IN A SINGLE-SCREW CARGO SHIP. 165 



speeds. The range covered corresponded to speeds of 5 to 12^ knots for the ship. 

 About forty different runs were made with each model, giving a corresponding num- 

 ber of points for plotting the torque, thrust, and revolutions per minute curves. 



The armature of the propeller dynamometer was especially designed to reduce 

 to a minimum the amount of magnetic thrust. This thrust increased with the torque 

 and amounted to 0.17 pound when the armature was displaced 7/16 inch and the 

 torque delivered to the shaft was 16 pounds-inches. Neglecting this at higher 

 powers would have caused an error in thrust measurements of about 1.4 per cent, 

 but would not have caused any error in the power measurements. However, this 

 magnetic thrust was separately calibrated, and corrections for it were made in work- 

 ing up the results of the experiments. 



It is interesting to note that experiments of this kind with a very heavy model, 

 in this case a displacement of 3,774 pounds, afford a very sensitive means of check- 

 ing the uniformity of speed of the towing carriage. If the towing carriage itself 

 were not very carefully adjusted, so as to eliminate slight variations in power re- 

 quired to drive it, owing to variations in the level of the tracks or uneven friction of 

 the driving wheels or g^ide wheels, the resulting small accelerations and retarda- 

 tions of the carriage were made very apparent. As the model, driven by its own 

 propeller, when supplied with a constant voltage, ran at very uniform speeds, it was 

 possible to see the towing carriage gaining or losing distance of a few inches in five 

 or ten seconds, as slight inequalities of speed due to small variations in the resist- 

 ance of the driving mechanism developed. In order to conduct self-propulsion ex- 

 periments successfully with the method followed, it was necessary to have the tow- 

 ing carriage running in excellent condition. 



Immediately after completion of the self -propulsion tests on the model the pro- 

 peller was removed and the runs to obtain the shaft friction and the thrust without 

 propeller were repeated. The model was next connected with the resistance dyna- 

 mometer on the towing carriage and the usual model resistance data taken. This in- 

 sured that the conditions of test both for self-propulsion and for the resistance of 

 the model would be uniform as regards conditions of the model, temperature of 

 water, etc. Experiments on the four models were run on succeeding days, and the 

 springs of the p^ropeller dynamometer were calibrated both before and after the 

 tests. 



In Fig. 3, Plate 98, is shown one of the dynamometer cards taken for a model 

 running at a speed corresponding to about 10% knots for the ship. It will be noted 

 that the deflection of the torque spring for this condition was 2.47 inches, for the 

 thrust spring 2.24 inches, and the interval over which the revolutions were measured 

 was 1.78 inches. With a little practice it was possible to read both the thrust and 

 the torque deflections within i/ioo of an inch. The revolutions could be read to an 

 accuracy of less than one-half of one per cent. Considering that the data taken 

 are plotted and the whole averaged by means of faired curves, it is estimated that 

 the results obtained as regards the three elements measured are correct within one 

 per cent. 



