would be but 13°7

--

2.17 = 11.53 lbs.; whereas, if the barometer stood at 31 inches, a perfect vacuum would be 15.2; and if the water was but 100°, its vapour would give a resistance of only 9, and, consequently, the highest attainable vacuum would be 15.2 — ·9 = 14°3 lbs., making a difference of 2'77 lbs.*

The table showing the force with which vapour or uncondensed steam resists the ascent or descent of the piston, according to the temperature of the water discharged from the air-pumps (Table II at the end of the work), and also the table of the variations of the weight of the atmosphere, as shown by the barometer (Table I at the end of the work), will make apparent the necessity of these observations.

315. Before concluding this subject, it is important to remark that in using an indicator for measuring work, that the distinction between indicated power-that is, the power developed in the cylinder-and "effective power," must never be lost sight of. The latter is the useful effect of the engine, while the former includes the power expended in working the pumps, overcoming the friction of the machinery, and all resistances which exist within the engine itself. In the marine engine it is very difficult to arrive at the amount of power really used in propelling a ship. That expended in driving the engine and pumps at different speeds can be approximated to by disconnecting the engines from the propeller shafting, and taking diagrams with the steam throttled down just sufficient for driving the engine at the required speed. Great care is necessary for starting the engines in taking these diagrams.

316. Further Remarks on Indicator Diagrams. It is only by a proper employment of this inestimable instrument—the indicator-that tolerably true deductions as to the internal working of an engine can be arrived at with any degree of certainty-that a correct diagnosis can be formed as to the real state of the iron patient. Such being the case, it is, perhaps, surprising that more attention has not been directed to certain influences that, in some cases, undoutedly distort an indicator diagram. These disturbing actions may be greatly modified by the construction of the indicator itself; they may be increased or diminished by a proper or improper construction of the instrument; but these influences will always exist to a greater or less extent, and some allowance should be made on this account, before the diagram be accepted as a perfectly true picture of the performance of the steam in the cylinder.

(a) The steam engine indicator is especially liable to several objections when used for determining forces acting with great velocity and suddenness.

* The vacuum shown by the indicator will generally vary from that shown by the vacuum gauge, when it is constructed with a glass tube, hermetically sealed at the top; for such gauges are designed to show the variation from a perfect vacuum, without reference to the weight of the atmosphere; but the vacuum shown by the indicator is affected by all its variations.

The inlet of the steam is generally throttled by the narrowed aperture of the cock, and the indicator piston in its sudden ascent has to meet the column of air above. In the ordinary instrument there is a long and tremulous spring, which, from its length and weakness, is somewhat liable to rub on one side of the casing. The working parts are also often made unnecessarily heavy. When steam of a high pressure is let into the indicator piston, the spring gives way and is suddenly thrown up; this action is aided by the momentum of the other moving parts; the pencil continues to oscillate during the remainder of the stroke; and the natural result is a jumping, serrated, zig-zag diagram (see Fig. 65, page 357). The total power indicated is, perhaps, not very far from exactitude, as the action and re-action are probably about equal; and as in interpreting the diagram the true curve is supposed to run midway between the crests and hollows of the wavy line produced by these oscillations-but the most important function of the indicator consists in the giving the true performance and progressive course of the steam in the cylinder, and it is evident that, under such circumstances, the lines of the diagram have not the accuracy that is demanded by the importance of the deductions the experimenter wishes to obtain.

(b) There can be little doubt that the larger the indicator the better. The friction of the piston and of the moving parts is less in proportion; the cock and pipe are larger, the steam is thus less liable to be throttled in its passage to the piston, &c., and the divisions on the scale and the diagram itself being of a larger size, there is less chance of error.* (See page 314).

The original indicators of BOULTON and WATT were all of a large size, and were made with a large powerful spring, while the stroke of the piston was much shorter than in the ordinary indicator as made at present. It is stated that WATT obtained the scale of pounds by actual trial of weights pressing down the spring; but these results were taken while the indicator was cold, and therefore in a very different state, and in a different relation of the parts by measure than when the instrument was doing its work. It is evident that a more correct way would be to apply the weights on the top side of the piston while the indicator is fixed on the cylinder of an engine, the instrument being then in the same state as when performing its proper functions, the spring is of the same size and of the same elasticity as when in use, the whole apparatus on it at a similar temperature, and the piston is lubricated with the same quantity of oil.

