416 



Rew = O.SSxIO 



:1.5 



FIGURE 23. Inconsistency of cavitation inception 

 outside the critical radius at low nuclei content. 

 Propeller C at 60% slip. 



like this is shown in Figure 22, but none or only 

 a few transient bubbles were seen. 



Electrolysis sometimes restores biibble cavitation 

 in this region, but in many cases it does not. 

 This inconsistency could even be foimd on the same 

 propeller in virtually the same condition when 

 tested repeatedly with long time intervals. In 

 one case an abundant amount of large bubbles was 

 visible without causing bubble cavitation, while 

 an amount of invisibly small nuclei did cause 

 bubble cavitation in the same condition. In Figure 

 19d it was seen that in the cavitation tunnel 

 cavitating spots at the leading edge were formed at 

 high Reynolds numbers . When the cavitation index 

 was lowered, bubble cavitation occurred in the 

 wake of these spots, while at radii in between of 

 the spots no bubble cavitation was observed. When 

 the cavitation index was lowered to about ai^rp=0.5 

 the spots were connected with intense biibble cavita- 

 tion, as shown in Figure 24. It can be seen that 

 the bubble cavitation is related to the spots at 

 the leading edge. Apparently the stream nuclei, 

 which were abundant in the tunnel at this low 

 cavitation index, did not create bubble cavitation, 

 while nuclei, generated by a cavitating spot 

 created intense bubble cavitation. The possible 

 relation between pressure distribution, boundary 

 layer, and nuclei distribution must be studied to 

 analyse these phenomena. 



VARIATION OF THE BOUNDARY LAYER BY ROUGHNESS 

 AT THE LEADING EDGE 



is larger in comparison with the boundary layer 

 thickness , there was a lower limit in the region 

 which had to be covered with carborundvim to cause 

 turbulent flow. For thin sections an evenly dis- 

 tributed layer of carborundum of say 0.5^ of the 

 chord was necessary to trip the boundary layer. 

 There was little difference between the effect of 

 30 urn and 60 pm carborundum. At thick sections 

 to be effective roughness was necessary until about 

 the minimum pressure point. At the pressure side 

 the boundary layer remained increasingly laminar 

 when the loading increased. At 1Q% slip the 

 the pressure side of the roughened blades was 

 completely laminar near the leading edge. 



Attention, given until now to the propeller 

 boundary layer, was focussed on the effect on 

 torque and thrust. Calculation methods to account 

 for Reynolds effects on open-water characteristics 

 are based on the assumption of turbulent boundary 

 layer flow on the propeller model [Lerbs (1951) ] 

 or on an empirical value in between fully turbulent 

 and fully laminar, as compiled by Lindgren (1972) . 

 From the paint tests however, we saw that the 

 turbulent region at the suction side strongly 

 depends on the propeller loading. The difference 

 between the dimensionless thrust and torque coeffi- 

 cients, therefore, will not only depend on the 

 Reynolds number, but also on the propeller loading. 



In order to eliminate the dependency of thrust 

 and torque coefficients on the Reynolds number, 

 turbulence stimulators have been used. Sasajima 

 (1975) used studs, Yasaki and Tsuda (1972) and 

 Tsuda et al. (1977) used trip wires at some distance 

 from the leading edge. Apart from changing the 

 boundary layer, these devices also have considerable 

 resistance of their own. Effects both on thrust 

 and torque are difficult to separate. The influence 

 of roughness at the leading edge on thrust and 



In all tests, at least one of the propeller blades 

 was roughened at the leading edge , as described in 

 Section 2. With paint tests, it was verified that 

 the laminar regions were changed into turbulent 

 ones. Although the grain size of 30 pm and 60 ym 



■'NT 



0.5 



FIGURE 24. Bubble cavitation in the wake of spots at 

 the leading edge. Propeller A at 30% slip in the cavi- 

 tation tunnel. 



