473 



pressure 



I 



JlO^ 



Pa 





«•■ '"'(;/'' VvH; 



Blade frequency period 



■^''X,,w> 





%v;i^^~" 



time (ms) 



10 



FIGURE 5. Pressure signal from a cavitating 

 propeller model. 



propeller. The signal corresponds to a spectrum of 

 the the type shown in Figure 4 and typically is a 

 rather slow variation of pressure interrupted by 

 sharp and fairly infrequent pulses. The pulses 

 are presumed to be generated during the final cavity 

 collapse and they provide the main contribution to 

 pressure levels at high frequencies. The pulses 

 are often higher than the low frequency variations , 

 but because of their low repetition frequency and 

 wide frequency content the spectrum levels at high 

 frequencies are lower than at low frequencies . 



To understand the scaling of cavitation noise 

 and how different types of cavitation noise are 

 generated, and perhaps can be reduced, it is 

 important to study the mechanism generating different 

 types of noise. A suitable way to obtain sucn 

 knowledge is to carry out high speed filming and 

 synchronous measurement of the cavitation noise. 

 The first idea was to carry out such measurements 

 with a propeller model. Because of high tip speed, 

 small dimensions, and the complicated geometry of 

 a propeller it was decided to take the first step 

 by performing such experiments with oscillating 

 hydrofoils. By suitable oscillation of a hydrofoil 

 it is possible to generate cavitation with approxi- 

 mately the same dynamic behavior as obtained from 

 a propeller operating in a wake. The experiments 

 with oscillating hydrofoils were supposed to shed 

 some light on the following questions that originated 

 from the search for methods of prediction and 

 reduction of propeller cavitation noise : 



1. Which are the characteristic properties of the 

 pressure pulses from some special types of 

 cavitation? 



2. Are strong pulses generated by an orderly 

 collapse of the whole cavity (e.g., a sheet 

 cavity) or do they originate from large or 

 small parts that separate from the main cavity? 

 What is the geometry before and during collapse 

 of cavities generating strong pulses? 



3. How is the pressure pulse related to the size 

 of the cavity? Is there, for example, any 

 relation between the maximum extension of a 

 sheet cavity and the final pressure pulse? 



4. Is rebound of cavities important for generation 

 of sharp pulses? 



5. What part of the cavitation period is of main 

 importance for the generation of different 

 types of noise (slow pressure variations, sharp 

 pulses, etc. ) ? 



6. Which are the characteristic properties of the 



flow field, osciJ-lation frequency, etc., causing 

 cavitation with violent collapse? 



7. To what extent is collapse time determined by 

 the oscillation frequency of the hydrofoil? 



8. To what extent does the cavity behavior seem 

 predictable by theoretical methods? How 

 realistic is it to think that a sufficiently 

 good scaling from model to full scale is 

 obtained for the most important cavitation 

 events? 



Experimental Set Up 



Cavitation Tunnel 



The tests were carried out in SSPA cavitation 

 tunnel No. 1 (the samller one) equipped with test 

 section No. 1 (500 x 500 mm). 



Oscillation Apparatus 



The hydrofoil was located horizontally in the test 

 section and attached to an oscillation apparatus 

 fixed to the test section wall (Figure 6) . The 

 hydrofoil was supported only at one end and forced 

 to oscillate (rotate) around an axis fixed spanwise 

 through the midchord point, i.e., the geometric 

 angle of attack oscillated around an adjustable 

 mean value, uq, (Figure 7). The axis was driven 

 by a connecting rod and an adjustable crankpin. 

 By setting the crankpin the oscillation angle, a, 

 could be varied from to 6°. With the hydrofoil 

 used in these tests the oscillation frequency, fosc' 

 was varied from to 15 Hz. The limits of water 

 speed, ag, a, and f^^^ were set by the strength of 

 the hydrofoil and the background noise generated by 

 the apparatus. One part of the background noise 

 from such an apparatus is knocking in shaft bearings. 

 To minimize this knocking, adjustable bearings were 

 used. The motor, which was not dimensioned for this 

 experiment, could deliver 16 kw at a maximum speed 

 of 50 r/s. 



The dynamic angle of attack, experienced by the 

 leading edge of the hydrofoil, is composed of the 

 geometric angle and of an angle caused by the motion 

 of the leading edge. The angle is also affected by 

 induced velocity. In the following only the geomet- 

 ric angle is considered (Figure 7) . 



The system with connecting rod and crankpin 

 results in an approximately sinusoidal oscillation 

 of the geometric angle of attack. This manner of 



