182 SOME GRAPHIC STUDIES OF THE ACTIVE GYRO STABILIZER. 



axis of both gyro rotors was horizontal and at right angles to the rolling axis 

 of the pendulum. The possible precession angles were limited to 60 degrees from 

 their center position by automatic switches opening the clutch circuit just before 

 the angles mentioned were obtained and mechanical stops, set for these limits, 

 provided a positive limit. 



The motor speed, and thereby the precessional velocity of the gyros, could be 

 varied. The value of one of the fundamental features of the active gyro, that of 

 establishing an absolute upper limit of the precessional velocity and therefore of the 

 stresses in the plant and also in the ship in the vicinity of the plant, was recognized 

 throughout the tests. The rotors of the stabilizer, as well as that of the control 

 gyro, were run at any desired speed by three-phase current obtained from a mo- 

 tor generator, the speed of which was under control. The automatic control was 

 of the type responsive to angular velocities and was represented by a third gyro 

 with its spinning axis also horizontal, athwartships, and able to precess about one 

 degree only to each side of the center position, in which it was held by adjustable 

 centralizing springs. An angular roll would first overcome the pressure of one 

 of these springs and then compel the slight precessional motion, which closed one 

 of the circuits of the magnetic clutch, causing precession of the main gyros in such 

 a direction that the stabilizing moments result in quenched roll in the direction 

 which closed the contact. The contacts can be positively closed by the angular ve- 

 locity of a ^-degree roll of 2.9 seconds period, thus insuring precession almost co- 

 incidentally with the beginning of even an incipient roll for close stabilization. The 

 stabilizing gyros, which are strictly non-pendulous and always locked, are by this 

 principle of control never allowed to precess in the wrong direction, increasing the 

 existing roll, nor allowed to precess too far, which would cause a reversal of roll 

 in its phase. Another important point is that such a stabilizer remains unaffected 

 by the positive couples set up by the gyros themselves, due to angular velocity of 

 any unquenched roll. 



This simple control is found to work well for wave impulses within the ca- 

 pacity of the stabilizer, and close stabilization is certain and reliable. Should an 

 impulse occur that is larger than the capacity of the gyros, a slight unquenched 

 residual roll results, the centralizing springs being overcome at a definite veloc- 

 ity of roll. The larger the roll entered upon, the sooner this velocity is attained 

 after the start of any individual roll. For maximum efficiency of the stabilizing 

 moment of the gyros their precession should be so timed that they pass through 

 their central position at the moment the vessel has attained its maximum velocity 

 in the middle of its roll, and it is found that the efficiency of the plant may be in- 

 creased slightly by retarding the action of the controller at times when the momen- 

 tary angular roll of the vessel exceeds a predetermined amount. It is found that 

 this may be accomplished automatically by the single augmentation of the central- 

 izing spring pressures as by the addition of a simple pendulum. 



The problem of main journals on the stabilizing gyros has been found to solve 

 itself. The frequent and complete reversal of the journal pressures with every half 