The weights being, one after the other, applied to the top of the piston, the atmospheric line and the gradual descent of the piston are then recorded by the pencil of the instrument, and the principal divisions are thus obtained for a complete subdivision, and a true scale is the result. Perhaps the most exact way would be to use a properly graduated mercurial column. There would certainly be some difficulty with this plan, with an indicator for very high pressure. Mr. PARKES, the civil engineer, found that the pressures for the scale of the indicator were more correctly obtained by comparing them with a mercurial column than by ascertaining them by means of weights. "The instrument being then heated, was then in precisely the same state as when it was in use." He also found that a certain amount of correction was frequently necessary, as both the spring and the amount of piston friction were affected by heat.

According to Professor MOSELY, the scale of ordinary indicators should not always be in equal divisions, because the wire of the spring being wound spirally into a screw of small diameter, the spiral obliquity of the thread of such a screw becomes more oblique to the direction of the bending force as the spring is stretched, and less oblique as the spring is compressed, and hence the scale of pounds per square inch, by which the curve should be measured for summing up the results, ought to be a scale of unequal divisions."

(c) The friction of the indicator, by directly opposing the motion of its piston and pencil, tends to make the indicated forward pressure less, and the indicated back pressure greater, thus the real forward and back pressures respectively, and so to make the indicated energy less than the real energy exerted by the steam on the piston, but to what extent is very uncertain. According to some experiments by Mr. HIRN (Bulletin de Mulhouse, vols. XXVII and XXVIII), the diminution of the indicated energy by the friction of the indicator agrees nearly with the work performed in overcoming the friction of the steam engine, so that the indicator shows, not the whole energy exerted by the steam on the piston, but very nearly the useful work of the steam engine; but it is doubtful how far this principle is generally applicable; and other experiments, especially those on screw steamers, are at variance with it.

(d) As is well known, the part of the engine giving motion to the string of the indicator ought always to be as near the instrument as possible, as otherwise the string will slacken, and the barrel will thus be stationary for an interval, of course, the result being a wrong diagram. As it is of such great importance that the string attached to the indicator barrel should neither be too slack nor too tight and liable to stretch, it has been suggested that it might be advisable to use a kind of string worked up with fine wire, similar to the tape measures that are sold for measuring lengths of 30 or 40 feet.*

(e) There is also sometimes a danger to the exactitude of the diagram, from the steam getting condensed. The most convenient way of taking a double diagram, as already stated, is by fixing the indicator to a pipe leading from the top to the bottom of the cylinder. This pipe being fitted with stop cocks, the two diagrams can be taken simultaneously. The great objection, however, to this plan, consists in the natural tendency of the steam to get condensed in this small pipe.

(f) It appears that the steam pressure undergoes some fall during the passage from the boiler to the cylinder. The amount of such fall varies greatly in different engines, but the general result is that the highest indicated steam pressure, before expansion begins, is some two or three pounds less than the boiler pressure.

Among the points to be noticed are (1) the resistance of the steam-pipe through which the steam passes, (2) the resistance of the regulator or throttlevalve, (3) the resistance due to the ports and steam passages; and here, also, the bends or sharp angles as well as the imperfect coating of the steam-pipe must be taken into account.

RANKINE says that in the present state of our knowledge it is impossible to calculate separately the losses of pressure due to these causes, and if it

* Mechanics' Magazine, vol. 9, New Series.

were possible the resulting formula would be too complicated to be of much use. An observation of this kind has a wide application. It may be pointed out that steam which has been lowered in pressure by the resistance of passages, or has been wire-drawn, as we have termed it, is to some extent superheated by the friction of its molecules, the tendency of all friction being to produce heat.

(9) There is, in practice, a rounding of the angle at which the expansion curve begins in the theoretical diagram. This is called wire-drawing at cut off. It is always to be seen where a slide-valve closing gradually is employed, and is reduced to a minimum in a well formed diagram of a Corliss engine. Speaking generally, it may be said that the steam begins, as it were, to work expansively a little before the valve is completely closed; or that the energy exerted is nearly the same as if the valve had closed instantaneously at a somewhat earlier period of the stroke, which point may be termed the effective cut off. Such a point is easily obtained by carrying the expansion curve a little higher, and by prolonging the probable steam line to meet it.

