10 



Thus, some care must be exercised in relating the results of these different types of experi- 

 ments to practical systems in which the actual onset is of interest. It would seem to be of 

 rather significant interest to know the maximum possible critical cavitation number for any 

 given system and, further, to know a "critical" cavitation number at which adverse effects 

 (noise, loss in performance, etc.) first begin. (These numbers are not, of course, necessarily 

 the same. As a matter of fact, this is only the question of the definition of cavitation onset 

 since, presumably, nuclei will be affected by any pressure field whether or not they are seen 

 or, as a result of the motion, radiate noise.) 



A REMARK ON THE QUESTION OF GAS 

 CONTENT AND NUCLEI CONTENT 



It seems clear from the foregoing remarks and the discussions in Reference 3 that the 

 inception of cavitation in technical liquids is intimately associated with the presence of 

 nuclei whether as free gas or vapor bubbles or bubbles trapped on solid particles. That the 

 role of gases in liquids is well recognized is reflected in the now almost standard procedure 

 of measuring air-content in experiments in which cavitation phenomena are involved. However, 

 the measurements that have been reported so far have all been of the total air and gas content, 

 i.e., both dissolved, and entrained as individual bubbles. On the basis of the previous dis- 

 cussion, it seems unlikely that completely dissolved gases can play an important role in the 

 inception of cavitation* since, in this situation, very high tensions would be expected. Thus, 

 it is the opinion of the writer that further progress on inception may depend, to a considerable 

 extent, on the development of theoretical and experimental methods for the characterization 

 and measurement of the nuclei content of liquids.** These studies should eventually result 

 in descriptions of the size distribution and quantity of undissolved gas nuclei— in other words, 

 a "spectrum" of nuclei. 



With such information available, it would then be possible to undertake correlation 

 studies of inception of cavitation as a function of the nuclei spectrum. Judging from the ex- 

 periments that have been made and remembering the many factors which must be considered 

 (turbulence, physical-chemical properties of the liquid, other foreign materials, etc.), it seems 

 not unlikely that under a given set of conditions, it will not be possible to prescribe a unique 

 critical cavitation number but only a "most probable" critical cavitation number. 



*Except insofar as the equilibrium conditions between dissolved gas and undissolved nuclei are concerned. 

 **Such studies are now underway at the David Taylor Model Basin. ; 



