62 ■ SOME EXPERIMENTS WITH MODELS HAVING RADICAL 



ment, 2,500 pounds in fresh water; midship section coefficient, .98; longitudinal 

 coefficient, .6122; block coefficient, .60. 



Fig. I, Plate 13, shows the original lines, including stern A and the curve of sec- 

 tional areas for all models. This was one of the models tested last year, the re- 

 sults being included in my paper of last year on "Relative Resistances of Some 

 Models with Block Coefficient Constant and other Coefficients Varied." Sterns B, 

 C, D, E and F are shown in Figs. 2, 3, 4, 5 and 6, Plate 14. Fig. 7, Plate 15, shows 

 the curves of total resistance and the estimated wetted surface resistances, and Fig. 

 8, Plate 16, shows the residuary resistances. 



The underlying idea of stern B was a shape which would start the water up 

 more quickly than in the standard type. The idea of stern C was to carry right aft 

 sections derived from the midship section (98 per cent coefficient), beam only being 

 decreased ; in other words, to make a wall-sided model. Stern D was stern C dis- 

 posed horizontally instead of vertically. The beam was kept constant and draught 

 only was changed, as we went aft, the sections having 98 per cent coefficient. In 

 other words, stern D was an exaggeration of the flat-stern type. Stern E was a 

 combination of sterns C and D, the attempt being made to combine them half and 

 half as it were, and coax the water up along a bilge diagonal. In stern F not only 

 was the coefficient .98 retained for all sections, but they were made geometrically 

 similar to the midship section, beam and draught both being decreased as necessary 

 to obtain the required area. For sterns C, D, and F the sectional coefficient is con- 

 stant, .98, as already stated. 



Fig. 9, Plate 17, shows how this constant sectional coefficient contrasts with the 

 curves of sectional coefficient of A, B, and E. It is typical of V-sections that the 

 sectional coefficients are below .5, but when we consider that in the conventional 

 type the after sections include actual or virtual dead wood, it is evident that the 

 numerical value of a sectional coefficient aft has small significance as indicating the 

 nature of the form with which the water has to deal. 



The outstanding result of the experiments was the good showing of stern F. 

 Of course this was partly due to small wetted surface, but stern F shows up very 

 well in the curves of residuary resistance, indicating that even with a conventional 

 dead wood added it would still be somewhat the best. 



Stern D, the exaggerated flat stern, undoubtedly makes the worst showing in 

 Figs. 7 and 8, Plates 15 and 16, which is at first sight surprising in view of the fact 

 that the wide, flat stern is used with success on many fast boats. But the wide, 

 flat stern as used is usually of type somewhere between D and F, and Figs. 7 and 

 8, Plates 15 and 16, do not go up to torpedo-boat speed. For 20-foot models torpedo- 

 craft speeds correspond to 8 knots or more, and it is quite possible that at 8 knots 

 stern D would show the least resistance. The models used were not of high-speed 

 type, having four or five times as much displacement for a given length as a high- 

 speed torpedo-boat model, and hence were not tested to high speeds. It is interest- 

 ing to note that the excessive resistance of stern D at speeds a little below five 

 knots is really due to a characteristic favorable to high speed. The virtual or ef- 



