Nevertheless, there is considerable appeal of the idea that a more 

 realistic appraisal of the actual features of the real viscous flow around 

 bodies of interest may help considerably in improving our knowledge of 

 cavitation inception and the development of better scaling laws. This, in 

 turn, leads us to reflect on just what the locations of inception might be 



(and hence the value of c there) and the strength of any unsteady 



loc 

 fluctuations there — without, for the moment, worrying about how the cavi- 

 tation might originate. With this in mind, we summarize some of these real 

 fluid features that might influence inception on smooth bodies: 



1. The laminar-separation reattachment region. 



2. A "natural" transition. 



3. Stimulated or "tripped" boundary layers. 



4. Freestream effects, e.g., turbulence, particulates, 

 polymer solutions. 



We cannot deal extensively with these well-known fluid mechanics 

 problems; much of the presently-known structure of laminar-separation 

 bubbles may be found in Gaster (1966) and the principal new features here 

 are the astoundingly large pressure fluctuations as already mentioned. 

 There has been recent progress in determining the length of these 

 separations, and of the turbulent transition process seen on the free shear 

 layer of these flows (Van Ingen 1975) . But it is a well-known fact that 

 the steady, laminar flow on smooth bodies gives way to a turbulent flow at 

 a sufficiently large Reynolds number and that, if this transition is 

 completed upstream of a laminar separation, this separation will disappear. 

 Flow transition may be stimulated by many methods if the Reynolds number is 

 not high enough for a "natural" one — as it, indeed, has been done histori- 

 cally in naval hydrodynamics. And, if by this process, a prior laminar 

 separation is for stalled, one would expect major consequences for the cavi- 

 tation inception process there. Other traditional, and some nontraditional, 

 issues naturally arise when transition in laboratory facilities occurs. 

 Of these, surely that of freestream turbulence has been one of the most re- 

 curring ones. In hydrodynamic applications, solute molecules and particu- 

 late suspensions may have profound effects on both laminar and turbulent 



31 



