497 



IdB 



— o— - cavity on the V 

 pressure -side of 

 the leading edge 



— • — - cavity on the 



suction-Bide of 

 the leading edge 



FIGURE 5. Development of noise and visual manifesta- 

 tion of cavitation on both pressure- and suction-side 

 in a uniform flow at constant pressure. 



unsteadiness of even small size cavities due to 

 closure behind the maximum suction zone. 



The change in the relative noise intensity of 

 sheet and bubble cavities depends upon the fact 

 that in the case of the bubble cavity structure the 

 unsteadiness varies but slightly, whereas the volume 

 of sheet cavities begins to sever ly pulsate. Passing 

 over to the unsteady flow, we may even observe the 

 reduction of bubble cavitation noise. This occurs 

 when one portion of the propeller gets free from 

 the cavity whereas, on the other portion thereof, 

 the intensive development of cavitation is not 

 accompanied by an increase of noise due to a satura- 

 tion effect. 



Individual points on the graphs shown in Figures 

 2 to 5 indicate moments of the first visual detection 

 of cavitation. As is seen, in a large cavitation 

 tunnel where the measurements were made, the above 

 conclusion that the noise comes ahead of the visual 

 detection of cavitation is to a variable degree 

 valid for any type of cavitation. 



a decrease of pressure, p^ , from 1 to 0.4 ata. It 

 is also emphasized that at sufficiently low p the 

 collapse of cavities is not necessarily accompanied 

 by shock wave generation. 



Vacuum noise measurements, when performed in 

 ship hydrodynamics laboratories engaged in cavitation 

 research, show inadmissible noise absorption in 

 the facility water unless measures are taken to 

 insure additional removal of gaseous nuclei of cavi- 

 tation from the water. By intensified vjater degassing 

 the absorption may be reduced to an acceptable level , 

 but the resulting growth of cavitation resistance 

 of water leads to a drastic change of conditions 

 for inception and development of cavitation [Gorsh- 

 koff and Lodkin (1966)]. In view of the complicated 

 character of absolute pressure effects on the 

 coefficient of cavity energy transformation into 

 noice it appears to be good practice to perform 

 cavitation noise measurements at a full-scale value 

 of pressure. 



That the Froude similarity will not be fulfilled 

 under these conditions, can be accepted provided 

 that adequate means are available for the description 

 and reproduction of the conditions of flow non- 

 uniformity behind the hull. This approach, used 

 in a large cavitation tunnel in combination with 

 correlation methods recommended by Levkovsky (1958), 

 Sturman (1974) , has shown that overestimated cavita- 

 tion noise levels are predicted in this case. This 

 was found to arise from the fact that the coefficient 

 of cavity energy transformation into noise is 

 approximately proportional to the rate of pressure 

 growth leading to the cavity collapse. In modelling 

 by the Froude method this pressure growth rate 

 decreases as v'lT 



In case of large-scale modelling the comparison 

 of model-test and full-scale data may not have 

 revealed this discrepancy among other more pro- 

 nounced ones. One can use pressure gradient instead 

 of the rate of pressure variation with time. Then, 

 for Froude similarity, the noise level model-to-full- 

 scale extrapolator coincides with that used by Sturman 

 (1974) . Not so with modelling at full-scale absolute 

 pressure. Here the proportionality of the transform- 

 ation coefficient both to the velocity of pressure 

 variation with time and to the pressure gradient in- 

 volve the same extrapolator. Giving up the construc- 

 tion of dimensionless parameters of which, with a 

 great number of constants involved, there is ample 

 freedom of choice, the extrapolator suggested by 

 Sturman (1974) 



<P^ 



R^p N 







(7) 



3 . MODEL-PROTOTYPE CORRELATION AND COMPARISON OF 

 MODEL-TEST RESULTS WITH FULL-SCALE DATA 



It is usually assumed [Levkovsky (1968) and 

 Sturman (1974)] that the fraction of the cavity 

 potential energy converted into cavitation noise 

 (coefficient of transformation) is the same for 

 the model and the full scale ship. Experience con- 

 firms the validity of the conflicting conclusions 

 [Beniaminovich et al. (1975)] that are indirectly 

 confirmed in some works. The coefficient of cavity 

 energy transformation into noise proves to be 

 strongly dependent on the absolute pressure, p . 

 It is this fact, that was used by Beniaminovich 

 et al. (1975) for explaining the reduction of the 

 transformation coefficient by several orders with 



can be substituted by the following: 



<P" 



o o o o 



T^L 



(8) 



Here 



No 



is the square of the acoustic pressure, 



is the distance to the point of noise 



measurement, 



is the number of cavities collapsing in 



unit time. 

 If we assume in the regular way that the similar- 

 ity of cavity patterns is observed and the noise is 

 measured at similar points of the flow, then 



R 



= L 



Mr 



