5.1 Cavitation induced
signatures can take the form of either discrete harmonics (blade orders)
or have a broadband character. The signals are also non-steady and,
therefore, the measurement needs to take account of this. The non-steady
character may be a function of any propeller blade to blade geometry
differences, temporal and spatial wake fluctuations, and water quality
and environmental changes.
5.2 Time series data
are most commonly reduced to their constituent frequency content using
Fast Fourier Transform (FFT) analysis techniques. Such processes generate
frequency spectra of pressure amplitudes and phases averaged over
the length of the time series records which contain several hundred
blade passages. Such a procedure does not distinguish the contributions
made by individual blades or account for discrete bursts of activity.
5.3 To establish the
nature of the pressure signal time histories, in particular beating
type effects, the measurements should be viewed as time series to
present the relationship between pressure and propeller angular position.
The pressure signatures should be considered over both short and long
time frames.
5.4 To distinguish
short time frame events alternative methods of analysis are necessary.
Such methods would include Joint Time Frequency Analysis (JTFA) in
which segments of the signal are analysed individually and the frequency
content is obtained as a function of time. Such procedures can be
extended to the separate analysis of the signatures from each blade.
5.5 When available,
synchronised high-speed video and pressure signals should be presented
to assist in connecting cavity and pressure events to identify the
source of the higher harmonic content of the signals.
5.6 Pressure spectra
and 1/3rd octave band analysis should be presented in order to supplement
the amplitudes extracted at multiples of the blade passing frequency.
5.7 Hull pressures
should be given as single amplitude peak values (half the peak-to-peak).