Borden 



0.60 



0.40 



- 0.20 



f 0.10 

 g 0.08 



0.06 

 0.04 



^^ ^^ ^ ^^^ 



0.02 



0.04 



0.08 0.1 0.2 0.4 



BOUNDARY LAYER THICKNESS IN INCHES 



0.6 0.8 1.0 



Fig. 12 - Effect of the pressure coefficient and the boundary layer 

 thickness on a three-dimensional cylindrical roughness (U„, = 26 

 knots, H^ = 15 ft, m = 7, T = 54° F) 



CONCLUDING REMARKS 



The foregoing analysis has not discussed such physical properties of the 

 fluid as dissolved air content, air nuclei, surface tension, and various chemical 

 and thermodynamic properties of the water. This paper is merely an analysis 

 of two sets of experiments performed in different laboratories. In both sets of 

 experiments, the data from cavitation inception on bluff bodies show a definite 

 dependence on Reynolds number. The Reynolds number effect was particularly 

 large on the cylinders and cones but was smaller on the hemispheres. The two- 

 dimensional triangular roughness elements showed less dependence on Reynolds 

 number than the bluff three-dimensional elements. The circular-arc rough- 

 nesses had a small height-to-length ratio and were not bluff enough to cause 

 separation. Cavitation inception on these roughnesses was essentially inde- 

 pendent of Reynolds number. 



Although both studies indicate that small changes in air content did not af- 

 fect cavitation inception speeds significantly, these physical parameters should 

 be investigated further. The experiments should be repeated in other flow fa- 

 cilities, in other types of water, and at different air contents and temperatures. 



Despite the limitations of the roughness data disclosed here, the results 

 point up the need for considering viscous effects in the scaling of cavitation on 

 bluff bodies or on roughness elements where flow separation may occur. Dis- 

 continuities from peeling paint or lapping seams can be likened to the 



198 



