Panel Discussion 



torque fluctuations, and it is hoped to report about this in a year's time, when 

 the ITTC meets in Rome in 1969. This relation, of course, holds only as long 

 as Ug /Ug is smaller than unity. This means also that the foregoing observa- 

 tions do not allow any direct conclusions regarding the scale effect on the volu- 

 metric mean value of the wake. But this can be easily obtained by integrating 

 over the disk area. 



If the scale effect is considered as a wake problem, it is found that the maxi- 

 mum wake behind a strut or shaft bracket on the ship is approximately equal to 

 the maximum wake on the model, which can be measured, times the ratio of the 

 drag coefficient C^g for the ship appendage to the drag coefficient C^^ of the 

 model appendage, to the power of 1/2. This is shown in books on boundary- layer 

 theory, such as Vol. 3 of Durand's "Aerodynamic Theory" in Division G— The 

 Mechanics of Viscous Fluids. Now, this would not be very serious if the two 

 drag coefficients had near enough the same value, but it can happen that the 

 Reynolds number of the model appendage is of the order of 8 ^ IC* and the Reyn- 

 olds number of the ship appendage may be about 10^. For a shaft strut with a 

 thickness-chord ratio of 20% (which is a little high, perhaps) the drag coefficient 

 of the model strut could be 0.07 and the drag coefficient of the full-scale strut 

 about 0.01, which means a ratio of seven to one. The appendage drag is thus 

 very sensitive to scale effect, and is very often overestimated. (This is a very 

 important matter and the effects can be estimated from the diagrams given in 

 Figs. 4 and 5.) 



The chairman announced that there were no other written contributions whose 

 authors were present and declared the meeting open for anyone to talk about any 

 aspect of propeller-hull interaction. He was sure that many questions in this 

 field had come up in the past, and here was an opportunity to voice them and 

 hear what other people think about them. 



H. B. Lindgren (The Swedish Tank, Gothenburg) said he would like to come 

 back once more to the question raised by Dr. van Manen a while ago dealing with, 

 in his opinion, the very important question of the possible scale effect on the thrust 

 deduction factor. His reaction was very similar to the one previously expressed 

 by Bavin. He had a strong feeling that Dr. van Manen must have overestimated 

 the importance of this question. In this connection, he did not think it necessary 

 to build such an extremely big new cavitation laboratory to find the solution to 

 this problem. In Gothenburg they were just now finishing a new cavitation tunnel 

 in which it was possible to install a complete ship model of a little more than 8 

 m in length, and in that it will be possible to study the propeller effects behind 

 this big ship model. The Swedish Tank has been carrying out tests for a long 

 while with models in the existing cavitation laboratory, and if there were such 

 very big influences of cavitation on the thrust deduction factor, Lindgren was 

 quite sure that they would have been detected when carrying out those experi- 

 ments, because they were made under atmospheric pressure as well as under 

 cavitating conditions. 



The chairman believed that there was still a lot of vagueness about the thrust 

 deduction. A great deal of model work done in the past has had very contradictory 

 findings. The work on the Victory ship models at the Wageningen Tank showed a 

 very large scale effect on thrust deduction t , increasing quite materially and 

 rapidly with increase in size of model, and this continued right up to the 72-ft 



1655 



