It is the aim of this article to present to the service an elementary treatment of the principles of the so-called impulse and reaction turbines, and in a later article the aim will be to point out the manner in which these principles are applied to marine turbines and turbine auxiliaries in use to-day, especially in our navy. No claim is made to originality in the manner of treatment. The paper is merely a presentation, in condensed form, of data and explanations found in Southern, Moyer, Cox, Stadola, Kent, Thomas and others.
The laws of steam state that the greater the range between the initial and final pressures of steam, in its expansion, the greater will be the work obtained from a given amount of steam; and that the rapidity of the flow of steam from a point of any pressure to a point of lower pressure is increased materially because of the property of expansion possessed by steam. Throughout the study of the principles of the turbine it will be found that it is this rush of steam from which all the work is obtained. The power of doing work which rushing steam possesses is known as its kinetic energy, the amount of which depends directly upon the square of the velocity of the steam.
Requirements of a Turbine.
The practical steam turbine, to be efficient, must be designed so that it will accomplish the following:
1. Transform the available heat energy into the maximum kinetic energy.
2. Absorb from the available kinetic energy the maximum power.
3. Rotate at a speed that will be economical and safe.
4. Combine simplicity of construction with reliability of action.
Classifications of Turbines.
All practical steam turbines are grouped into two general classes, known commercially as either impulse or reaction; but these terms are confusing, because of the fact that at least one turbine in use at present is operated by both impulse and reaction, several by impulse alone, but none by reaction alone. These terms will be retained to avoid further confusion, but will not be considered as literal terms descriptive of the principles involved.
The impulse turbine is sometimes known as the velocity turbine, and the reaction as the action-reaction or pressure turbine.
Simple and commonplace examples of the two types of turbines are found in the windmill, where the impulse of the wind upon the vanes causes rotation; and in the pin-wheel, where the reaction of the continual explosions of the powder acting tangentially causes rotation.
Professor Thomas gives one of the best definitions of impulse and reaction:
Impulse is the total dynamic pressure exerted by the steam or jet passing over a blade or bucket surface and experiencing a change in direction of flow. Reaction is to be understood as the pressure opposite in direction to that of flow resulting from and accompanying a change in velocity of the steam.
With the simple impulse turbine it is known (and will be proved later) that for maximum efficiency the circumferential or peripheral velocity of the rotor should be one-half the velocity of the rushing steam. Efficient expansion of steam from a pressure of 170 pounds absolute to 2 pounds absolute gives a velocity to the steam of above 4000 feet per second. It is therefore evident that some means must be adopted to utilize the energy by steps, rather than in one single expansion, if the turbine be intended for marine propulsion. The manner in which this is done will be shown later for each type of turbine.
The Single-Stage Impulse Turbine.
The DeLaval turbine is not used for marine propulsion, but it • is described here, as far as the action of the steam is concerned, for the reason that it is the best example of the impulse turbine, in its simplest form.
Simply stated, the steam is constrained by a nozzle to impinge upon buckets arranged radially upon the periphery of the wheel attached to the shaft of rotation.
Fig. 1.
Fig. 1 shows this single-stage impulse turbine, but for a better understanding of the internal action Fig. 2 will be used in the description.
(1) and (2) are two views of the turbine nozzle and wheel; (3) is a velocity and pressure diagram. At a the steam is of high pressure and low velocity and is about to enter the neck of the nozzle. After passing the neck of the nozzle the steam enters a passage of gradually increasing cross-sectional area, expanding uniformly to the limit established by the pressure of the exhaust, with the result that the velocity has increased to correspond to the increase of volume, and, if no losses have occurred, the heat energy originally possessed by the steam has now been transformed to kinetic enrgy in the form of steam rushing along the axis of the nozzle
While pursuing this line of direction the steam impinges upon the cupped blades upon the periphery of the moving wheel, exerting upon them an impulse due to change in direction of the jet in conforming to the curvature of the blade.
This energy being applied tangentially, as shown in Fig. 1, causes the rotation of the wheel about its axis. After passing through the single wheel, the steam has completed its function and is condensed or exhausted to the atmosphere, according to the installation.
. The action of the steam in its passage through the nozzle and blades will now be taken up in detail.
AJozzles.—The divergent or expanding nozzle is the result of many experiments and varies in the degree of divergence and the length according to the extent of expansion desired and the uses to which it is to be put. It is so designed, however, that the steam, undergoing expansion in its passage through the nozzle, will acquire a velocity as the result of a constantly increasing acceleration, and that, while buildingupthe velocity,the losses from friction, heat losses, current eddies, etc., will be reduced to a minimum.
