applied only to those areas of the foil tip whose streamlines contributed directly 

 to the forming tip vortex at the foil tip. The paint flow patterns at design a, 

 (Figure 13) indicated that only the extreme tip area of the foil pressure side — less 

 than 0.7 percent of the planform area — had to be treated. In addition, the visual 

 observations — Figure 15 — of the TVC revealed that the TV attachment point remained 

 relatively fixed on the very tip of the foil suction side. On this basis, it was 

 felt that roughness also applied in this area — less than 0.4 percent of the planform 

 area — would act in a beneficial manner to reduce the TV core velocities without in- 

 ducing substantial drag penalties. The roughness grain sizes were large enough to 

 provide substantial increases in the tip boundary layer thickness, but also were 



small enough to avoid local surface cavitation. Since the tip vortex is dependent 



4 17 



upon the tip boundary layer thickness, the roughness size should scale according 



to Reynolds number. Figure 8 illustrates the parent foil tip, pressure side, area 



which was treated with artifical roughness. 



The initial roughness experiments involved the evaluation of the three config- 

 urations: pressure side only, suction side only and both pressure and suction sides. 

 A 400-micron grain size roughness was selected and the roughness was applied as dis- 

 cussed above. The effects of the roughness on pressure and suction sides on the foil 

 lift and drag were well within the acknowledged 2 percent accuracy of the experiment. 

 These results are not totally unexpected, considering the relatively small areas that 

 were treated. Similar force results were obtained for the other two roughness con- 

 figurations, suction side only and pressure side only. 



The cavitation characteristics of the three roughness conf igurafions are pre- 

 sented in Figure 22. These data correlate well with Equation (1), when compared to 

 the parent foil, the indicated k values give corresponding increases in TVC inception 

 speeds of 41 percent for pressure side roughness, 28 percent for suction side rough- 

 ness, and 67 percent for pressure and suction side roughness. Note that the improve- 

 ment in cavitation inception for the suction side only and pressure side only rough- 

 ness appear to be additive. Based upon these results, the optimum roughness pattern 

 required both pressure and suction side roughness at the tip. 



The next phase of the roughness experiments dealt with the optimization of the 

 roughness grain size. For these experiments the standard or optimum roughness 

 pattern was used for each of the grain sizes: 90, 400, 600, and 800 microns. 



31 



