Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-17T20:44:11.364Z Has data issue: false hasContentIssue false

An energetic signature for breaking inception in surface gravity waves

Published online by Cambridge University Press:  23 March 2023

Daniel G. Boettger*
Affiliation:
School of Mathematics and Statistics, University of New South Wales, Sydney 2052, Australia
Shane R. Keating
Affiliation:
School of Mathematics and Statistics, University of New South Wales, Sydney 2052, Australia
Michael L. Banner
Affiliation:
School of Mathematics and Statistics, University of New South Wales, Sydney 2052, Australia
Russel P. Morison
Affiliation:
School of Mathematics and Statistics, University of New South Wales, Sydney 2052, Australia
Xavier Barthélémy
Affiliation:
School of Mathematics and Statistics, University of New South Wales, Sydney 2052, Australia
*
†Email address for correspondence: d.boettger@student.unsw.edu.au

Abstract

A dynamical understanding of the physical process of surface gravity wave breaking remains an unresolved problem in fluid dynamics. Conceptually, breaking can be described by inception and onset, where breaking inception is the initiation of unknown irreversible processes within a wave crest that precede the visible manifestation of breaking onset. In the search for an energetic indicator of breaking inception, we use an ensemble of non-breaking and breaking crests evolving within unsteady wave packets simulated in a numerical wave tank to investigate the evolution of each term in the kinetic energy balance equation. We observe that breaking onset is preceded by around one quarter of a wave period by a rapid increase in the rate of convergence of kinetic energy that triggers an irreversible acceleration of the kinetic energy growth rate. This energetic signature, which is present only for crests that subsequently break, arises when the kinetic energy growth rate exceeds a critical threshold. At this point the additional kinetic energy convergence cannot be offset by converting excess kinetic energy to potential energy or by dissipation through friction. Our results suggest that the ratio of the leading terms of the kinetic energy balance equation at the time of this energetic signature is proportional to the strength of the breaking crest. Hence this energetic inception point both predicts the occurrence of breaking onset and indicates the strength of the breaking event.

Information

Type
JFM Papers
Copyright
© The Author(s), 2023. Published by Cambridge University Press
Figure 0

Figure 1.Schematic of our numerical wave tank. Waves are generated at the paddle boundary and travel down the tank in the positive $x$ direction before being absorbed by the numerical sponge layer. A typical chirped wave (enlarged for clarity) is shown, with tank dimensions normalised by the deep-water wavelength $\lambda _p$ derived from the paddle frequency $\omega _p$.

Figure 1

Table 1.Summary of experiments included in this study. The model was configured using a range of mesh refinement levels, wave packet size $N$ and water depth $d/\lambda _p$. For each configuration the amplitude of the paddle $A_p/\lambda _p$ was varied to generate an ensemble of breaking and non-breaking crest cases.

Figure 2

Figure 2.Summary of the individual crests included in the ensemble. The model was configured using a range of mesh refinement levels ($2^x$), wave packets $N$ and water depth $d/\lambda _p$. For each configuration the amplitude of the paddle $A_p/\lambda _p$ was varied to generate an ensemble of crests with different energy levels. Non-breaking (breaking) crests are indicated by hollow (filled) symbols. The representative crests shown in subsequent figures are labelled NB1, B1 and B2. (a) The maximum kinetic energy $E_k$ as a function of local wave steepness $S_c$. (b) The value of $B$ at $t^*=0$ (non-breaking crests with $B < 0.8$ not shown); the threshold $B_{th} = 0.855 \pm 0.05$ reported by Barthelemy et al. (2018) is shown by the hatched region.

Figure 3

Figure 3.The evolution of $E_k$ for the (a) NB1 (near-breaking), (b) B1 and (c) B2 (breaking) representative crests as they progress through the growing and decaying phase. Snapshots are equally spaced at intervals of $0.09T_0$ and the vertical axis is exaggerated by a factor of 7 : 1. The $+$ indicates the location $\boldsymbol {x_+}$ where $E_k$ has its maximum value, and in (d) the temporal evolution of $E_k$ at this location is shown. Periods of active breaking are indicated in (d) by the dotted lines. The time of each snapshot (A–D) corresponds with the vertical dashed lines. Snapshot $C$ occurs at the time that $E_k$ at $\boldsymbol {x_+}$ has its maximum value for the near-breaking crest, and the time of breaking onset for breaking crests.

Figure 4

Figure 4.The rate of change of the local kinetic energy density $E_k$ at the location $\boldsymbol {x_+}$ (figure 3) and its contributing terms from (3.3) for the representative near-breaking (NB1) and breaking (B1, B2) crests. Positive (negative) values indicate a source (sink) of $E_k$. Periods of wave breaking are indicated by the solid dots.

