• flow instabilities
• cavitatin inducers

This thesis is focused on the identication of instabilities manifestation in the autocorrelation created from pressure signals data of sensors installed in a turbopump casing. In the autocorrelation or spectra diagrams, it is not always feasible to identify and quantify overlapping uid dynamic instabilities, because of the presence of many peaks which are
not surely attributable to each one of them.
Very often, an autocorrelations rich in peaks can be read incorrectly, because some peaks of different frequencies can be
believed to belong to the same instability, instead of being caused by different phenomena, with relative different amplitude and frequency.
This thesis is developed implementing this process:
ˆ Correlation implementation;
ˆ Optimisation implementation;
ˆ Optimisation error calculation;
This procedure was utilised in this work especially with the three blades inducer RAPDUD pressure autocorrelation.
The autocorrelations were chosen with operating conditions such that rotating cavitation and surge cavitation are clearly visible, in order to improve and test the theoretical model with optimisations with a real experimental autocorrelation.