RUDDER TRIALS, U. S. S. STERETT. 329 



a turn, when the helm was put over to the desired angle, at a fairly rapid 

 rate (3° per second), the vibration did not increase materially up to the time 

 when the dynamometer showed a maximum stress (approximately 1 2 seconds 

 for 35° helm). Therefore it was considered that for the condition of maxi- 

 mum stress, the whole force of friction, as shown on curve I, was effective 

 in opposing the movement of the rudder. Accordingly the reading of the 

 dynamometer (maximum stress) was diminished by an amount taken from 

 curve I (Fig. 17, Plate 127) for the given maximum helm-angle — 



^T=D-{iR+B). 



On the other hand, when the helm was steady a-starboard and the ship 

 had turned through approximately 90°, with the dynamometer indicating 

 a practically uniform stress, there was a marked increase in vibration of 

 the ship's stern, possibly due to the impact of masses of water thrown 

 successively against the rudder by the port propeller. It is certain that 

 this violent vibration would materially diminish any static frictional forces 

 in rudder bearings. Therefore it is believed that on steady helm, the 

 correction for rudder friction (fi?) amounting to -f-2 per cent (^Wir) of the 

 maximum stress obtained, should be neglected. Furthermore, the static 

 force in lead B would be reduced by continued vibration, and could not 

 again reach its full value without further motion of the helm to starboard. 

 The force in B could not act at steady helm, with the full value shown on 

 curve II (Fig. 17, Plate 127). Accordingly, since we have no means of 

 approximating the value of B in this condition of violent vibration on steady 

 helm, and since neglecting this force entirely is to err on the safe side, over- 

 estimate, 



D = ^T+B-hR, 



no frictional correction has been applied to the "steady" stresses obtained. 

 The dynamic friction of the leads on 35° helm, as measured afloat, but 

 without vibration, is 1,000 pounds. Assuming even as much as half of 

 this to act under violent and continued vibration, the error introduced 

 would be ^oVt, which corresponds to a correction of —6 per cent. This 

 correction, combined with the correction (+2 per cent) for rudder friction, 

 gives a maximum net correction of —4 per cent which has been neglected. 

 The error is probably not so great as this, and is on the safe side. 



On the backing tests, the vibration was so violent as to shake down 

 the boiler brick walls. Consequently it is believed that no value could be 

 found in an analysis of friction forces under the backing condition. 



