the sea are known, the ship model experimenter will not be aware of the basic conditions 

 that should be simulated in his propeller cavitation tests. 



The malignant disease known as propeller erosion, first recognized about 60 years 

 ago, is still with us despite a good deal of improvement in propellers as regards hydro- 

 dynamic design, metallurgy, and the properties of the material. 



The root of the propeller is the region most frequently affected, and close atten- 

 tion is directed toward shaping the root sections, root fillets, and propeller boss, accord- 

 ingly. 



Collapse pressure of bubbles and corrosion, separately or in combination, have 

 been adduced as causes and with different emphasis, according to the particular investi- 

 gator concerned. Perhaps the author would comment on this, and on the recent claims 

 to the effect that the life of the propeller can be prolonged by local cathodic protection. 



Most of the bubbles in the boiling water around a ship's propellers are so minute 

 as to provoke the thought that it should be possible to devise means for their elimination. 

 An ad hoc attempt by myself to achieve this by injecting silicones into the water to 

 reduce the surface tensions was not at all encouraging. Perhaps the author would remark 

 on the possibility of such direct attack on the bubbles to remove them as soon as they 

 are formed. 



K. F. Herzfeld 



I would like to make some comments on the nuclei which are necessary as 

 origin for a small bubble which will then grow, either in boiling or cavitation. The 

 first question is whether bubbles start preferentially at a smooth wall, as compared 

 to the interior of the liquid. While there does not, to my knowledge, exist any sys- 

 tematic investigation of the subject, I have an impression that a smooth wall is not a 

 favorite origin for bubbles. On the other hand, some photographs by Dr. Snay seem 

 to contradict that; here cavitation bubbles form on the smooth lucite wall, either upon 

 reflection of the shock wave or in the negative phase of the bubble pulse. On the 

 other hand, a wad of glass wool is a favorite starting place [1], and it is known that 

 some rough pieces intiate boiling. 



Next, what is the nature of the nuclei in the liquid? Fox and I had proposed 

 [2] that they are microscopic gas bubbles, covered by an organic skin, which prevents 

 them from dissolving. This could explain why the previous application of pressure 

 prevents the start of cavitation, since it would crush the skin. However if this were 

 so, there should be a lower limiting pressure — that necessary to crush the skin — below 

 which previously applied pressure would have no effect on cavitation. This disagrees 

 with experiment. According to Strasberg [3] there is no such lower limit, but the 

 tension necessary for cavitation to set in under influence of ultrasonic waves increases 

 smoothly with the amount of pressure previously applied. These and other experi- 

 ments seem to exclude the assumption of gas bubbles with organic cover as necessary 

 cavitation nuclei. 



This leaves as the other alternative solid particles. These must be non-wetted 

 by the liquid and carry undissolved gas, probably in a crack, which acts as the starter 

 of a growing bubble. This assumption has been mainly proposed by E. Meyer at the 

 University of Goettingen and experiments carried out with iron oxide seem to confirm 

 it. 



A few theoretical remarks should be added on the origin of cavitation damage. 

 On principle a collapsing bubble produces a very high pressure and a high temperature. 

 If the shock wave produced in the metal were plane I doubt that it would produce 

 pitting; a purely compressional wave does not produce such damage. On the other 

 hand, due to the local concentration, there is probably some shear involved, which 

 might be responsible for damage. 



If the mechanism is due to high temperature, it might be threefold: direct, 

 by softening or melting; by sudden liberation of gases previously dissolved in the metal; 



319 



