Methods of Determining the Dryness of Steam.-Report of the Committee, consisting of Sir F. J. BRAMWELL, Bart. (Chairman), Professor T. H. BEARE, Mr. JEREMIAH HEAD, Professor A. B. W. KENNEDY, Professor OSBORNE REYNOLDS, Mr. MAIR RUMLEY, Mr. C. I. WILSON, and Professor W. C. UNWIN (Secretary).

THE determination of the quality of steam, or its wetness, has come to be of very great importance both in boiler trials and engine trials. It was in the classical researches of G. A. Hirn at Logelbach in Alsace in 1854-57, and in subsequent experiments by MM. Grostête, Hallauer, and Leloutre, and other members of the Société Industrielle of Mulhouse, that the great importance of the action of water in the steam-engine cylinder was first clearly recognised. It was perceived that the water present in the cylinder, and which may amount to 50 per cent. of the steam introduced, was due to three distinct causes: (a) initial wetness of the steam; (b) condensation in the mass of steam due to adiabatic expansion; (c) the action of the cylinder wall. An accurate discussion of the results of an engine test is impossible if the initial condition of the steam is unknown, both because the total heat per pound of steam received by the engine depends on its dryness, and because water introduced with the steam increases the prejudicial action of the cylinder wall.

In boiler tests the evaporative efficiency is measured in pounds of water converted into steam per pound of coal. But if part of the feed water supplied leaves the boiler as water entrained in the steam, the real evaporation is less than the apparent evaporation. The entrained water must be allowed for in estimating the evaporative efficiency of the boiler. Further, a boiler which supplies wet steam is a bad boiler, because wet steam is prejudicial to the efficiency of the engine, and it should be one object of a boiler trial to determine whether the steam is of good quality.

The earliest attempts to determine the amount of moisture in steam which have come to our knowledge were made during some boiler trials by a committee of the Société Industrielle of Mulhouse. This committee tried three methods: 1. A Method of Separation.-The whole of the steam was taken through a spare boiler, without fire, in the expectation that the moisture would be deposited. The increase of water in the spare boiler during a trial was measured, and that part of the increase due to condensation by radiation was determined by a special experiment. 2. A Condensing Method suggested by Hirn. Some of the steam from a boiler under trial was condensed and its total heat determined. By comparing the total heat per pound with that of dry saturated steam the amount of moisture was ascertained. 3. A Chemical Method.-Salt was introduced into the boiler, and the diminution of saltness of the boiler water during a test determined. It was assumed that moisture in the steam would be of the same saltness as the boiler water, and therefore the amount of salt removed measured the wetness of the steam. In these early trials only the second method appeared to give reasonable results. But the committee did not place full reliance on any of the methods tried.

Origin of the Water Suspended or Entrained in Steam.-There are

three sources of the water found in steam :—

1 Report of MM. E. Burnat and E. Dubied,' Bulletin de la Société Industrielle de Mulhouse, 1859.

(1) By ebullition boiler water is projected into the steam space. Part falls back, but part is carried on in the steam current. The extent to which wetness may be thus produced depends on the activity of the ebullition, the area of the water surface, the volume of the steam space, the position of the steam valve, the density of the steam, and other circumstances. In certain conditions of the boiler water, it foams, and the steam space is filled with vesicles. Under such circumstances the mechanically produced priming may be excessively severe.

As to priming of this kind some observations of Mr. Thornycroft on a boiler with glass ends are very instructive. He states that 'waters which cause priming on boiling produce foam, consisting of a mass of bubbles of various sizes. Water which is very bad produces bubbles so durable as to remain a considerable time without breaking, and by them the steam space of a boiler may be entirely filled. So soon as this takes place, instead of simply steam leaving the boiler, the discharge consists of foam, which is broken up in its rapid passage through the steam pipe.' With pure water, steam retains no film of liquid for sufficient time to be

seen.

(2) The steam in the boiler is subject to variations of pressure. Bubbles formed under water rise to a region of less pressure. Fluctuations of pressure arise from the intermittent demand for steam. During expansions water must be formed as mist throughout the mass of the steam. It is difficult, however, to suppose that any great quantity of moisture is thus produced.

(3) The steam in the steam space of the boiler, and when flowing through the steam pipes, loses heat by radiation from the boiler roof and the surfaces of the pipes. To this must correspond condensation of part of the steam. Probably in some cases very considerable amounts of moisture are produced in this way.

