4.1 Wave spectra
The JONSWAP spectrum should be used as this describes fetch
and duration in limited seas, which correspond to the majority of
the conditions worldwide. In this respect it is important that not
only the peak period of the wave train is verified but also that the
zero crossing period is correct.
It is required that for every test run the wave spectrum
be recorded and documented. Measurements for this recording should
be taken at the probe closest to the wave making machine.
It is also required that the model be instrumented so that
its motions (roll, heave and pitch) as well as its attitude (heel,
sinkage and trim) are monitored and recorded.
It has been found that it is not practical to set absolute
limits for significant wave heights, peak periods and zero crossing
periods of the model wave spectra, therefore an acceptable margin
has been introduced.
4.2 To avoid interference of the mooring system
with the ship dynamics, the towing carriage (to which the mooring
system is attached) should follow the model at its actual drifting
speed. In a sea state with irregular waves the drift speed will not
be constant; a constant carriage speed would result in low frequency,
large amplitude drift oscillations, which may affect the model behaviour.
4.3 A sufficient number of tests in different
wave trains are necessary to ensure statistical reliability, i.e.
the objective is to determine with a high degree of confidence that
an unsafe ship will capsize in the selected conditions. A minimum
number of 10 runs are considered to provide a reasonable level of
reliability.