Neglecting all losses, the velocity attained by the expansion of steam can be determined by the formula:
.----- V" — 2 gh,
3 c h
where V is the theoretical velocity in feet per second; £ = 32.2 (gravitational constant) ; and /i = heat drop, in foot-pounds, i. c..
total heat of initial steam minus total heat of exhaust steam (as found from the steam tables) multiplied by 778, the ratio of a foot-pound to a British Thermal Unit.
With a known velocity and a known angle of the nozzle with the plane of the turbine wheel, Fig. 3 shows a theoretical velocity- diagram. V is the absolute velocity and AB the absolute direction of the jet of steam issuing from the nozzle, and a is the angle it makes with the plane of the moving wheel DB.
u is the velocity and CB the direction of rotation of the turbine wheel drawn to the same scale as V. Having given AB, BC, and the included angle a, by completing the triangle, V, represents
Fig. 3.—Theoretical Velocity Diagram, Impulse Turbine.
*
the relative velocity and AC the relative direction of the entering steam when viewed from the moving wheel.
For the sake of simplicity, the relative entering angle /3 and the relative exit angle /?' will be assumed to be equal. As losses have been neglected, the steam will have the same relative exit velocity ( By laying off EF equal to u but in the opposite
direction (wheel and steam are moving in opposite directions), and completing the triangle CEF, CF will represent the absolute direction and V2 the absolute velocity of exit jet of steam. To differentiate: Absolute velocity is the velocity when regarded from a fixed point, and relative velocity is the velocity when regarded from a moving point, as in this case, the turbine wheel.
Fig. 4 shows the velocities V and Vas the absolute velocities of entering and exit jet of steam meeting at the blade B at angles a and /?, respectively. By resolving these velocities into their rectangular components, EB and BD will represent the components acting together toward moving the blade in the plane of rotation FB, and AE and CD will represent the components of the impulse acting against each other (as shown by the arrows) in producing thrust to the shaft containing the blades. In the actual DeLaval turbine the blades are so designed that the components AE and CD are equal, thus counteracting the thrust com-
/?
Fig. 4.—Thrust Diagram, Impulse Turbine.
ponent AE of the entering jet by the equal and opposite thrust component CD of the leaving jet.
The Reaction Turbine.
The Parsons turbine is an example of the impulse and reaction principles, combined with another characteristic feature, that of the gradual expansion of the steam in each of a long series of turbine elements. Contrary to the DeLaval, there are no nozzles in which the steam is expanded before acting upon the moving blades, but the first row of fixed blades acts much in the same way as the nozzles. The feature of the gradual expansion of the steam and the corresponding increase in velocity and volume necessitates a long series of moving blades to absorb the energy
with any degree of efficiency, and for each row of moving blades there is one row of fixed blades.
Bv this arrangement the steam velocity is always comparatively low.
The idea can be grasped more easily by following the description in Fig. 5. (1) shows an elevation in section, with alternating
rows of fixed blades or guides a and moving blades or vanes b, each pair of which is known as a stage; (2) shows a diagrammatic sketch of the course of the steam through the blades; and (3) shows a graphic steam velocity and pressure diagram. It is necessary to follow the description in the three parts simultaneously to obtain a clear understanding of the internal operation of the Parsons turbine.
At A the steam is of high pressure and low velocity. Upon entering the first row of guides a, the velocity is increased slightly, (3) a, and the direction of the flow changes to conform to the blades, (2) a. Upon issuing from the guides a the steam impinges upon the vanes b and exerts an impulse upon the vanes tending to cause movement. Due to the continual drop in pressure that takes place in each row of blades, (3) a, b, c, with the accompanying increase in the velocity, and also to the fact that the passage between the adjacent blades of a single row converges, as shown in Fig. 8, there is a force of reaction present, tending to cause movement of the vanes in the same direction as that due to the impulse. This increase in velocity is due to the drop in pressure and to the oridinary nozzle effect of the converging passage. Upon leaving the vanes b and entering the guides c, none of the velocity is given up, as the guides have no motion. This results in a still further increase in the velocity, (3) b to c.
After traversing a number of rows of tbe same height (A-B), the gradual expansion of the steam has increased the velocity to such an extent that the ratio of the velocity of the steam to the speed of the blades, which for maximum efficiency should be about as ten is to seven, has been greatly exceeded. To re-establish this ratio, the height of the blades is increased in a group, (1) B, with the result that the increased area of passage for the steam reduces the absolute velocity of the steam, and the increased height of the blades gives a greater peripheral speed of blades for a constant number of revolutions (radius being greater). Each group of blades of the same height is known as a barrel, or preferably, an expansion.