(h) There is a rounding of the expansion curve when release begins before the end of the stroke, and it is recommended that the point of release should be so advanced that one-half of the fall of pressure takes place at the end of the forward stroke, and the other half at the beginning of the return stroke. When the release is small the expansion curve is continued to the end of the diagram, and in such a case the exhaust line slopes gradually downwards as the piston returns, instead of being nearly horizontal.

(i) The general effect of water in the cylinder, from whatever cause produced, but which we will suppose to be present in some degree throughout the stroke, is to lower the steam line in the first portion of the stroke, and to raise it in the latter portion. On this subject it is very easy to propound theories, but the subject-matter lies so much within the region of experiment, that any theoretical deductions, which are nearly all that we can be said to have at present, may indeed be interesting, but would perhaps admit of being classed among conclusions in which nothing is concluded.

(j) Another cause which affects the form of the expansion curve, and which has been the subject of much observation, is that of the partial liquifaction of the steam at the beginning of the stroke, and immediately after being cut off; and again, the re-evaporation of this condensed steam towards the termination of the stroke. This is most noticeable in engines in which the steam is greatly expanded. The effect is shown upon the diagram by the lessening of the initial pressure by the falling of the pressure immediately after cut off below the hyperbola, and again by the rise of the pencil above the hyperbolic line towards the end of the stroke. A partial explanation appears to be that the incoming steam is cooled and partially liquified by contact with the sides of the cylinder and passages, which have just been

cooled by communication with the condenser, and that towards the end of the stroke the steam is re-evaporated by the heat conveyed by the sides of the cylinder, which is then at a high temperature as compared with that of the expanded steam. ZEUNER, however, a well-known investigator, has given calculations to show that the approach of the line to the hyperbola towards the end of the stroke is affected by the initial condensation of the steam as to dryness approaching it the more nearly the greater the quantity of moisture it originally contained. ZEUNER concludes, therefore, that the action of the sides of the cylinder has been over-rated,*and there are practical engineers who ascribe the rise of the pressure solely to leakage through the valves, contending also that it is absurd to suppose that the amount of heat transmitted to the mass of steam and water in so short a time would reevaporate it to such an extent as to account for this phenomenon. It is to be remembered, however, that leakage through the pistons, in this case, tends to counteract the effect of leakage through the valves, and the rise of pressure is so very marked in engines of the best make, when working with great expansion, that there can be little doubt, after making allowance for other disturbing causes, that the steam is affected in the manner stated above.

In addition to the action of the sides of the cylinder, the presence of water due to the liquifaction of steam during expansion also affects the diagram, and with respect to this Prof. RANKINE states:-That liquifaction does not, when it first takes place, directly constitute a waste of heat or of energy, for it is accompanied by a corresponding performance of work. It does, however, afterwards, by an indirect process, diminish the efficiency of the engine; for the water which becomes liquid in the cylinder, probably in the form of mist and spray, acts as a distributer of heat and equalizer of temperature, abstracting heat from the hot and dense steam during its admission into the cylinder, and communicating that heat to the cool and rarefied steam which is on the point of being discharged, and thus lowering the initial pressure and increasing the final pressure of the steam, but lowering the initial pressure much more than the final pressure is increased, and so producing a loss of energy which cannot be estimated theoretically.

(k) Clearance will modify the form of the expansion curve of steam by removing backwards through a small space the zero line of volumes. And, as we have seen, if the steam be completely exhausted from the cylinder during the return stroke, the effect of clearance is to waste during the double stroke. But, inasmuch as it is possible to compress a portion of steam in the cylinder during the return stroke, the loss above referred to may be greatly,

* ZEUNER in a paper, "Ueber die Wirkung des Drosselus und den Einfluss des schädlichen Raumes auf die bei Dampf-maschinen verbranchte Dampfmerge," Civil Ingénieur, vol. XXI, p. 1, published in 1875, states his belief that a more powerful cause must be found for some of the effects observed, and finds it in certain consequences of wire-drawing. See further on this subject "The Steam Engine considered as a Heat Engine." By JAMES H. COTTERILL, M.A., chapter X.