LOGARITHMIC SPEED-POWER DIAGRAM. 339 



Peabody's paper "Experiments on the Froude") were run at 5 knots speed 

 of advance. To construct Fig. 8, Plate 136, the curves of propeller thrust 

 were reduced to effective horse-power, and for convenience, slip (s) was 

 reduced to the coefficient K, which is 101.3-^(1—5). With the sheet super- 

 posed upon Fig. I, Plate 129, the points are marked at 5 knots speed corre- 

 sponding to the respective values of K and of efficiency. Through the former 

 points diagonal lines are drawn parallel to the line on Fig. i, Plate 129, 

 marked F^ and each line marked with its respective value of K. The scale 

 for efficiency is now constructed by projection from the points for efficiency 

 which have been marked at 5 knots, in a direction parallel to the lines for K. 

 Now, since experiment has shown that, within limits, a given propeller has 

 constant efficiency at constant slip, and that its power varies as the cube of 

 speed of advance, we may take any speed of advance for this propeller, and 

 any effective power, and read the value of K, and, by diagonal projection, 

 the value of efficiency. From K, speed of advance, V, and pitch, P, the 

 revolutions per minute, R, are ascertained by the formula — ■ 



R=^XK. 



In the case of design, if it is found that the revolutions per minute so 

 found are too high, the diagram, Fig. 8, Plate 136, may be shifted upward on 

 Fig. I, Plate 129, giving a propeller of larger diameter, and for the desired 

 speed of advance and power the efficiency and revolutions are again deter- 

 mined, and the process repeated until a desired speed of revolution is obtained. 

 In this process, of course, the shifting line of Fig. 8, Plate 136, is kept super- 

 posed on line V on Fig. i, Plate 129. In case of designs where the revolu- 

 tions per minute can be made to suit the propeller, it is only necessary to 

 shift until the line of best efficiency passes through the required speed of 

 advance and effective power. 



A sheet such as Fig. 8, Plate 136, represents the performance of only one 

 series of propellers, all geometrically similar. If the pitch ratio or the width 

 ratio is changed it is necessary to construct a new sheet. It is not difficult, 

 however, to conceive an arrangement by which a series of types may be 

 shown on the same sheet. 



In the application of this propeller data to the ship certain precautions 

 are necessary, as in all methods of propeller calculations, three of which will 

 be mentioned. 



(a) The speed of advance of the propeller through the surrounding water 

 is not equal to that of the ship, due to the wake. This can only be estimated 

 by comparison with other cases, or determined by model tank experiment. 