As the steam expands adiabatically (without gain or loss of heat), the velocity increases correspondingly; that is, the velocity will follow a curve which is the reverse of the adiabatic curve of expansion. This curve gives the theoretical blade height, as shown in Fig. 6.
It is true that bv the increase of blade heights bv groups, there is a loss in each row of blades that does not coincide with the curve. This loss could be eliminated by having each row of blades of slightly greater height than the next preceding one, but in the opinion of the designer, the losses are not of sufficient magnitude to warrant the increased cost of construction.
The increase in the height of the blade is limited by the struc-
Steam Turbines.
tural strength of the blade material. In the turbine shown in Fig. 6, this point of maximum blade height is reached after the fourth expansion of the H. P. rotor. By passing the steam to a second cylinder of larger diameter (L. P. rotor), it is possible to reduce the actual blade height by an increase in the diameter of the drum or rotor containing the moving blades and still maintain an efficient ratio between the steam velocity and the speed of blades.
Throughout the expansion of the steam down to the seventh expansion of the L. P. rotor, the distance between the blades,
FORWARD
©
_ © |
|
r-—-------------------- 4>---------------- |
|
tuuum |
|
©. ... |
TTTTTTTttI
AFT
Fig. 7.—Wing Blading Reaction Turbine.
the curvature of the blades, and the entrance and exit angles have remained constant, the whole increased volume of steam being accommodated by the increase in the blade height. As shown in Fig. 6, as the pressure approaches zero, the volume increases to such an extent that further increase in blade height is out of the question; also it is impossible to exhaust the steam to a cylinder of greater radius, as was done in the previous case, because of the excessive dimensions called for to accommodate steam of such great volume. It becomes necessary therefore to resort to other means; this is done as shown in Fig. 7, in the seventh and eighth
expansion of the L. P. rotor, by decreasing the number of blades in each row and decreasing the curvature of the blades. These altered blades are known as wing blades.
Referring to Fig. 8, a theoretical velocity diagram, the steam of pressure P enters the guides and, because of the slight drop in pressure to Plt the resulting expansion and the form of the guides, the steam moves along the line AB with a velocity V^; but
P
Fig. 8.—Theoretical Velocity Diagram, Reaction Turbine.
by compounding with the speed of the blade u, as in Fig. 3, AC is found to be the relative direction and vx the relative velocity of the steam. As already explained, the impulse and the force of reaction cause the movement of the blades. Again, the steam of pressure Pt undergoes the same expansion in dropping to pressure P2, and as the blades are merely reversed, the steam moves along DE with a velocity v2. By compounding as before, DF is found to be the absolute exit direction and V2 the absolute exit velocity of the steam leaving the moving blades. From this figure the converging of the passage between the blades is shown by a,
the width of the entering jet of steam, and b, the width of the exit jet of steam ; the usual ratio is about a — 2.b to a—3b.
In actual turbine practice, the expansion of the steam in the moving blades increases the velocity of flow to such an extent that, in spite of friction and other losses, the relative exit angle a and velocity DE, and the absolute entrance angle a and velocity AB, are about equal. The difference in pressure on the two sides of any row (Fj—P2) of blades has the effect of exerting an axial thrust in the direction of flow which may or may not be corrected, depending upon the character of the work to be performed by the turbine.
Fig. 5 shows the annular passage that exists between the end or tip of the guides and the rotor, and between the tipof the vanes and the casing. As the pressure of the steam drops continually from one row of blades to another it is evident that there will be a continual leakage of steam around the tips of the blades, with a corresponding loss of the energy contained in the leaking steam. This is called tip leakage. There is one feature of this leaking steam that counteracts in a small way the loss by tip leakage ; that is, the heat generated by the expansion of the steam while passing around the tips has the tendency of superheating the steam that has passed through the blades.
The action in subsequent rows of blades is similar to the action already described. The energy is absorbed proportionally at each stage, and after expanding to the exhaust pressure the steam passes to the condenser.
Contrast of the Impulse and Reaction Turbines.
The DeLaval and Parsons turbines have been selected to illustrate the two extreme principles involved in practical steam turbines of the present day. It is necessary that the salient features of these two be thoroughly understood, as all other practical turbines simply involve modifications of principles given heretofore.
