March i, 1900J 



NA TURE 



425 



potential energy of the high pressure steam into kinetic energy 

 of velocity in the direction of flow. 



Secondly, the crude paddle-wheel of Bianca is replaced by a 

 wheel of the strongest steel, fringed round the periphery with 

 little cupped blades of steel, somewhat analogous to the buckets 

 of a Pelton water-wheel. 



Lastly, the steel wheel is mounted on a long and somewhat 

 elastic shaft, to allow of its easy and free motion, and on one 

 extremity of this shaft is mounted the pinion of the spiral 

 reduction gear. 



The speeds of revolution of the steam wheels of De Laval's 

 turbine are from 10,000 to 30,000 revolutions per minute, 

 according to the size, involving peripheral speeds up to 1200 

 feet per second, or about one-half the speed of the projectile 

 from a modern cannon. Such speeds are necessary to obtain 

 power economically from the high-pressure steam jet, issuing 

 at from 3000 to 5000 feet per second, as calculated by Rankine. 

 It is somewhat remarkable that not till a century after Bianca, 

 the piston or ordinary reciprocating engine made its first 

 ppearance, in about the year 1705, and has since become one 

 t the chief factors in the great mechanical and engineering \ 

 !4rowths of the last century. During this period the steam 

 turbine seems to have been, practically speaking, neglected, 

 which is somewhat remarkable in view of the numerous attempts 

 of inventors to construct a rotary engine, attempts which had 

 no practical results. 



In the year 1884, the advent of the dynamo-electric machine, 1 

 and development of mechanical and electrical engineering, 

 created an increased demand for a good high-speed engine. 1 

 Engineers were becoming more accustomed to high si>eeds of 

 revolution, for the speed of dynamos was at this time from 1000 

 to 2000 revolutions per minute, of centrifugal pumps from 300 

 to 1500, and wood-working machinery from 3000 to 5000; and 

 Sir Charles Wheatstone had made a tiny mirror revolve at a I 

 speed of 50,000 revolutions per minute for apparatus for j 

 measuring the velocity of light. The problem then presented 1 

 itself of constructing a steam turbine, or ideal rotary engine, 

 capable of working with good economy of steam at a moderate 

 speed of revolution, and suitable for driving dynamos without ' 

 the intervention of reduction gearing. To facilitate the problem, I 

 the dynamo was also considered with the view of raising its | 

 speed of revolution to the level of the lowest permissible speed 

 of the turbine engine. In other words, to secure a successful 

 combination, the turbine had to be made to run as slowly as 

 possible, and the dynamo speed had to be raised as much as 

 possible, and up to the same speed as the turbine, to permit of 

 direct coupling. 



In 1884 preliminary experiments were commenced at Gates- 

 head-on-Tyne, with the view of ascertaining by actual trial, the 

 conditions of working equilibrium and steady motion of shafts 

 and bearings at the very high speeds of rotation that appeared 

 to be essential to the construction of an economical steam 

 turbine of moderate size. Trial shafts were run in bearings of 

 different descriptions up to speeds of 40,000 revolutions per ■ 

 minute ; these shafts were about i^ inches in diameter and 2 

 feet long, the bearinsjs being about % inch in diameter. No 

 difficulty was experienced in attaining this immense speed, 

 provided that the bearings were designed to ha%e a certain small 

 amount of "give" or elasticity ; and after the trial of many 

 devices to secure these conditions, it was found that elasticity, i 

 combined with frictional resistance to transverse motion of the I 

 bearing bush, gave the best results, and tended to damp out I 

 vibrations in the revolving spindle. This result was achieved ! 

 by a simple arrangement ; the bearing in which the shaft re- ' 

 volved was a plain gun-metal bush with a collar at one end and ; 

 a nut at the other ; on this bush were threaded thin washers, ! 

 each being alternately larger and smaller than its neighbour, 

 the small series fitting the bush and the larger series fitting the 

 hole in the bearing block, these washers occupying the greater 

 part of the length of the bush. Lastly, a wide washer fitted ! 

