The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

448 ing frame with involute teeth is unique. From the foregoing it will be seen that the elasticity of the I-beams al- lows freedom of movement of the axis of the pinion and floating frame in the vertical plane. But by warp- ing the webs of these beams the axis of the floating frame might also turn slightly in the horizontal plane. It is, of course, many times more rigid- ly mounted' against this displacement than against that inthe vertical plane, but still the rigidity. is by no means complete. Now. it« may appear that this is not specially from curve A B, Fig. 6, it would seem that such an angular displacement in the horizontal plane, even if of meas- urable amount, causes no important disturbance. But that is not-the only consideration, asthe question then arose as to whether with freedom of 'zontal planes the position of the axis of the floating frame would be stable or unstable; that is to say, if forcibly displaced from exact alignment to a will the forces at the tooth contacts tend to bring it back to perfect align- ment or to further displace it? By consideration of the slight changes of direction of the forces at the tooth contacts produced by such displace- ment, it can readily be shown that, with involute teeth, the position of the axis of the floating frame is un- stable, and a slight displacement aris- ing from any cause will tend to in- crease. With the I-beams alone, there- fore, there might arise excessive hort- zontal errors of alignment, and trif- ling causes might produce sudden and violent changes of the horizontal er- ror, either of which effects would en- danger the whole mechanism. The floating frame has therefore 'been de- prived of all freedom of motion in the horizontal plane by means of two horizontal struts standing transversely between the floating frame and bed- plate. These bear on the floating frame at D and Dt, Fig. 2, and one is clearly shown in Fig. 1. Obviously these struts in no way interfere with the freedom of movement of the floating frame in the vertical plane. Where they bear in the bedplate there is an adjusting and locking mechanism which greatly facilitates the true set- ting up of the floating frame, and is to be seen in the perspective view in the front of the bedplate, near the end next the turbine. ; With epicycloidal teeth in perfect adjustment, the position of the float- ing frame axis can be shown to be important, as rotation both in the vertical and hori- small angle in the. horizontal plane, -and between the teeth Tae Marine RevIEW stable without struts. But this is only an apparent advantage, as this stability would at once be upset by widening the gear and pinion centers, or by the slightest wear. But even if this stability could be trusted, it is believed that epicycloidal teeth are far too sensitive to errors of adjust- ment to be used, and the question of stability or instability in no way af-. fects this consideration. Provision for OSnious Lubrication. The only condition which "cause serious disturbance of the tooth pressures is-an excessive heating of the pinion above the temperature of the gear. But the design provides for a copious application of. lubricating and cooling oil, especially to .the pinion, which has most tendency to heat. The cover, also,.is.so arranged as to draw in air at the ends and at A, Fig. 4, by the fan action of the gears, and discharge it by openings (not shown) to the cross-section. Besides, water can be cir- culated, by means only partly shown in the figures, between the pinion and the flexible shaft S. But it is not anticipated that the heat to 'be. re- moved per minute will be so. great as has been provided for. The gears of heavy electric trains work under much higher stress both in the metal in contact-- very much higher, as these stresses are accentuated 'by errors of align- ment shown by rounding of the teeth through wear--and with practically no lubrication. Yet they run for long 'distances. In the reduction-gear these unfavorable conditions are reversed-- there is good lubrication and very uniform contact. If the coefficient of friction at the tooth contacts is 1/10, it. can readily be shown that the frictional loss at the teeth will be under 1 per cent... So. that even if the gear were transmitting 6,000 horsepower--the very highest power hoped for--this' frictional loss would be under 60 horsepower. As at least one-half will go to the large-gear, where it will readily be dissipated from the large surface, there is left well under 30 horsepower to remove from the pinion, which, with the various means, provided, should pre- sent no difficulty. If then this con- trol of the heat is assured, there will, it is claimed, 'be prectically per- fect tooth contacts; this, with the fine pitch and the spiral gears (avoiding sudden entering or leaving contact of one whole tooth at a time), will, it is 'asserted, insure reasonably quiet run- ning of the gears. pressure of 350 Ib. could: per. minute. multiplied _ "by _ preéssir -meafs' of : 'comparing: these ro cases, right of this: November, 1909 At 1,500 revolutions per minute of pinion and 453 Ib. per in. of tooth contact, the gear, will transmit 6,000 horsepower. The pitch-line speed would then be 5,500 ft. per minute, and the mean speed of sliding about one-tenth this, or 550 ft. per minute. There is on record a case of a right- angled worm-gear run successfully at 15,000 ft. per minute with a contact per, inch. The average speed of sliding will here be greater than the pitch-line 'speed by over 41 per cent, or, say, 21,000 ft: Taking. speed of sliding : "rough We Cel. Reduction gear 5 500% 453 1 Right-_ oo = 'angled worm gear, "2h 000 x 350: 29.5 The extreme constants chosen, therefore, Seem well within practical limits. Before the experiment is tried it is impossible to predict the limit of safe load, but- the following are two cases which it 'will ibe interesting to con- sider :--- 1. At 1,000 revolutions per minute of---pinion,~-and 283 Ib.» per -ins--of tooth contact--that is, as shown above- only 28 per cent of the intensity of pressure used with steel gears of 1% in. pitch at ordinary pitch-line speeds --the gear would transmit 2;500 horse- power. The famous merchant ship Aberdeen, .\by the engines of which Dr. Alexander C. Kirk introduced triple-expansion, was about of this power. Her main engines, without shafting, propeller, stern-tube, spare gear and fittings, weighed 221 tons. If we deduct from this the condensers, pipes, sea-valves, air-pumps, circulating pumps, donkey pumps, floors and gratings, etc., the weight left for the main engines proper will be about 150 to 160 tons--say, .150 tons. Hence the following comparison may be made :-- Westinghouse marine tur- bine of 3,000 brake horse-power, 1,000 revo- lutions per minute Wels: 36 27 Tons Reduction gear 25 a Tote, "so. 52 Aberdeen's engines' re- Pieced -o cee 150 a Rovio ee: 98 2 That is, 65 per cent of the weight is saved. If this design will trans- mit 2,500 horsepower, it is claimed that similar gears could 'be applied