Methods of Determining the Wetness of Steam.-Very different methods have been tried by different observers to determine the amount of moisture in steam. Some method is required sufficiently accurate for practical purposes, and not involving excessively delicate measurements or complicated apparatus. It is proposed to describe all the methods which have been tried which seem at all likely to be useful, and finally to give the results of some comparative trials made for the committee which throw light on their relative trustworthiness.

I. Weighing Method. A method of direct weighing of a known volume of a sample of steam has been proposed by Guzzi and Knight.3 A copper globe is used as a measuring vessel, which is placed in a receiver connected with the boiler or steam pipe. After filling, it is taken out and weighed. Let V be the volume of the globe, w the weight of wet steam in it. Let (1-x)w be the moisture and xw the steam in the globe, and let v be the volume in cubic feet per pound of dry steam in the same conditions of pressure and temperature. Then

The method is one obviously of excessive difficulty.

'Circulation in the Thornycroft Water Tube Boiler,' Trans. Inst. Naval Architects, 1894.

2 Revue Industrielle, 1878, p. 102.

3 Journal of the Franklin Institute, 1877, p. 358.

II. Separating Method. Attempts have been made to ascertain the amount of moisture in steam by measuring the quantity of water trapped in a separator, between the boiler and engine, through which the whole of the steam passed. Probably such separators are too small relatively to the amount of steam flowing through them to detain the whole of the moisture. Mr. G. A. Barrus, in using a small separator in connection with a superheating calorimeter, noticed that very nearly the whole of the moisture was caught by the separator. In that case the quantity of steam passing through the separator is small. More recently Professor Carpenter, of Cornell University, has brought out a form of separating calorimeter of small size which can be applied almost anywhere on a steam pipe or boiler, and which can be used with very great facility.

The separating calorimeter (fig. 1) consists of a vessel, A, about 12 in. x 3 in., consisting of an inner chamber and a jacket. The steam from the steam pipe, S, passes first into the inner chamber, where the moisture is separated, and thence into the outer chamber. The separating chamber is consequently almost perfectly protected from radiation. As the water accumulates in the inner chamber its level is shown by a gauge glass, g, and the amount can be read off on a scale. A very small orifice at the bottom of the outer chamber regulates the amount of steam discharged. The steam as it escapes passes through a flexible tube to a simple form of condenser, C. The increase of weight of the condenser in any given interval of time is noted, and the amount of water deposited in the same time in the separator. If x is the dryness fraction of the steam, w the weight of water caught in the separator, and W the weight of steam condensed,

100

The scale on the separator is graduated to ths of a pound. There is also a gauge glass and scale on the condenser, graduated to read pounds. and tenths at a temperature of 110° F. But as the variation of volume of the water with temperature affects the readings considerably, it is best to weigh the condenser, or, at any rate, to determine a correction of the scale for temperature. Professor Carpenter states that the dryness of the steam after passing the separator was tested in the laboratories of Sibley College by several observers, and with steam carrying from 4 per cent. to 60 per cent. of moisture. In every case the separation of the water from the steam was complete and perfect. Other tests have been made with moderately dry steam, using the separating and throttling calorimeter simultaneously, and the results were practically identical. The instrument is very simple to use, and requires no pressure gauges or thermometers.

III. Condensing Method. Suppose a known weight of the steam condensed and its total heat determined. By comparing this with the total heat of an equal weight of dry saturated steam, according to Regnault's Tables, the amount of moisture in the steam can be determined.

The apparatus designed and used by Hirn in the Mulhouse trials in 1859 (fig. 2) is, so far as we know, the most convenient and perfect apparatus hitherto used in trials of this method. It consists of an iron vessel, C, about one foot in diameter, furnished with a loose cover. This

'See Barrus, Boiler Tests, Boston, U.S.A., 1891, p. 258.

See Experimental Engineering, by R. C. Carpenter, New York, 1892, P. 400.

is the condenser. A small pipe and cock in the steam pipe deliver steam, through a small orifice near the steam pipe, into the pipe S. An agitator, g, and a sensitive thermometer, i, are provided. The condenser is suspended

from a hydrostat, H, which permits extremely accurate weighing of the amount of steam condensed.

Let x be the dryness fraction of the steam, w the increase of weight of the calorimeter. Then the condenser has received xw lb. of dry steam