Malavard 



of which 0.056 corresponds to the thrust provided by the nozzle and 0.202 to that 

 of the propeller. , 



A model of the ducted propeller given in Fig. 26 has been tested at the 

 Bassin des Carenes, Paris. The results obtained were very encouraging, since 

 the thrust of the nozzle corresponded well to what was expected, as did the ef- 

 ficiency of the overall propeller. The conclusions drawn were 66% experimental 

 instead of 70% theoretical. The total thrust, however, was only attained to within 

 16%. In any case, with regard to the so-called nozzle itself, the study showed 

 the advantage of the method of calculation used: if improvements should be 

 sought, they ought to be concerned with the calculation of the fan and of the 

 straighteners. To support this argument it may be noted further that the com- 

 parison of the efficiency of this nozzle with a Wageningen no. 9 nozzle fitted with 

 a K 4.55 propeller, which was considered to give the best performance, resulted 

 in a preference for the former. With a practically equal diameter, the gain in 

 efficiency of the first nozzle is about 5 to 13% higher depending on the power. 



STUDY OF FLOW AROUND THICK BODIES 



In most studies already described the bodies are supposed very thin. In 

 this case the linearization hypotheses are valid. Nevertheless, when the rela- 

 tive thickness of the bodies is important it is not possible to simplify the bound- 

 ary conditions over their surface. Thus, if we consider the perturbation veloci- 

 ties potential 4- = ^ - x, with a harmonic function in x, y, and z, it is conven- 

 ient to write, for the whole surface S limiting the body, the tangential velocity 

 condition as - - 



— = f(Xi,yi,z.) , 

 an 



where f(x.,yj,z.) is a known function of points M(x. ,y . , z- ) over the surface 2. 

 This function depends on the local slope of the body and its motion. 



The body can be slightly immersed beneath, or can traverse, the free sur- 

 face. In general, it is possible to simplify the equilibrium condition of this 

 surface by supposing that the perturbation induced by the body is not very im- 

 portant. The linearized boundary condition on the free surface still holds good 

 and can be written in the same form that in the General Equations subsection, 

 described earlier 



^+ K ^- 



Far upstream we have the condition 



lim grad cp - 



x-»-oo 



This problem can then be solved according to the method described in the 

 subsection on Subcavitating Hydrofoils. The (p function, which is the solution of 



406 



