Nowaekt and Sharma 
theoretical behavior in potential flow (see Appendix D) . However, in 
a real flow the decrease in effective wake with increasing loading can 
be explained qualitatively by a supposed contraction of the viscous wake 
due to propeller suction as first pointed out by Dickmann (1939) , see 
also next section . Fourth , all propulsion factors vary slowly and al- 
most monotonically with changesinloading, so that the arbitrary choice 
of one particular loading (e.g.that corresponding to the self-propulsion 
point of a ship of \ = 80) for further investigation is not liable to hide 
any important phenomena . 
Fig. 25 shows the various propulsion factors as functions of 
Froude number over the range 3.5 <¥% <12.5, all evaluated at the 
self-propulsion point of a smooth ship of Y = 80 . ( This choice of 
scale ratio is arbitrary , but not crucial as just pointed out ). The 
following features deserve special mention . First , all factors depict- 
ed exhibit a significant and oscillatory dependence on Froude number . 
Second , the self-propulsion point advance coefficient J,, , and con- 
sequently the equivalent open water efficiencies ng , depend mainly 
on hull resistance , and hence reveal humps and hollows in inverse 
phase to the coefficient of wave resistance (compare Fig. 5) as expect- 
ed . Third , contrary to common belief , the thrust deduction and 
effective wake fractions vary significantly with Froude number , the 
most remarkable feature being the sudden drop around Yo= 5 . The 
hull efficiency zz merely shows their combined effect . Fourth, the 
relative rotative efficiency np is exceptionally low but approaches 
normal values at higher Froude numbers . Fifth , there is an unusually 
large discrepancy between thrust and torque identity points , but it 
tends to decrease with increasing Froude numbers . The last two 
effects are presumably due to strong nonuniformities in the viscous 
wake of the hull , which would also explain why they are relatively 
weaker at higher Froude numbers . 
Iit."G. 2: ~“Wailee 
The next step in interaction analysis was an attempt to corre- 
late by theory the measured wake and thrust deduction . This required 
first the generation of a mathematical model of the propeller in the 
behind hull condition . Again the computer program described in 
Appendix C was used . The inputs to the program were the advance 
coefficient J), at the ship self-propulsion point , the corresponding 
thrust coefficient Ky, the radial distribution of measured nominal 
wake w (R), and the two-dimensional foil characteristics already 
established on the basis of open water characteristics ( see Propeller 
Analysis ) . In order to account for the difference between nominal 
and effective wake the program was allowed to determine by trial and 
error a wake corrector k,, , with which the nominal wake w(R) was 
1864 
