Ellis 



Fig. 4 - Bubble show- 

 ing asymmetrical col- 

 lapse and shock wave 

 centered off the sur- 

 face (200,000 frames/ 

 sec) 



such general features as jets might well exist in 

 the latter case as well. 



The brief outline of cavitation damage re- 

 search given here is merely intended to show mo- 

 tivation for work presently under way. Very com- 

 plete and excellent summaries are available 

 elsewhere, notably those of Flynn (27) and Eisen- 

 berg and Tulin (28). 



It was felt by the author that a logical attack 

 on the complete problem should first consider 

 those situations where comparison of theory and 

 experiment might be possible though difficult. To- 

 ward this end, experimental methods of generating 

 and observing collapsing cavities both in flow and 

 in stationary liquid were studied. The first step 

 was to duplicate the Naude'- Ellis work but in a 

 flowing system. Accordingly a glass-lined blow- 

 down type of water tunnel was constructed which 

 has been described elsewhere (29). This tunnel 

 was designed to provide velocities of over 100 feet 

 per second through a 2-3/4-inch-square working 

 section. Very-low-speed flows can also be gen- 

 erated covering a Reynolds number range from 

 zero to over two million. Initial cavitation experi- 

 ments included an initially hemispherical spark- 

 generated bubble at the low pressure point of a 

 circular-arc hydrofoil. The radius of curvature 

 and chord length was chosen such that the ideal 

 fluid pressure distribution would correspond to 

 that of the 1.5 -caliber ogive of Fig. 1, since this 

 distribution had yielded large well-defined bubbles 

 in the axially symmetric case. 



Figure 5 shows the resultant behavior. It can 

 be seen that even though the initial shape is hemi- 

 spherical the shape near collapse is not. The 

 center figure shows a shape very much like a 

 theoretical shape calculated by Naude' and Ellis for 

 an initially nonhemispherical bubble without flow. 

 The situation 10 microseconds later (bottom fig- 

 ure) is seen to consist of a smaller and more asym- 

 metrical bubble remnant and a radiated shock wave. 

 Although rather speculative, it seems likely that 

 the presence of flow caused a perturbation from 

 the hemispherical shape to develop in later stages 

 and that a jet occurred, hitting the solid at an 



142 