 both the bush and block, forming a fulcrum on which the bush I 

 rested ; while a spiral spring between the washers and the nut 

 on the bush pressed all the washers tightly against their neigh- 

 bours. It will be seen now that, should the rotating shaft be 

 slightly out of truth (which it is impossible to avoid in practice), 

 the effect is to cause a slight lateral displacement of the bear- 

 ing bush, which is resisted by the mutual sliding friction of each 

 washer against its neighbour. The shaft itself being slightly 

 elastic, tends to centre itself upon the fulcrum washer before 

 mentioned, under the gyrostatic forces brought into play by 



NO. 1583, VOL. 6l] 



the rapid revolutions of the shaft ahd influenced by the fric- 

 tional resistance of the washers, and so the shaft tends to assume 

 a steady state of revolution about its principal axis, or the axis 

 of the mass, without wobbling or vibration. This form of bearing 

 was exclusively u.sed for some years in turbine engines aggregating, 

 some thousands of horse power, but it has since been replaced 

 by a simpler form fulfilling the same functions. In this later 

 form the gun-metal bush is surrounded by several concentric 

 tubes fitting easily within each other with a very slight lateral 

 play ; in the interstices between the tubes the oil enters, and its 

 great viscosity when spread into thin films has the result of 

 producing great frictional resistance to a rapid lateral displace- 

 ment of the bearing bush ; the oil film has also a centring action, 

 and tends under vibration to assume a uniformity of thickness 

 around the axis, thus centring the shaft, and like a cushion- 

 damping out vibrations arising from errors of balance. This 

 form of bearing has been found to be very durable and quite 

 satisfactory under all conditions. 



Having tested the bearings up to speeds above those con- 

 templated in the steam turbine, the next problem was the 

 turbine itself. The laws regulating the flow of steam being 

 well known (which was not the case in Hero's time), various 

 forms of steam turbine were considered, and it appeared de- 

 sirable to adopt in principle some type that had been both 

 successful in the water turbine, and also easily adapted to % 

 multiple or compound formation, a construction in which the 

 steam should pass successively through a series of turbines one 

 after the other. 



Fixed Bl«oes. 



Mo 



rir«, Bladc'j. K 



^ 



lYlov 



Fl aED BlbdES 



n(^&LHOEe>. 



Fig. I. — Fixed and moving blades of turbine. 



The three best known of water turbines are the outward ffow, 

 the inward flow, and the parallel flow, and of these the latter 

 appeared to be the best adapted for the niultiple of compound 

 steam turbine, for reasons which will afterwards appear. 



The object in view being to obtain a good coefficient of 

 efficiency from the steam with a moderate speed of revolution 

 and diameter of turbine wheel, it becomes essential that the 

 steam shall be caused to pass through a large number of suc- 

 cessive turbines, with a small difference of pressure urging it 

 through each individual turbine of the set, .so that the velocity 

 of flow of the steam may have the proper relation to the 

 peripheral velocity of the turbine blades to secure the highest 

 degree of efliciency from the steam, conditions analogous to 

 those necessary for high efficiency in water turbines. A large 

 diameter of turbine wheel, it is true, would secure a moderate 

 speed of revolution, but this may be dismissed at once for the 

 simple reason that the frictional resistance of such a disc re- 

 volving at the immense peripheral velocity, in the exhaust steam, 

 would make it a most inefficient engine. 



In the year 1884, a compound steam turbine engine of lo horse- 

 power and a modified high speed dynamo were designed and 

 built for a working speed of 18,000 revolutions per minute. 

 This machine proved to be practically successful, and sub- 

 sequently ran for some years doing useful work, and is now ii> 

 the South Kensington Museum. 



This turbine engine consisted of two groups of fifteen suc- 

 cessive turbine wheels, or rows of blades, on one drum or shaft 

 within a concentric case on the right and left of the steam inlet, 

 the moving blades or vanes being in circumferential rows pro- 

 jecting outwardly from the shaft, and nearly touching the case. 



