Panel Discussion 



his firm had made quite a few propellers weighing about 50 tons and costing 

 $120,000, a serious matter for the customers. They found lower efficiencies 

 with these very large ships and also that the propellers had a lower efficiency 

 in the tank compared with the computer design. Tests at the NSMB at different 

 Reynolds numbers showed that this effect could be fully explained by the low 

 Reynolds number at which the propellers were tested in open water. Van Gun- 

 steren believed that the ship models were too short for the large tankers, and 

 that to get a correct prediction the model must be certainly larger than 8 m, 

 giving a scale ratio of more than 1 to 50 for the propeller. 



Lindgren agreed that this was a problem and also with the need of higher 

 Reynolds numbers for the big tanker models. The philosophy at Gothenburg was 

 that with the new cavitation laboratory it will be possible to study the influence 

 of Reynolds number up to appreciably higher values than at present because 

 there will be no free water level in the cavitation tunnel. On the other hand, he 

 would like to know a little more about what Van Gunsteren meant about this scale 

 effect due to low Reynolds number. Did he mean that laminar separation is 

 present when carrying out propulsion experiments or does he mean that laminar 

 flow occurs around the propeller profile? What was his hypothesis? 



Van Gunsteren replied that conditions are under- critical if there is laminar 

 flow at the root sections, and referred to a diagram given in the famous work of 

 Troost, Van Lammeren, and Van Manen. Dr. Gutsche proposed a critical Reyn- 

 olds number based on propeller diameter and revolutions. If the propellers for 

 models of very large tankers are plotted in this diagram, it will be found that 

 they are in the region where, based on this Reynolds number, the flow is going 

 from turbulent to laminar. So the propellers are in the under-critical condition. 

 Moreover, the effect is not the same for all types of propellers. If another kind 

 of section or a different type of propeller is used, open water tests in such a 

 critical range may show that one propeller is better than another, while behind 

 a ship both propellers might be equal or the relative efficiencies may even be 

 reversed. 



Lindgren said that the Gutsche under-critical Reynolds number means that 

 there is laminar separation on the profiles and is related to the performance in 

 homogeneous flow in the open-water condition. He did not think that the Reyn- 

 olds number in the open condition could be compared with that calculated using 

 the wake fraction in the behind condition. They are operating in quite another 

 degree of turbulence. What that really means is not known but it must have a 

 very marked influence on the performance. 



Dr. Morgan wished to make one point about the scale or Reynolds number ef- 

 fect on propellers. There seemed to be some confusion here and it might be as 

 well to point out that the work that Gutsche did for propellers in uniform flow 

 does not apply to the entirely different conditions behind the model. The friction 

 coefficient variation with Reynolds number is something like that shown in Fig. 6 

 with a transition from laminar to turbulent flow. This transition on a propeller 

 blade of airfoil section is roughly about 5 > 10^ at 0.7 radius. However, if the 

 propeller is operating in turbulent flow or is rough, it falls upon the upper tur- 

 bulent line. The friction coefficients on model propeller blades may be three 

 times higher than those on the full scale, and it is necessary to be very careful 

 in analyzing the data when comparing different propellers. Such a problem was 



1659 



