SOME GRAPHIC STUDIES OF THE ACTIVE GYRO STABILIZER. 185 



The scale for the records given was obtained by forcibly tilting the pendulum 

 definite angles and holding the same momentarily with the recorder operating. 



It will be noted in the formula previously given that the impulsive perform- 

 ance on the part of the gyro is assumed as datum or lOO per cent. This being an 

 engineering impossibility from a practical standpoint, it has been suggested to 

 adopt a more practical unit, consisting of the efficiency obtained by the constant 

 precessional velocity of two times the phase velocity of the ship, for the reason 

 that this phase velocity is so slow, especially for large ships, that it may be at- 

 tained almost instantaneously. This, therefore, will be used herein as the basis on 

 which the efficiency obtained in these experiments will be calculated. 



Amongst others, the following characteristic curves were produced : 



1. Extinction Curves, made by tilting the pendulum to 20 degrees and releas- 

 ing it with one or two gyros in action (see Figs. 2 and 1 1, Plates 1 13 and 1 14). The 

 quenched increments were for a = 2°, ^ = 2.69°/ for a = 3°, 5' = 4. 36°, when deducting 

 the small corresponding friction increments, taken from the natural extinction 

 curve, this indicates an actual efficiency of .94 for a = 2°, and .96 for a = 3"; 

 when based on the efficiency described above. 



2. Rolling Curves (Figs. 3 and 12, Plates 113 and 114). These were made 

 by reversing the control contacts, tilting the pendulum about one degree to insure 

 an initial movement, and then releasing it. The roll increments produced when 

 adding the corresponding friction increments were 2.75°, 4.7°, and 2.25°. The ef- 

 ficiency indicated was .965, .99 and .95, on the same basis. 



3. Curves showing the performance of the stabilizer when the pendulum was 

 subjected to a succession of uniform waves. For these experiments the weight 

 producing the wave moment was adjusted until the maximum moment produced 

 an inclination of a° corresponding to the maximum moment of a synchronous 

 wave of a slope. Figs. 4-23, Plates 113 and 114, show the effect of one gyro and 

 of two gyros against 2 degrees and 3 degrees wave slope, respectively. In curves. 

 Figs. 6, 7, 18 and 19, the stabilizer was stopped, and, after several synchronous in- 

 crements, thrown in again to show that the stabilizer actually worked close to its 

 capacity as the resulting extinction curve nearly coincides with the natural fric- 

 tion extinction curve, which emphasizes the close correspondence between the fig- 

 ured and actual roll quenching power, and also the striking fact that close stabil- 

 ization can be relied upon even when the figured roll quenching power is actually 

 less than the wave increments. 



Since the mass moment and frictional moment of the stabilizer are entirely 

 eliminated by active control, the closeness of the stabilization obtained will depend 

 only on the sensitiveness of the control, which, for the definite speed of the control 



gyro, is proportionate to-=, where T is not the period of the vessel or pendulum, 



but of the wave. This is due to the fact that a stabilized vessel or pendulum will 

 no longer roll in its own phase, but will, almost immediately after the stabilizer is 

 thrown in, fall into forced roll and adopt the wave period. Closer stabilization 



