Hostname: page-component-7b7567c8c-wxvwp Total loading time: 0 Render date: 2026-11-16T21:42:17.320Z Has data issue: false hasContentIssue false

Resolving wave and laminar boundary layer scales for gap resonance problems

Published online by Cambridge University Press:  18 March 2019

H. Wang*
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
H. A. Wolgamot
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
S. Draper
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
W. Zhao
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
P. H. Taylor
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
L. Cheng
Affiliation:
Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
*
†Email address for correspondence: hongchao.wang2013@gmail.com

Abstract

Free surface oscillations in a narrow gap between elongated parallel bodies are studied numerically. As this represents both a highly resonant system and an arrangement of relevance to offshore operations, the nature of the damping is of primary interest, and has a critical role in determining the response. Previous experimental work has suggested that the damping could be attributed to laminar boundary layers; here our numerical wave tank successfully resolves both wave and boundary layer scales to provide strong numerical evidence in support of this conclusion. The simulations follow the experiments in using wave groups so that the computation is tractable, and both linear and second harmonic excitation of the gap are demonstrated.

Information

Type
JFM Papers
Copyright
© 2019 Cambridge University Press 
Figure 0

Figure 1.Definition sketch of experimental and numerical (mesh A3) set-ups.

Figure 1

Figure 2.Comparison of numerical and experimental incident wave groups for cases A and B.

Figure 2

Figure 3.Boundary layer profiles in the gap at various phases over a nominal cycle $T$ (from 8.12 s to 9.10 s) computed numerically (thin lines) and using (2.1) (circles) for meshes (a) A1 and (b) A3, at $y=0$, $z=-0.43D$.

Figure 3

Figure 4.Response at gap centre for meshes A1–A4 (a) and that from mesh A3 compared to Zhao’s experiments (b) for case A (linear excitation). The time axes are as in figure 2.

Figure 4

Figure 5.Snapshots of numerically computed response wave field for mesh A3 (in the region $2.5~\text{m}, $-4.0~\text{m}) captured at different times. Note that the vertical axis is stretched, such that 1 unit in the vertical equals 5 units in the horizontal, to more clearly show the free surface motions. To this end, the boxes are shown as transparent.

Figure 5

Figure 6.Comparison of experimentally and numerically (mesh B3) determined harmonic responses: (a) long-wave difference component; (b) first harmonic; (c–f) second harmonic; (g) third harmonic; and (h) fourth harmonic, with scaled envelope squared of the second harmonic (dashed grey lines). The time axes are as in figure 2.

Figure 6

Figure 7.Non-dimensional damping coefficient for each mode versus frequency (discontinuous axis) for each numerical simulation compared to experimental results. The modes shown are modes 1, 3 and 5 (from left to right). The grey dashed vertical lines and values above represent the resonant frequencies obtained from the potential flow software DIFFRACT.

Figure 7

Figure 8.Dissipation rate time series for control volumes $V_{1}$ and $V_{2}$, mesh A3, case A and (inset) cross-section diagram of control volume $V_{2}$, for which $\unicode[STIX]{x1D6FF}=3$  mm.

Figure 8

Figure 9.Radiation pattern, in terms of energy loss per unit modal amplitude squared per radian, from gap mode 1 (a), mode 3 (b) and mode 5 (c) determined using linear potential flow software DIFFRACT in an unbounded domain. The definition sketch of $\unicode[STIX]{x1D703}$ is shown in (a) and $\unicode[STIX]{x1D703}=\unicode[STIX]{x03C0}/2$ is along the gap direction.

Figure 9

Figure 10.Amplitude spectrum (a), and high-pass and low-pass time series (b) of viscous dissipation rate from $t=-3$ to $t=29$  s. In (a), double-frequency peaks corresponding to the three main gap modes can clearly be seen between 2 and 2.5 Hz, and the red dashed line represents the cut-off frequency for the filters.

Figure 10

Figure 11.Numerical fit of high-pass (a) and low-pass (b) viscous dissipation rate time series, using the method of Kumaresan & Tufts (1982) and a least-squares method, respectively.

Figure 11

Table 1.Values of $\unicode[STIX]{x1D709}\times 10^{3}$ determined from ‘global’ $\unicode[STIX]{x1D702}$ (corresponding to the damping coefficients in figure 7) and ‘local’ $\unicode[STIX]{x1D707}$ analyses.

Wang et al. supplementary movie

Movie 1 (for mesh A3) shows the numerically computed free surface elevation around the boxes during passage of the incident transient wave group for Case A (in the region 2.5 m < x < 8.5 m, -4.0 m < y < 0 m, which is not the full extent of the NWT). Note that the vertical axis is stretched, such that 1 unit in the vertical = 5 units in the horizontal, to more clearly show the free surface motions. To this end, the boxes are shown as transparent.

Download Wang et al. supplementary movie(Video)
Video 9.3 MB