Harbour
Resonance: Comparison
with the Analytical solution given by Ijima

Contours of normalized wave height in the rectangular harbour.
The displayed contours are the responses near peak harbour resonance.
Ijima
et al. (1981) gave the analytical solution of this rectangular harbor with
a harbor length = 2h, harbor width = h, a constant water depth outside the harbor
(h = 0.1 m), and a constant slope within the harbor. On the left side of
this computing domain, the given boundary condition was specified. Along
the harbor perimeter, the total reflection B.C. (
= 0) was specified. At
the top and bottom open boundaries, the total passing through B.C. was assigned
(
= 1). The incident wave height is given as Hi.
When the ratio of harbor length to wave length approaches 1/4 (kh
1.2),
the ratio of He/Hi
reaches 8 and demonstrates the first resonance; the second resonance occurred
at kh
3.7. Because an approximation (i.e., the rectangular open boundary)
of the exact geometry for the open boundary was used, the model-calculated resonance
responses are not as accurate as the analytical solutions. As pointed
out by Chen (personnel communication 1997), a semi-circle open boundary is needed
for an accurate simulation of the resonance. The calculated contours of
the normalized wave height at a near resonance frequency further demonstrate
that the scatter waves propagated outward from the harbor entrance and showed
why a semi-circular open boundary is needed. For details, see Maa
and Hwung (1998).

Comparison of the model calculated wave height at the harbour end with the analytical solution.