and (1-x)w lb. of water at the steam temperature. Let W be the weight of water initially in the condenser plus the equivalent reckoned as water of the condenser itself. Let t be the temperature of the steam, and i and f the initial and final temperatures of the condenser water. Then xw (1116-0-71t)+w(t-f)=W(f—i) W(f-i)-w(t-f) x= w(1116-0-71t) degree To arrive at satisfactory results the temperatures must be read to at least, and the weight of steam condensed determined very accurately. Further, it is desirable that the temperature i should be about as much below the temperature of the place where the test is made as the temperature ƒ is above it. A correction is easily made for the small fall of temperature by radiation during the time the agitator is used to secure uniformity of temperature. Suppose the agitator is used for two minutes, and the temperature ƒ then noted. The temperature is again noted two minutes later. The small observed fall is the loss in the interval by radiation, and is added to the observed temperature ƒ as a correction. In the Mulhouse experiments the initial and final temperatures of the condenser were generally about 60° and 110° F. The following table gives the results of the Mulhouse tests in 1859. They were made very carefully, and Hirn himself directed the tests. The committee at Mulhouse, in discussing these results, point out their variability, and remark that a mean of numerous observations should be taken. This is probably true when the test is made on a small scale. In Mr. Willans's tests one hundredweight of steam was condensed in each test, and this gives what is virtually a mean of numerous smaller tests. In the case of Trial 23, the Mulhouse Committee thought that the calorimeter underestimated the amount of priming water. About 1860, Joule used a calorimeter very similar to the above, in order to determine the total heat of steam and verify Regnault's results. The paper describing the experiments is interesting as showing that fairly accurate results can be obtained by this method, used on a small scale, when the observer is sufficiently skilful. The amount of steam condensed was only from 8,700 to 14,100 grains in each test. Temperatures were read to three decimal places. The weight of water in the condenser was about 140,000 grains. The steam flowed into the condenser for two minutes. The steam must have been almost perfectly dry, as the following (mean) results show : The method has been often used in determining the wetness of steam in boiler trials, but very often the arrangements have been so rough that no reliance can be placed on the results. A calorimeter of this 'The Scientific Papers of J. P. Joule, vol. i. p. 482. BOILER TRIALS AT MULHOUSE. Trials with Hirn's Calorimeter on the Quantity of Water entrained in Steam. Omitting Trial 23, the mean amount of water carried by the steam was 44 per cent. kind is called in America a barrel calorimeter. The barrel test is carried out thus. An ordinary oil barrel is used, fitted with an outlet valve. A set of tests are made following each other rapidly, and an average of the results is taken. The barrel is filled and a trial made, the results of which are discarded. The object seems to be to heat up portions of the apparatus and the barrel itself. The barrel is then refilled, weighed, and the temperature of the water taken. A steam pipe connected to the boiler is then blown through to warm it, and a movable pipe is attached leading to the barrel. The steam valve is then opened and the steam condensed in the barrel, till the temperature has risen to about 110° F. The valve is shut, the movable pipe removed, the water stirred, and the barrel again weighed and its temperature taken. It is obvious that the proceeding is a very rough one. Losses of heat by radiation, &c., tend to make the observed total heat too small, and therefore the moisture in the steam too large. On the other hand, evaporation from the barrel tends to make an error of the opposite kind. It is known that in using this method in a rough way negative results are sometimes obtained. Mr. P. W. Willans used this method also to determine the wetness of the steam in his engine trials. Mr. Willans used a very large con densing tank weighing full about 3 tons, and placed on the platform of a weighing machine. The tank was first balanced so that the weighing machine lever just rose. Then a standard hundredweight was placed on the platform of the machine, and the poise adjusted so that the lever again just rose. The hundredweight was then removed, the temperature of the condenser was noted, and steam condensed in the tank till the lever rose (making an electric signal). The steam at that moment was 1 Economy Trials of a Non-condensing Steam Engine,' by P. W. Willans, Proc. Inst. Civil Engineers, vol. xciii. p. 23. turned off, and the tank contained exactly one hundredweight of condensed steam. The condensing water was then stirred and the temperature taken. The thermometers were sensitive (1° C. corresponded to 1 in. in some of them), and they were carefully compared with standard thermometers. A very small radiation correction was made. The value of the dryness fraction found on different days was 9996, 9638, 9949, 9646, 9976, 9893, 1.0072, 1.0048, 9987. The mean of all is 0.9911, showing an average of about 1 per cent. of moisture in the steam. In spite of the large scale on which the experiment was tried and the care taken, the results are rather discrepant, and two of them give values which it would seem must be erroneous. IV. Continuous Condensing Method. The difficulties of the ordinary condensing method have led Mr. Barrus, Mr. Hoadley, and other observers to adopt a process of continuous steady condensation. The steam may be condensed (a) in the condensing water, or (B) in a surface condenser. The first method may be carried out by the apparatus shown in fig. 3. Steam passes from the steam pipe, S, to a small injector, i. The condensing water is drawn from the tank A, and the mixed water and condensed steam are discharged into the tank B. If W, is the decrease of weight of A in any interval, and W, the increase of weight of B, then w=W2-W, is the weight of steam condensed. Thermometers t, t, give the temperatures of the water entering, and mixed steam and water leaving the injector. Let t1, to be these temperatures, and t the temperature of the steam. Then, if x is the dryness fraction, Mr. Barrus has used a surface condenser (fig. 4), consisting merely of a short vertical pipe in a vessel through which cooling water circulates. The condenser is supplied with cooling water at a fixed rate of flow. Then, as the condensing surface is constant, the rate of condensation is constant, and the rise of temperature of the cooling water is constant. Condensation is carried on for any convenient period, and the condensed steam run off for weighing. Simultaneously a series of readings are taken of the temperature of the water entering and leaving the condenser, and of the condensed steam. The radiation correction can be determined by observing the fall of temperature in the condenser when the supply of steam and cooling water is stopped.2 The continuous condensing method seems likely to be more accurate than the ordinary method, but the arrangements are more complicated. V. Method by Superheating.-Mr. G. H. Barrus about 1890 devised a calorimeter in which the steam to be tested passed through a chamber jacketed with superheated steam. The steam was thus dried and super Carpenter, Experimental Engineering, p. 375. * See Peabody, Thermodynamics of the Steam Engine, p. 232; Carpenter, Erperimental Engineering, p. 380. heated. The quantity of heat required was inferred from the fall of temperature of the superheated steam. The orifices of the inner steam pipe and outer jacket were equal, so that the same weight of steam passed through each. Hence no weighing was necessary. If t, t, are the temperatures of the superheated steam entering and leaving the jacket, and t, t the temperatures of the steam entering and leaving the inner chamber, then the superheated steam in the jacket has lost t-t, degrees of temperature, and the sample steam has been dried and superheated t-t3 degrees. If x is the dryness fraction and L the latent heat of the steam in the inner chamber, 0·48(t,—t2-7)=L(1−x)+0·48(1, −13) I is a small correction for radiation. Only four thermometer readings are necessary for a determination. The calorimeter is only suitable for nearly dry steam, and appears to be superseded by the simpler throttling and separating calorimeters. A description will be found in Boiler Tests,' 1894. Ꭰ Ꭰ |