Figure 5

Figure 5.The evolution of the kinetic energy balance for the NB1 crest. The dominant terms of the local balance equation (3.3) are (a) $CON$, (b) $K2P$ and (c) ${\rm D}_b E_k / {\rm D}t$. Vectors in (a) show the relative magnitude of the $\boldsymbol {u - b}$ flux velocity at each snapshot. The $+$ indicates the location $\boldsymbol {x_+}$ where $E_k$ has its maximum instantaneous value. Snapshots are equally spaced at intervals of $0.09T_0$ and the vertical axis is exaggerated by a factor of 7 : 1, with snapshot $C$ occurring at the time that $E_k$ at $\boldsymbol {x_+}$ reaches its maximum value. The temporal evolution of each term in (a), (b) and (c) at $\boldsymbol {x_+}$ have been taken from figure 4 and shown in (d) for comparison.

Figure 6

Figure 6.As for figure 5 showing the B1 crest. Periods of wave breaking are indicated by the solid dots.

Figure 7

Figure 7.As for figure 5 showing the B2 crest. Periods of wave breaking are indicated by the solid dots.

Figure 8

Figure 8.Relationship between the two leading terms in the $E_k$ balance equation (3.3): $K2P$ $- \boldsymbol {u} \boldsymbol{\cdot} \boldsymbol {f}$ and $CON$, for the period $t^*=-1$ to $t^*=0$. Values above (below) the dashed line indicate an imbalance between these terms which leads to an increase (decrease) in ${\rm D}_b E_k / {\rm D}t$. The energetic inception time ($\star$) and the kinematic inception time (i.e. when $B=B_{th}$, indicated by a $\times$) for breaking crests are annotated, with the superposed black lines indicating the period for which $0.83 < B < 0.88$. An animated version of this figure is provided as supplementary material.

Figure 9

Figure 9.The magnitude of $CON$ as a function of (a) $K2P - \boldsymbol {u}\boldsymbol{\cdot}\boldsymbol {f}$ and (b) ${\rm D}_b E_k / {\rm D}t$, at the time that the critical point $p^{\star }$ occurs. The $5\,\%$ and $95\,\%$ confidence intervals are also shown. Breaking crests are coloured by the breaking strength indicator $\varGamma$, and non-breaking crests are coloured grey. The ${\rm D}_b E_k / {\rm D}t$ threshold separating non-breaking and breaking crests is shown by the hatched region.

Figure 10

Figure 10.The energetic inception time ($\star$) and the kinematic inception time ($\times$) for all breaking crests in our ensemble, relative to the time of breaking onset $t^*=0$. Crests are grouped by wave packet size and water depth and the representative breaking crests B1 and B2 discussed in the text are annotated.

Figure 11

Figure 11.The integrated energy of the water phase (excluding the sponge layer) for the simulation from which the near-breaking B2 crest has been taken (black) and for identical simulations with lower maximum grid refinement levels $2^n$. The total $\mathcal {E}$ (top), kinetic $\mathcal {K}$ (middle) and potential energy $\mathcal {P}$ (bottom) are shown. All values are normalised by $\mathcal {E}_0$: the total energy at the time which the wave packet has fully entered the numerical wave tank. Breaking onset time for the $2^{11}$ B2 crest is indicated by the dashed grey line.

Figure 12

Figure 12.Schematic of the control volume $V(t)$ bounded at the top by the interface $\eta (x,t)$, at the bottom by the horizontal plane $z_0(t)$ and on the left and right by the vertical planes $x_L(t)$ and $x_R(t)$. The free variables $z_0$, $x_L$ and $x_R$ are set to examine a chosen subregion of the crest.

Figure 13

Figure 13.Evolution of the rate of change of the integrated kinetic energy $\mathcal {K}$ and its contributing terms from (B5) for the representative near-breaking (NB1) and breaking (B1, B2) crests. Positive (negative) values indicate a source (sink) of $\mathcal {K}$. Periods of wave breaking are indicated by the solid dots. The top row (a–c) shows the local values taken at the location $\boldsymbol {x_+}$ (figure 4) and multiplied by the model cell volume. Each remaining row displays values integrated over a control volume that encompasses increasing amounts of the crest as shown by each crest schematic (left).

Boettger et al. Supplementary Movie

See "Boettger et al. Supplementary Movie Caption"

Download Boettger et al. Supplementary Movie(Video)
Video 973.1 KB
Supplementary material: File

Boettger et al. Supplementary Movie Caption

Boettger et al. Supplementary Movie Caption

Download Boettger et al. Supplementary Movie Caption(File)
File 922 Bytes