Propeller-Hull Interaction 



met with in work done some years ago at NSRDC on the model of the tanker Man- 

 hattan when trying to compare a nine-bladed propeller with a f ive-bladed pro- 

 peller of roughly the same blade-area ratio. The Reynolds number of the nine- 

 bladed propeller was about 1.5 x 10 ^ and that of the five-bladed propeller was 

 about 5.0 X 10^. It is clear from Fig. 6 that this would give rise to a consider- 

 able difference in the performance, whereas at full scale both propellers op- 

 erate in the rather flat area and should not show any significant difference. 

 Actually the comparison between the five-bladed and the nine-bladed models 

 would indicate that the nine-bladed was not as efficient relatively as it actually 

 would be full scale. Dr. Morgan came upon this problem when he was well along 

 in developing a design technique for propellers. He discovered that on a couple 

 of propellers, especially the nine-bladed, the performance was not as expected, 

 and the theory did not match the experiment results. In the theoretical calcula- 

 tions it was assumed that the drag coefficient was 0.008, but reference to airfoil 

 data showed that the drag coefficient in fully turbulent flow would have been 

 0.012, or 50% higher. By assuming this value, the theoretical and experimental 

 results matched on the model, but the difficulty is to know whether the flow is 

 fully turbulent, and this is a very important point in doing analysis. 



Lackenby made a brief reply to Dr. Kinoshita, Professor Prohaska, and Dr. 

 English. In regard to Dr. Kinoshita' s remarks, he pointed out that the results 

 given were essentially a resistance and propulsion investigation on a methodical 

 series, and he agreed entirely that if these results were used on an actual de- 

 sign there might well be restraints on the movements of the LCB due to vibra- 

 tion and directional stability considerations. As to how this sort of behavior 

 might apply to very large tankers, this work was done some time ago and was 

 aimed at a tanker of about 75,000 tons. Certainly in a 250,000 tonner, the pro- 

 peller today would be proportionally less in diameter in relation to the drait and 

 the behavior might be somewhat different. Again, of course, as time goes on, the 

 tendency will be to have slower-rotating propellers of much bigger diameter, 

 but this trend may not be sufficient to overcome the effects just mentioned. In 

 regard to the question of loading, raised by Professor Prohaska, the same load- 

 ing was used throughout all the five model tests. In Lackenby' s opinion, the 

 peculiar changes with LCB position were a function of the shape of the sections. 

 Some other results for a model of the same block coefficient of 0.85 in which a 

 somewhat similar variation in LCB position was made, but where the section 



TURBULENT FLOW 



Rn 



L5xlO^ 5,0 X 10^ 



FOR 9- FOR 5- 



BLADED BLADED 



PROP PROP 



SHIP PROPELLERS 



Fig. 6 - Variation of the friction coefficient 

 with Reynolds number 



1660 



