speeds. For the smaller grain sizes, (90 and 400 micron), the surface inception 

 data indicate lower as or higher inception speeds as compared to the corresponding 

 TVC inception data. However, for the larger grain size (600 and 800 micron), the 



surface cavitation inception numbers approach or exceed the corresponding TVC incep- 



21 

 tion numbers. This may not pose any problems because recent studies have shown 



that distributed surface roughness cavitation is weekly dependent upon Reynolds 

 number while the TVC is strongly Reynolds number dependent. On this basis, a quali- 

 tative statement can be made that, for higher Reynolds number, surface roughness 



cavitation would appear at lower a or higher inception speeds as compared to TVC. 



21 

 However, it should be emphasized that the referenced work is somewhat limited in 



scope; i.e., essentially flat plate — zero pressure gradient — conditions at relatively 

 low Reynolds numbers with wire screen roughness. Additional research will be re- 

 quired before this issue can be better quantified. 



Based upon the above discussion and the data presented in Figures 22-25, the 

 600-micron grain size roughness appears to be the optimum grain size for these exper- 

 imental conditions. The 600-micron grain size resulted in the largest TVC inception 

 speed increase of 92 percent with no loss in foil performance — no drag increase or 

 lift decrease. 



The final phase of the roughness experiments involved two efforts. The first 

 effort dealt with an attempt to improve the TVC performance by increasing the treated 

 area of the foil pressure side. For this experiment, the 600-micron grain size was 

 applied to the area 0.75 in. (0.019 m) inboard from the tip, corresponding to an 

 area of approximately 1.8 percent of the total pressure side area. This treatment 

 represented an area approximately twice that of the standard which represented the 

 area 0.5 in. (0.013 m) inboard from the tip. The suction side area was kept the same 

 as the standard with 600-micron grain applied to the area 0.25 in. (0.006 m) inboard 

 from the tip. The foil performance with the increased treated area indicated an 

 increase in foil drag of approximately 4 percent over that of the standard. The cor- 

 responding TVC inception data, shown in the upper two curves of Figure 26, indicated 

 a slight increase in U„/U compared to the standard pattern. These slight increases 

 in TVC performance are not considered significant in light of the increased drag 

 penalty, and the "standard" configuration is taken to be the optimum roughness pat- 

 tern. The second effort involved an attempt to improve the 600-micron standard 



36 



