Uncertainty in the prediction of overtopping parameters in numerical and physical models due to offshore spectral boundary conditions

Williams, Hannah Elizabeth (2015) Uncertainty in the prediction of overtopping parameters in numerical and physical models due to offshore spectral boundary conditions. PhD thesis, University of Nottingham.

PDF (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (11MB) | Preview


The accurate prediction of wave overtopping is one of the most important aspects in the design of coastal defence structures. This can be achieved by using three different approaches: by physical modelling using laboratory tests, by empirical formulae available in literature derived from physical modelling and field tests, or by numerical simulation of the hydraulic response of the structure.

All of these prediction methods are subject to a certain level of uncertainty. One source of this is the requirement of a defined free surface elevation and velocity time series seaward boundary condition in any model. Often, these are not available but the modeller is instead provided with an incident energy density spectrum. A time series will then be reconstructed from this spectrum to be used as boundary conditions. Since the energy density spectrum provides only information on the amplitude of the components, it is usually assumed that the phases of these components are randomly distributed. To create the randomly generated phases, an initial seed value is required to generate a population of uniformly distributed random phases. By varying this value for each simulation a different time series will be produced. The overall objective of this research is to quantify the uncertainty in the prediction of overtopping due to this process.

This research involved carrying out two sets of laboratory experiments. Firstly, those carried out in the 2D wave flume at HR Wallingford, which provided a reference case for the validation of a numerical model, as well as a measured incident wave spectra for the generation of the population of reconstructed offshore boundary time series. The second set of experiments was carried out in the smaller 2D flume at the University of Nottingham to investigate the effect of random seeding to generate the time series at the wave paddle on the resulting overtopping parameters. This was also carried out to allow a comparison in the variability between the physical and numerical results.

It was found in the work, that when a measured free surface elevation is used as the input, good agreement between the numerical solver prediction and the overtopping measurements was observed. Subsequently, when a Monte Carlo approach was used to generate the population of reconstructed offshore boundary time series from the measured incident spectra the statistical analysis of the results showed that the variability was higher for the small numbers of overtopping waves and decreases as overtopping becomes more frequent.

To allow for more generalised conclusions on the uncertainty, further numerical tests were then carried out with synthetic spectra allowing different hydraulic and structural parameters to be considered. These showed good agreement with the findings of the initial statistical analysis.

Finally, the results from the physical model tests carried out at the University of Nottingham were analysed. The influence of laboratory effects were studied and analysis was carried out to establish the magnitude and sources of variability in these results. As with the numerical results, the characteristics of the distribution of the predicted overtopping parameters were also studied.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Briganti, Riccardo
Dodd, N.
Keywords: Waves, Mathematical models, Shore protection
Subjects: T Technology > TC Hydraulic engineering. Ocean engineering
Faculties/Schools: UK Campuses > Faculty of Engineering
Item ID: 30339
Depositing User: Williams, Hannah
Date Deposited: 21 Apr 2016 10:29
Last Modified: 17 Sep 2016 02:21
URI: http://eprints.nottingham.ac.uk/id/eprint/30339

Actions (Archive Staff Only)

Edit View Edit View