167 



2 O 



1-9 



+ MODEL EXPERIMENTS 

 PREDICTIONS 



°/o PROPELLER RADIUS 



FROM HULL. ^.^^ 2309 



4192 PREDICTION FOR 

 230., I2.5W 12.5-/0 RADIUS. 



1-8 



I-7 



u. 1.5 



1-4 



1-2 



I- I 



l-O 



t 



UNPOWERED 



MODEL SELF 

 PROPULSION 



10 II 1-2 1-3 1.4 



PROPELLER DIFFUSION RATIO. 



FIGURE 17. Relative magnitude of the secondary flow 

 for vehicle B. 



I- I 



+ MODEL EXPERIMENTS 

 PREDICTIONS 



°/o PROPELLER RADIUS 

 FROM HULL. 



l-O 



0-9 



0-8 



0-7 



^UNPOWERED 



MODEL SELF 

 PROPULSION 



10 II 1-2 1-3 



PROPELLER DIFFUSION RATIO 



FIGURE 18. Relative magnitude of the velocity defect 

 for vehicle B. 



10 



0-9 



2o-8 



O 

 £0-7 



Ul 



a 



< 0-6 



UJ 



> 

 0-4 



MODEL DATA I VEHICLE A 

 MODEL DATA % VEHICLE B 



°/o PROPELLER RADIUS 

 FROM HULL 

 33-5 gi 12-5 I 250 



65'0 



sEMPIRICAL VALUES 

 FOR APPENDAGE 

 GEOMETRY. 



I 



_1_ 



O 0-2 0-4 0-6 0-8 l-O 



DISTANCE FROM CONTROL TRAILING EDGE 



LOCAL CONTROL SURFACE CHORD 



FIGURE 19. Variation of model velocity defect with 

 distance from the control surface. 



if this is used for the predictions the above 

 differences become + 1 percent and - 2 percent 

 respectively. Thus the speed trial and full-scale 

 wake data become compatible and both suggest a 

 scale effect on the flow velocity for the vehicle 

 A with the finer stern much smaller than predicted. 

 This may be due to the fact that the full-scale 

 vehicle is hydraulically rough at all but the very 

 lowest speeds while the prediction method assumes 

 hydraulically smooth conditions. 



The process of adding the predicted mean circum- 

 ferential velocity changes to all measured veloci- 

 ties are described in Section 4 naturally leads to 

 a change in the ratios used herein to describe the 

 relative magnitudes of the velocity defect and 

 secondary flow. For the velocity defect Figures 15 

 and 18 show that the predicted magnitude decreases 

 slightly with increasing diffusion ratio such that 

 at model self-propulsion the relative magnitudes 

 are 3 percent higher than measured for body A and 

 up to 6 percent for body B . The predicted relative 

 magnitude of the wake defect at the full scale 

 condition is within 2 percent of that measured, 

 although as already noted the absolute velocities 

 are 15 percent and 9 percent higher than measured. 

 The use of a smaller scaling effect based on the 

 equivalent Reynolds number discussed above would 

 slightly reduce the above error in predicted 

 velocity defect. 



The predicted relative magnitude of the secondary 

 flow can be seen from Figures 16 and 17 to decrease 

 with increasing diffusion ratio but at a slower 

 rate than actually measured on the models. Thus, 

 the propeller is having an influence on the develop- 

 ment of the secondary flow in addition to the simple 

 change in relative magnitude arising from the 

 propeller induced velocity. It is clear that at 

 model conditions, the difference between the 

 measured and predicted secondary flow is much 

 greater for the blunter afterbody form of vehicle 



