0. > -c hemisphere body (9) 



1 ~" P- 



and 



a. < -c ITTC body (10) 



1 _ P„ 



Although there is some improvement when the reattachment pressure co- 

 efficient is used,* the two bodies are seen to behave quite differently. 

 And the original observations of Parkin and Kermeen (namely that the local 

 pressure exceeds vapor pressure yet the cavitation still occurs) on the 

 hemisphere body are confirmed once again. 



Before moving on to questions raised by these data, it is appropriate 

 to point out that in this section we have been dealing with attached forms 

 of cavitation which are seen to occur within a boundary layer separation. 

 It seems clear that the sheets, or band forms of cavitation, are due to this 

 separation. In the Caltech experiments so far reported, this type of cavi- 

 tation occurs without travelling bubble cavitation. But we see, for example 

 on the ITTC body in Figure 6, that other forms of cavitation also occur and 



it should not be supposed that the approximate correlation, o. - -c , is 



s 

 valid for them. Indeed, the great variability of the ITTC inception 



results of Figure 5 makes it plain that each of these forms of cavitation 



inception will have its own governing dynamics. Of course, one is left 



with the problem of determining, in a particular flow situation, the 



particular form of cavitation, so that the appropriate physical model may 



be used to determine the scaling procedure. 



Conditions within the Separation 



There is still the problem of explaining cavitation on the hemisphere 

 body when the local pressure (in the separation region) exceeds vapor 

 pressure. (We exclude gaseous cavitation here by the slowness of the 

 diffusion process.) The implication is clear that there must be local, 



*See Arakeri (1973) for these measurements. 



27 



