is an essential parameter in the design of beach nourishments and structures for coastal safety. The higher its value, the lower the long wave height is at the dike toe, leading to lower overtopping and impact force. Modification of existing prediction formulas is ongoing. However, there is little known about the amount of directional spreading actually occurring nearshore during storm conditions along the Belgian coast. More analysis of existing field measurements is strongly recommended, in addition to continued and more dedicated field measurements.
  16. First order wave generation at the offshore boundary in nearshore experimental and numerical models introduces spurious, non-physical long waves, which affect the maximum individual overtopping volume and the mean wave overtopping discharge. This is especially true for mean overtopping discharge values in the order of 10 l/m/s and lower. Second order wave generation prevents such spurious long waves and is therefore recommended.
  17. The numerical model SWASH is able to provide an accurate estimation of the maximum force per impact event on dike-mounted vertical walls, by assuming hydrostatic pressure only for the calculation of the force on the vertical wall. Including non-hydrostatic pressure effects might improve results further, particularly for dynamic wave impacts. However, spurious pressure/force oscillations are observed when including the non-hydrostatic pressure. No explanation for this numerical effect has been found yet.
  18. The numerical model SWASH significantly underestimates the impulse of the force per wave impact event on a dike-mounted vertical wall in shallow foreshore conditions, indicating that the wave impact flow is not modelled correctly. More detailed Navier-Stokes models such as OpenFOAM and DualSPHysics are necessary for a more accurate flow modelling along the vertical wall, leading to a better estimation of the duration of wave impact forces.
  19. Maximum individual wave overtopping and impact is affected by the wave generation method (seed effect). This effect was tested for a mean overtopping discharge, q, of about 15 l/m/s, and is expected to increase even more for smaller mean overtopping discharges (e.g. q ≈ 1 l/m/s, currently the limit used in the safety assessment). Additional in-depth research into this issue is necessary.
  20. Modelling beach morphodynamics during a storm is a key aspect in understanding and accurately predicting wave overtopping. The sand transport along the beach profile during a storm (beach morphodynamics) triggers profile changes which need to be included in the modelling of wave overtopping over a dike with a very shallow foreshore (very relevant to the Belgian coast).