Modern Development.— Steam-turbines are heat engines, converting the calorific energy of the steam into directly-available mechanical en ergy. From another point of view they are analogous to hydraulic turbines, and in design, construction and operation are controlled by the same ultimate principles. To-day the tur bine is more nearly a "perfect than any other known tFpe, so far as its design and construction are concerned. As in the case of hydraulic turbines the steam-turbine is di vided into classes; the two principal ones being action and reaction turbines; and each of these classes is subdivided accordingly as the turbine is composed of a single wheel or of several wheels, traversed successively by the steam in course of expansion. Among reaction turbines (compared in hydraulics to the Jonval turbine) of which, as multiple machines, the prototype is the Parsons turbine, the steam is only partially expanded in the distribution and acquires its full expansion in the movable wheel. The steam, therefore, acts on the blades at once by its pressure and its velocity. In the action tur bine, the steam only acts on the movable wheels by its velocity. Each wheel revolves in a casing in which the pressure is uniform. Action tur bines revolve at a less velocity than reaction turbines. There are now recognized three ways of utilizing steam for the transformation of heat into useful work: (1) By directing a jet of steam on all movable objects, such as the blades of fans, or, in reaction wheels, by allowing the steam passage itself to revolve; (2) by con densing steam in a cylinder and allowing the external atmospheric pressure to force a piston inward; (3) by cutting off slices of high-pres: sure steam and expanding in the cylinders of the modern reciprocating engine. The day steam-turbine designer, however, is merely concerned with the first and oldest method.
The Impulse Steam-Turbine.— The plest possible steam impulse turbine must con sist essentially of a fixed nozzle guiding a jet of steam upon vanes arranged at a periphery of a rotating wheel in a very similar way to 'lie buckets of the water-wheel; but to turn this wheel efficiently it is necessary not only to make the nozzle of suitable dimensions and shape for the work to be done, but to form the vanes or buckets of the wheel so that they make the best possible use of the steam as delivered from the nozzle. The function of the nozzle is to convert the whole available energy of the steam into mass velocity in the required direction. It should, therefore, be designed so as to expand the steam to the same pressure as that of the wheel chamber before the delivery on the vanes. If the nozzle is improperly formed, either the expansion will not be complete and the jet will burst into a cloud at the orifice, or eddy currents there will be retardation of flow due to the skin friction, and if C is too short the steam will not have time to expand completely. The area at A, then, depends on the quantity of steam to be passed; the area, B. on final pressure required, and the length, C, on the velocity of flow due to the difference between initial and final pres sures. The best shape for wheel vane faces is undoubtedly as nearly a semi-circle as is per mitted by the angle of the nozzle in Fig. 3. This angle is usually about 20 degrees, which is nearly the maximum possible. While being of sufficient area to reverse the direction of current without shock or eddy, these vanes will be set up within the nozzle itself and re tardation of flow will be the result. To arrive
at the best shape of steam nozzle, the ratio of expansion required should be first decided and the area of entrance — A in Fig. 1— should bear a relation to the orifice, B, in direct propor tion to the increased volume of steam when ex panded. The curve of nozzle walls is calculated to allow gradual expansion, until maximum ve locity, due to initial pressure energy, is attained. Fig. 1 shows an example of what is found to take place in an incorrectly shaped nozzle. The steam particles rebound from the walls as shown by the dotted lines, and meeting at a a a a form points of greatest pressure. Turning again to Fig. 2, if the length, C, is too great, should be short, so as to minimize skin friction, and the edges should be sharp to avoid blanket ing the nozzle. The most efficient mean speed of vanes is just under half the velocity of the steam current, thus leaving the exhausted mole cules relatively motionless. For this reason single disc turbines must revolve at enormous speed and be geared down for driving ordinary machinery. For example, a 10-inch De Laval turbine working at 70 pounds' pressure revolves 14,000 times per minute.
The Multiple-Step Plan.— Geared motors are avoided by most engineers, and so the mul tiple-step steam-turbine was devised as a means for reducing the speed of rotation and driving machinery direct. This idea was first developed by C. A. Parsons during 1884-87. The earliest Parsons parallel flow turbine was a collection of zigzag nozzles whose walls were formed by pro jecting rings of blades intermeshing and so ar ranged that the azigsp were fixed to the inner circumference of a stationary hollow cylinder and the uzagsa to the outer circumference of a rotable cylinder. The modern Parsons turbines from which such notably economical steaming results have been obtained differ from the ear liest — apart from improvements in mechanical and constructional detail—only in the alteration of the contour of the steam passages or nozzles from a zigzag to a sinuous shape. Instead of sharp corners the tam ion flowing curves and the cross-sectional area of steam channels increases in nearly exact tion to the growing volume of steam as it to a certain extent permitting correct mainte nance of expansion ratios for varying steam pressures and loads. Professor Curtis argues that his arrangement of a• few large nozzles and wheels in series lessens the percentage of waste through spilling of steam. In order to reduce the revolutions per minute as far as possible, Mr. Parsons appears to advocate a large num ber of turbines in series of small diameter, while Professor Curtis believes in few wheels of large diameter. Prof. A. Bateau of Paris has worked out a multiple step impulse steam turbine, consisting of a number of Pelton or Laval wheels arranged in series on a shaft, each \\ hccl fig in a ;-.Lra,atc r tributing nozzles convey the steam to the wheel vanes, and the whole of the expansion is ac complished within the fixed nozzles. The steam expands to the exhaust pressure, the collective areas at each stage fulfilling more closely the conditions for a perfect nozzle, as previously mentioned. But as expansion takes place both within fixed and moving blades, as the whole passage is one continuous nozzle, the clearances and workmanship must be of the finest to min imize leakage, which in the earlier machines caused serious trouble.