aay be that firsts the effect of viscosity lo not included in the confutation, 

 and second, the test was done with turbulence stimulation. The top figure 

 lndicatea that for an angle of attack of 4 degrees, the computed lift 

 coefficient is about twice the measured value when the submergence is large. 

 When the ratio of submergence to chord is smaller than 0.5, the agreement 

 between computation and experiment is good. For small submergences, the lift is 

 more affected by the free surface than viscosity, and when the submergence is 

 large, the reverse is true. On the other hand, compared with the experiment, 

 the results of Wadlln and Christopher show better agreement than those by the 

 present method when the submergence ratio is larger than 2. 



When the location of the sail plane or stern plane of a submersible Is at a 

 depth of two or three times of chord, a better lift computation can be expected 

 with the method developed in Reference 17. 



Figure 16 shows forces and moment for Model 4621 with and without stern- 

 planes at deep submergence. The computed and experimental axial forces are 

 fairly steady at different trim angles. The vertical forces for the bare hull 

 are computed to be significantly less than those of experiment. However, the 

 results of vertical forces with sternplane agree very well with those of experi- 

 ment except at a»12°. The reason for the good agreement for the case with 

 sternplane is that the vertical force (in this case lift) of the control plane 

 g'one is over-estimated at deep submergence as shown In Figures 14 and 15; and 

 this over-estimation is compensated with the under-estlmation of the vertical 

 forces of base hull. The moments are computed to be larger than the experimen- 

 tal values. 



Figure 17 shows the results of vertical force of a spheroid with L/DW near 



18 

 a wall. The computed results are compared with the results of Newman. 



Newman's method was developed using slender body theory and with the assumption 



of L/D » i and h/D << !. The vertical forces computed by the present method 



are smaller than these computed by Newman. It is unkown which method is more 



accurate. 



25 



