88 



TABLE I Turbulence Levels in Some Water Tunnels 



Ottawa, Canada 0.75% 



Kriloff No. 2 

 Leningrad, USSR 0.4% 



NPL No. 1 

 Feltham, UK 0. 5% 



MIT 

 Massachusetts, USA 0.77% 



6" Tunnel 

 Minnesota, USA 0.8% 



ORL 

 Pennsylvania State, USA 0.8% 



HSWT 



California Institute of 

 Technology, USA 



0.25% 



LTWT 

 California Institute of 

 Technology, USA 0.05% 

 (present work) to 3.75% 



carried out at the Garfield Thomas Water Tunnel 

 (GTWT) noted tJiat there appeared to be no laminar 

 separation on a hemisphere nose body when polymer 

 was added to the water, but no direct flow visual- 

 ization was done. Later van der Meulen (1975) 

 verified this speculation with the clever use of 

 schlieren holography to observe simultaneously the 

 viscous flow and cavitation inception on a 10 mm 

 diameter hemisphere nose body. His results showed 

 clearly that when polymer was injected into the 

 boundary layer that the laminar separation was re- 

 moved, van der Meulen suggested that the polymer 

 removed the separation by causing an early transi- 

 tion to a turbulent non-separating boundary layer. 

 He then attributed the suppression effect to the 

 removal of the large pressure fluctuations associ- 

 ated with the transition zone of the free shear 

 layer [Arakeri (1975)]. 



Freestream Nuclei 



It is generally accepted that inception begins at 

 the nuclei in the liquid and that there are two 

 sources for these nuclei — the test body surface 

 and the incoming flow. At one time "surface nuclei" 

 received considerable attention [e.g. Acosta and 

 Hamaguchi (1967) , Holl and Treaster (1966) , Holl 

 (1968), Peterson (1968) and van der Meulen (1972)]. 

 While on the one hand it was shown that under certain 

 circumstances, but not in normal cavitation testing, 

 surface nuclei could exert a controlling influence 

 upon inception. It seemed evident on the other hand 

 from the results of the ITTC tests that freestream 

 nuclei were the more important. Further, the de- 

 velopment of the concepts of cavitation event count- 

 ing [Schiebe (1966) ] in conjunction with Johnson 

 and Hsieh's (1966) trajectory calculations, the 

 idea of "cavitation susceptibility" [Schiebe (1972)] 

 and the development of equipment to measure free- 

 stream nuclei populations have led to more interest 

 in the influence of freestream nuclei versus surface 

 nuclei. In particular, the experiments of Keller 

 (1972) have prompted considerable interest in mea- 

 suring and relating freestream nuclei populations 

 to inception. 



Morgan (1972) has reviewed the various types of 

 instruments available for measuring freestream 

 nuclei populations and Peterson et al. (1975) have 

 made an experimental comparison of three of these, 

 namely; light scattering, microscopy, and holog- 

 raphy. At the moment holography seems the best 



in that no "calibration" is required, a permanent 

 record is obtained, a large volume is sampled, and, 

 as Peterson observed, one can determine if the 

 nuclei are solid particles or micro-bubbles. 



There is seen to be ample reason then to pursue 

 these various freestream factors in inception re- 

 search. Two are primarily fluid-dynamic in nature 

 and of these the questions concerning freestream 

 turbulence levels are of historic interest in fluid 

 mechanics and naval architecture. The cavitation 

 nuclei however are directly involved in the cavita- 

 tion inception process and the recent experimental 

 progress cited above make one hope for a more quan- 

 titative predictive ability than in the past inso- 

 far as inception is concerned. The present work is 

 in the mainstream of these observations; briefly we 

 report on observations made in two different flow 

 facilities having widely different freestream nuclei 

 distributions on identical bodies. In one of these, 

 the freestream turbulence level is varied over nearly 

 a factor of 100 (but not in a condition of cavita- 

 tion then) and we confirm and extend the observa- 

 tions of van der Meulen on the polymer effect. 

 Schlieren photography is extensively used to visu- 

 alize thermal boundary layers on the test bodies 

 used and in-line holography is used to determine 

 nuclei populations in the working section. 



Before discussing these effects we should com- 

 ment briefly on the means used for the determination 

 of tJie actual inception observation. A standard 

 procedure has been to observe the test body under 

 stroboscopic light and to say that inception occurs 

 when macroscopic cavities or bubbles become visible 

 on the model. However, this method is observer- 

 dependent and the trend now is to use cavitation- 

 event counters free of human judgment. Ellis et al. 

 (1970) and Keller (1972) have developed optical 

 techniques which count interruptions of light beams 

 which are adjusted to graze the model surface where 

 inception has been observed to occur. Peterson 

 (1972), Brockett (1972), and Silberman et al. 

 (1973) have also determined inception acoustically 

 by locating a hydrophone inside the test model . 

 There are problems of identifying the types and 

 location of the cavitation phenomena occurring with 

 these "events." Aside from the question of tech- 

 nique, there is also the question of selecting 

 appropriate threshold levels at which an event be- 

 comes countable and also the event rate at which 

 inception is defined to occur. At present there is 

 no universal agreement of just what these values 

 should be. 