The Impulse Turbine Possesses (Single Stage):
1. Complete expansion of steam in the nozzle from initial to final pressure.
2. Single turbine wheel to absorb kinetic energy.
3. Exceedingly high velocity of steam (3000-4000 feet per second) and high peripheral speed of moving blades (1000-1200 feet per second).
4. Simple absorption of kinetic energy in the moving blades.
5. Same steam pressure before and after passing through moving blades.
6. Absence of axial thrust.
7- No tip leakage.
The Reaction Turbine Possesses {Compound):
i- Gradual expansion of steam, in each row of blades, moving as well as fixed.
2. From 30 to 150, or more, rows of moving blades to absorb kinetic energy as it is generated.
3. Comparatively low velocity of steam (250 feet per second) and low peripheral speed of blades (125 feet per second).
4- The generation as well as the absorption of kinetic energy in the moving blades.
5. Drop in pressure at each row of blades.
6. Axial thrust in the direction of the flow.
7. Tip leakage at each row of blades.
To contrast still further the two types of turbines, and to draw a comparison as to the relative speed of rotation necessary for theoretical maximum efficiency, supplementary velocity diagrams are shown in Figs. 9 and 10.
In considering the action of the steam in the cases" shown in Figs. 9 and 10, it must be remembered that the single-stage impulse turbine and the single-stage pure reaction turbine (no impulse) are the two types under discussion.
To review slightly, it is known that in the impulse turbine the steam expands in a separate nozzle, impinges upon the blades with a certain relative velocity, and, if losses be neglected, leaves the blades with a relative velocity equal to that of the entering jet.
Neglecting all other considerations, it is evident that to extract all the energy from the jet of steam, it should emerge from the moving blades with an absolute velocity whose component in the direction of rotation is zero. Such a case for the impulse turbine is found by constructing Fig. 9 with V the velocity of the jet,
with the rotation component DB, and u the speed of rotation taken as one-half of DB, or «= ^ cos a. By compounding the velocities V and u and the angle a, E1 is the relative entering velocity. Being a symmetrical blade j8 = j31( V2 shows the relative exit velocity, and CE the relative exit direction, of the steam. By
Tfieoref/c#/ l/e/oc/fy JP/ayra/v. fmptf/se -wfiere
further compounding with u, V3 is found to be the absolute exit velocity and direction; and by a review of the construction of the figure, it is evident that V3 has a zero component in the direction of rotation. Therefore the full energy of the jet of steam has been extracted by' having the speed of rotation equal to one-half
V
of the rotation component DB of the velocity V, or u= — cos a.
The fact that the rotation component DB does not equal V, the velocity, but E cos a, is necessitated by the mechanical reason that
for the steam to flow in an axial direction it is necessary for the steam to be introduced on one side of the blade, as in this case at the angle a, and the blades must be so shaped that the steam will discharge on the other side. But to illustrate a theoretical case: as a is decreased, /3 and decrease in a certain ratio, until, at the point where a = o, AB (V) coincides with DB (Fcosa), (cos o° — 1; therefore V = V cos a=2u. Also, with a approaching zero, AD and DE approach zero, thus reducing V., to zero.
In contrast to the impulse turbine, take the theoretical case of
7y?co/ie//ca/ l/e/oc/fy JD/op/a/n.
/fee/cf/or? —wAere ^U/= \A cos ex.
/?
the pure reaction turbine, where the steam of comparatively low velocity enters the vanes of the moving wheel and, while traversing the passage, imparts to itself a velocity, by expansion, which exerts upon the vanes a force of reaction.
As was the case in the impulse turbine, to extract the full energy from the steam jet, it must leave the vanes with a velocity whose component in the direction of rotation equals zero. For the determination of such a case, construct Fig. 10 so that u (speed of rotation) — V cos a (rotation component of V). Then, by compounding the two velocities and the angle a, V1 is found to be the relative entering velocity and direction. By expansion through blades of same curvature (a = a1), the relative exit
velocity V2 equals V. By again compounding with u, V3 is found to be the absolute exit velocity. As before, by reviewing the construction of the figure, V3 is seen to have a rotation component equal to zero. Therefore all the energy has been extracted b> employing a speed of rotation equal to the rotation component of the velocity of the steam jet, or u=V cos a.
For the same reasons as stated in the case of the impulse turbine, it is necessary to introduce the steam on one side of the blade and to have blades of such a construction that the steam may exhaust on the other side; but, as also stated before (to illustrate a theoretical case) : as a approaches zero, AB more nearly coincides with BC, and when a=o, V — u, and Vz o.
