Physical properties of breaking waves in unidirectional sea states
File(s)
Author(s)
Craciunescu, Constantin Cosmin
Type
Thesis
Abstract
This thesis concerns the study of wave breaking in the context of unidirectional water
waves. Breaking waves are of key engineering interest as they are characterised by
significantly higher velocities than their non-breaking counterparts, resulting in large
environmental loads on offshore structures.
Historically, breaking thresholds have been quantified using geometric parameters.
These rely on estimates of wavenumbers obtained from measured time-histories
of surface elevation. Using focused wave groups, the work herein shows that the linear
dispersion equation can be 30% inaccurate, while Hilbert transform approaches
incorrectly estimate the wavelength by 20% for broad-banded sea-states. Hence, the
present thesis gives recommendations on obtaining best estimates of wavenumbers
and also provides empirical corrections to the dispersion equation.
This thesis is the first to illustrate the competing effects of nonlinear spectral
broadening, spectral phasing and dispersion reduction on the evolution of crests
speed. Specifically, rapid nonlinear spectral broadening acts to reduce the crest
speed, while nonlinear increases in components’ phase speeds act to increase it; an
overall nonlinear reduction in crest speed is observed. This improved physical understanding
helps explain the success of a newly developed dynamic breaking criterion.
An experimental investigation was performed to parameterise energy dissipation
in breaking waves based on JONSWAP spectra. Experimental results indicate that
an 85% increase in the eddy viscosity is needed to account for breaking dissipation
in realistic sea-states.
Finally, a High-Order Spectral model is improved to include a novel breaking
detection methodology and the new breaking dissipation parameterisation, creating a
model capable of efficiently simulating realistic sea-states, including breaking waves.
It is demonstrated that the model can accurately capture the effects of nonlinear
amplification and wave breaking on crest elevation statistics. The thesis is the first
to offer a comprehensive validation of the model against high-quality laboratory data
across a wide range of sea-states.
waves. Breaking waves are of key engineering interest as they are characterised by
significantly higher velocities than their non-breaking counterparts, resulting in large
environmental loads on offshore structures.
Historically, breaking thresholds have been quantified using geometric parameters.
These rely on estimates of wavenumbers obtained from measured time-histories
of surface elevation. Using focused wave groups, the work herein shows that the linear
dispersion equation can be 30% inaccurate, while Hilbert transform approaches
incorrectly estimate the wavelength by 20% for broad-banded sea-states. Hence, the
present thesis gives recommendations on obtaining best estimates of wavenumbers
and also provides empirical corrections to the dispersion equation.
This thesis is the first to illustrate the competing effects of nonlinear spectral
broadening, spectral phasing and dispersion reduction on the evolution of crests
speed. Specifically, rapid nonlinear spectral broadening acts to reduce the crest
speed, while nonlinear increases in components’ phase speeds act to increase it; an
overall nonlinear reduction in crest speed is observed. This improved physical understanding
helps explain the success of a newly developed dynamic breaking criterion.
An experimental investigation was performed to parameterise energy dissipation
in breaking waves based on JONSWAP spectra. Experimental results indicate that
an 85% increase in the eddy viscosity is needed to account for breaking dissipation
in realistic sea-states.
Finally, a High-Order Spectral model is improved to include a novel breaking
detection methodology and the new breaking dissipation parameterisation, creating a
model capable of efficiently simulating realistic sea-states, including breaking waves.
It is demonstrated that the model can accurately capture the effects of nonlinear
amplification and wave breaking on crest elevation statistics. The thesis is the first
to offer a comprehensive validation of the model against high-quality laboratory data
across a wide range of sea-states.
Version
Open Access
Date Issued
2022-05-26
Date Awarded
01/10/2023
License URL
Advisor
Christou, Marios
Sponsor
Imperial College London
Engineering and Physical Sciences Research Council
Grant Number
1855727
Publisher Department
Civil and Environmental Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)