PREDICTION OF CAVITATION INCEPTION 

 SPEEDS ON ROUGH HYDRODYNAMIC 



BODIES 



Avis Borden 

 Naval Ship Research and Development Center 

 Washington, B.C. 



ABSTRACT 



Experimental studies of cavitation inception over two- and three- 

 dimensional roughness elements in a boundary layer (made at the Ord- 

 nance Research Laboratory (ORL) of Pennsylvania State University and 

 at the David Taylor Model Basin, respectively), have been used to derive 

 scaling laws for predicting cavitation on these elements under arbitrary 

 boundary layer conditions. It has already been shown that the cavitation 

 number of three-dimensional elements, computed in terms of the veloc- 

 ity of the oncoming flow in the boundary layer at the height of a rough- 

 ness element, is a function of the local Reynolds number, based on 

 roughness height and local velocity. In the present study, the ORL data 

 have been analyzed in the same way, and analytical curves have been 

 fitted to both sets of data. Methods are derived for computing cavitation 

 inception speeds on rough hydrodynamic bodies. Sample calculations 

 show the degrading effect of the various types of roughnesses on the 

 cavitation inception speed of a typical sonar dome. 



INTRODUCTION 



The prediction of cavitation inception on hydrodynamic bodies, such as ship 

 hulls, appendages, sonar domes, and propellers is one of the important unsolved 

 problems in naval architecture. The cavitation inception speeds predicted by 

 model tests and theoretical analyses are never attained under full-scale oper- 

 ating conditions. This disparity is partly due to the physical size of the body and 

 its surface finish. Recent experiments have shown that cavitation inception on 

 bluff bodies and those having sharp pressure minima is primarily a function of 

 Reynolds number. Reference 1 describes experiments on families of ogive and 

 disk models at several water temperatures in which Reynolds number trends 

 were definitely established. Reynolds number effects have also been observed 

 in the scaling of tip vortices from propellers (2). 



In a series of experiments recently completed at the David Taylor Model 

 Basin (3), what appeared to be a Reynolds number scaling was found for 



183 



