solution of integral equations representing the flow conditions over the head and tail 

 contours H and T and over the free stream surface M. The vortex sheet is, of course, 

 of varying strength over H and T, and of constant strength over M. 



Vandrey did not apply his technique to any case of cavity flow but he showed 

 that the method gave satisfactory agreement with experiment in three cases of fully 

 wetted axi-symmetric flow; it is clear from these examples that his iterative method 

 of solving the second order integral equations will, after a few iterations, usually lead to 

 a satisfactorily convergent solution although the computational labour involved may be 

 considerable. This work of Vandrey was followed up shortly afterwards by the investi- 

 gations of Armstrong and Dunham (1953) who formulated a similar method for 

 obtaining an exact iterative solution of the Riabouchinsky model for axi-symmetric 

 cavitating flow around a flat disc held normal to the main stream. In order to simplify 

 the numerical work associated with their method they made use of certain similarity 

 principles between the two-dimensional and three-dimensional flows. The results of 

 these calculations give cavity dimensions which agree well with the available experi- 

 mental data; Fig. 7 shows satisfactory agreement between the calculated cavity length 



Figure 7. Variation of length coefficient with cavitation number for flow past disc. 



and the observations of Rouse and McNown (1948) in the Iowa cavitation tunnel. 

 The modified drag coefficient C D r — C D /{\ + Q) calculated by Armstrong and Dun- 

 ham was in good agreement with the Fisher theory; the numerical values, were, how- 

 ever, some 3% greater than those calculated by Plesset and Schaffer using their simili- 

 tude technique which will be described later (Fig. 8). 



During the last year Garabedian (1955) has described further investigations of 



222 



