296 ELECTRIC PROPULSION OF MERCHANT SHIPS. 



The pitting is clearly a case of fatigue of metal from repeated overstrain, and, furthermore, 

 experience witli ships has shown that this pitting occurs periodically, and every time you 

 take a pit out of the gear you reduce the effective face of the tooth. Thus the more a gear 

 is pitted the more it is subject to additional pitting. 



My statement that gears of all types have shown deterioration in cargo ships has been, 

 criticised. That statement is absolutely justified, and my personal observation justifies it; 

 that is, I have seen gears of the very lowest degrees of pressures in which the faces were 

 badly pitted, and subsequently received reports from these ships that the pitting had become 

 very much worse. There was one notable case of a ship in which the lowest pressures were 

 used of any gears produced in this coimtry, and we had a report from one of our engineers 

 to the effect that the ship had been run 40,000 miles, and the original surfaces of the low- 

 speed pinions were pretty nearly obliterated; that is, they were badly pitted all over, and 

 will have to be replaced at an early date. You are concentrating force on the reduced area 

 when the gear is pitted. The point I want to make is this, that this thing is in the nature 

 of strains which come occasionally through the effects of the sea, and the ship runs along 

 well for a long time, and then suddenly there is some condition where the gears get in- 

 jured. I claim, if a 100 per cent performance of gears is to be expected, that these gears 

 must be so proportioned that they do not pit at all. If they pit at all, under any condition, 

 it means in time they will go down, and if they do go down it is a serious matter. The 

 ship will be laid up to renew them, it will cost a great deal of money, and it would very 

 soon eliminate any advantage of the gear if you had to make many renewals. They should 

 virtuall}^ last forever. 



The speed in reversing was mentioned. Now the test referred to in this paper was at 

 the full speed of the turbine, but the test I mention, where we got 500° F. instantaneous rise 

 by reversal, was made when we reversed at the normal reversing speed, running the tur- 

 bine at such speed as is produced in service by the reversing of the turbine. We got 500° 

 rise of temperature at certain points in the main turbine. 



Mr. Reid's criticism of this design is a true one. This drawing is not correct and 

 shows a bad arrangement. It was done in a hurry, and there were certain things I ob- 

 jected to, when the drawing was made, but we did not have time to change it. I do not 

 like the arrangement of the motor there. There are four bearings too close together, and 

 they might tend to jam against each other. The motor should be a little further forward 

 and a little more room for flexibility allowed. His criticism, however, of the absence of 

 thrust is not correct, because that little wart on the forward end of the motor is supposed to 

 be a thrust bearing — I think that it is not quite a big enough wart, that it should have been 

 made bigger to make room for the thrust bearing. 



The question of putting the turbine above is absolutely beyond criticism. The par- 

 ticular design of turbine shown in that paper could be put up on 2 by 4 scantlings if they 

 would hold the weight. The turbine would run perfectly. We run even our larger tur- 

 bines on structural iron supports with perfect success, so you can put the turbine where you 

 please, put it in the deck-house or off to one side alongside the boilers, or anywhere where 

 you have room for it. I think it would be rather convenient to place it as we have shown 

 it here — the condenser in athwart ship' position, and then the drainage from the condenser 

 is convenient, the delivery of steam from the condenser is convenient, and there is an en- 

 tire absence of strain to distort the turbine. With the condenser so supported, I am sure 

 the apparatus installed as shown will work very well. 



