53 



droplets during the oscillation which cause damage on impact. Before definite 

 statements as to the origin of damage in this case can be made, however, it 

 will be necessary to investigate the accelerations that can be produced in the 

 liquid during the oscillation and whether detached water droplets can acquire 

 sufficient momentum toward the solid material. 



It has been suggested to the writer that the collapse of such cavi- 

 ties might be associated with condensation phenomena within the vapor itself. 

 Such conditions might be likened to the conditions leading' to condensation 

 shocks observed in high-speed air or steam flows. 80 ' 81 » 82 » 83 However, such 

 condensation is associated with conditions of supersaturation in the vapor 

 phase and it is not likely, in cavitating flows of the type discussed here, 

 that the expansions required for supersaturation can occur. It is possible 

 that at various points throughout the vapor phase, supersaturated conditions 

 may exist momentarily, but equilibrium would be quickly established without 

 condensation throughout the vapor phase. 



In the case of the cavitating venturi a possible mechanism of the 

 observed oscillations, and thus of the collapse process, has been described 

 in the foregoing — both for the transverse and for the longitudinal motion. 

 Here, again, however, it is not clear that the longitudinal and transverse 

 oscillations are of sufficient strength to cause damage. If they were, damage 

 might be expected over the entire length of the nozzle, whereas, to the 

 writer's knowledge, the damage is usually observed near the end of the cavity, 

 at least for these experiments in which the position of the end of the cavity 

 is fairly stable. Whether the damage in this case is again associated with 

 the small jets and drops which must be present when the central core of liq- 

 uid impinges on the filled portion of the diffuser remains to be investigated. 



IDEAS ON ELECTROCHEMICAL ORIGIN OP DAMAGE 



There has been some discussion recently of cavitation damage due to 

 electrochemical phenomena associated with the rapid formation of new surfaces 

 in a liquid. This theory associates the formation of surface double-electric 

 layers with a subsequent spark discharge and dissociation which supposedly con- 

 tributes to damage. 84 It should be noted, when considering these theories, 

 that the mechanical theory of corrosion is supported by experiments in which 

 cavitation damage was obtained on plate glass and on a lead plate without loss 

 of weight. 59 It is not clear how a spark can originate during cavitation in 

 the interior of a liquid since it would seem that the orientation of dipoles 

 at the cavity surface would lead to no electrical unbalance. In the case of 

 cavitation at a solid, it is again not clear that a current would flow across 

 the "gap" when a path of much lower resistance from the water to the solid is 



