THE PROBLEM OF THE HULL AND ITS SCREW PROPELLER. 
By Rear Apmirat C. W. Dyson, U. S. Navy, MEMBER. 
[Read at the twenty-eighth general meeting of the Society of Naval Architects and Marine Engineers, held in New 
York, November 11 and 12, 1920.] 
1. Introduction.—In 1915 Sir Archibald Denny made the statement that ‘in 
the future the rules for the correct designing of propellers should be derived from 
data carefully taken from the trials of smooth bottom vessels carefully run over 
accurately measured deep water courses.” 
Several years ago Admiral Taylor wrote to the effect that if theoretical for- 
mulas for design are adhered to, it would be necessary to compare each formula 
with experimental results and select that one which seemed to agree more closely. 
Then, using this as a semiempirical formula, with constants deduced from experi- 
ments or experience, problems could be satisfactorily dealt with. 
Both of these statements practically agree. Denny boldly casts all theo- 
retical treatment of the propeller into the discard, while Taylor, by implication, 
takes the same action, only retaining enough of the theory to save its face, to use 
a Chinese expression. 
To accept these statements of such eminent authorities at their full value 
throws us at once into the field of empirical results from which to derive data and 
formulas for the determination of hull forms and characteristics of propellers to 
fit any particular form of hull and the resistance which it may be desired to 
overcome. 
2. Difficulties Encountered in Obtaining Reliable Data——Upon entering the 
empirical field for propeller data, the investigator is at first encouraged by the 
large, apparently large, amount of data available, but it does not require any 
extravagant amount of time for him to ascertain that fully 95 per cent of the 
available data is absolutely worthless, this statement of course referring to condi- 
tions where general formulas and constants for design covering the entire field of 
the combined problem, ships and their propellers, are being sought. 
The principal causes of worthless data are:— 
(a) Lack of model-tank trial curves for the ship. 
(6) Variations in conditions of ship’s bottom. * 
(c) Deep sea trials in place of measured mile. 
(d) Variation in displacement and trim from those of model. 
(e) Measured mile trials run in a tideway with insufficient number of runs 
for each spot. 
(f) Trials run in shallow water. 
(g) Model hull, when tried, not fitted with external appendages. 