In conclusion, it may be stated that for a single-stage impulse turbine, the maximum efficiency is obtained when the speed of rotation is one-half the velocity of the steam jet, while for a pure reaction turbine, the speed of rotation must equal the velocity of the steam. As the Parsons turbine is a combination of impulse and reaction, the speed of rotation must be about 70 per cent of the velocity of the steam for maximum efficiency.
Again by employing a single-stage impulse turbine with a peripheral speed of one-half the velocity of the steam, all of the energy is extracted from the jet in its passage through the moving blades. The most difficult problem in the design of turbines for marine use is that of reducing the speed of the blades, without enormous increase in 'weight and radial dimensions, and still maintaining efficiency in complete utilization of the total energy
available. .
This restriction is necessitated by the difficulty experienced in obtaining a propeller which will fulfill the requirements of efficient propulsion while rotating at a speed which will allow efficient operation of the turbines.
It will be remembered, from the proofs of Figs. 9 and 10, that for equally efficient operation, the single-stage impulse-reaction turbine must rotate at about 40 per cent greater speed than the single-stage impulse turbine, for an equal steam expansion.
The Parsons turbine, as already described, does keep a more or less constant ratio of steam velocity to the peripheral speed of the blade, by the very gradual expansion of the steam and the increase of blade height and rotor diameters; but for the impulse turbine the problem is dealt with by compounding, that is, by the compound impulse turbine.
Before embarking upon the description of the compound impulse turbine, it must be understood that if, in the single-stage impulse turbine, the steam velocity is reduced by twice the velocity of the wheel for maximum efficiency (Fig. 9), it is obvious from a study of Fig. 11 that if two rows of moving blades, with an
W/?ere st#— ct
.
IA///Z? /wo rows of
intervening row of fixed blades, be arranged in series, the blade speed need be but one-fourth the steam velocity for all the energy to be absorbed. This compounding for the absorption of all the energy contained in the jet of steam of the same pressure is known as compounding for velocity, each row of moving blades being called a velocity stage.
3IQ
Compound Impulse Turbine.
A simple example of the compound impulse turbine will lx described to show the principle of operation. This turbine combines the impulse feature of the DeLaval turbine, with a modi - cation of the compound feature of the Parsons turbine. Generally speaking, the steam is compounded for pressure by the use of a series of pressure stages, and in each pressure stage the steam
is again compounded for velocity by the use of a number of rows of blades. Stated' simply, the steam is compounded for both
pressure and velocity. .
Fig 12 shows a two-stage impulse turbine: (i) is a sectional view through the nozzles and blades; (2) is a diagrammatic sketch of the nozzles and blades; and (3) is a pressure and velocity line diagram.
The steam at a is of high pressure and low velocity, (3) a. By expansion through the nozzle, the steam velocity increases, (3) b, to the extent allowed by the drop in pressure, (3) a to b; and upon issuing from the nozzle, the jet of steam impinges upon the blades that come within the limits of that jet. The impulse of the steam here is similar to that which takes place in the DeLaval turbine; but upon issuing from the moving blades, the velocity, though reduced, is still capable of imparting energy to the moving blades; so, by the passage of steam through a segment of reversed blades, (2) c, of sufficient circumferential dimensions to include the jet of steam, the direction is changed to conform to the original direction. By the passage of the jet through the second row of moving blades the velocity is reduced by twice the peripheral speed of the blades, (3) d. The velocity has now been reduced to such a point that further extraction is impossible.
The steam now enters a set of nozzles of such a number that the increased volume of steam may be accommodated. These nozzles are arranged around the circumference, so that the jet of steam issuing therefrom may be a continuous band of steam throughout the segment rather than a number of smaller jets with their resulting eddy losses. In these nozzles of the second pressure stage the expansion brought by the drop in pressure, (3) d to e, generates a velocity for the second time, (3) d to e. In this pressure stage the steam velocity is absorbed in the same manner as in the first stage. After issuing from the last row of blades the steam is exhausted to the condenser or to the atmosphere, depending upon the installation.
The Compound Impulse Turbine Possesses:
1. A few well defined expansions from the initial to the final pressure.
2. Two or more rows of moving blades to absorb the energy in each pressure stage.
3. Moderate steam velocities and low speed of rotation.
4. Simple absorption of energy in the moving blades.
5. Same steam pressure throughout the whole of any one pressure stage.
6. No axial thrust on blades.
7. No tip leakage, as the steam is of uniform pressure in each pressure stage.