The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

142 'gas engine was running, working on the Otto cycle. The suction gas genera- ting plant, practically a large Ameri- 'can stove, the gas from which is util- 'ized instead of passing up the chimney, 'was then examined, and its construc- tion explained by Mr. Sellers. The dis- tinctive feature of this plant is. the patent exhaust evaporator, the ar- rangement exhaust "consisting of an THE /\aARINE REVIEW silencer fitted: with an internal non- pressure evaporator, in which steam 1s raised by the waste heat from the ex- haust. The plant is also fitted with a hot gas evaporator and* air heater, _which consists of a water jacket round pipe, immediately above the so that the water and air and steam is - formed by By means of. the. gas generator, are heated, the heat of the gas. ; June, 1909 these two evaporators, therefore, . prac- tically all the waste heat is utilized. Anthracite coal is -used, and in the case of the 25 B. H. P. producer in use at the works, it was stated that the 'cost averaged 1s per day in fuel for the ordinary working day. It is claimed that the usual cost of the suction gas | 'producer is one-tenth of a penny per H. P. per hour. Measurement of Marine Engine Power by Torsionmeters | T A MEETING of the Green- ock Association of Engineers and Ship . Builders, held in Greenock recently, a paper on~ "The Measurement of Power é ' oo% % by Means of 'Torsionmeters," was read by Joseph Menmuir, A.M, Inst. E.E., of the staff of Messrs. Kelvin and James White, Ltd. Glasgow. After referring to the introduction "and development of the steam _ tur- bine and to the inability of engineers 'to measure, by means of the steam indicator, the power developed in them, as in the case ot ordinary re- ciprocating. engines, the author out- lined the steps which have led to the invention of various forms of - tor- ~ sionmeters . for the purpose. Having ' explained the angle of twist, or "tor- sion," of a shaft rotating and trans- _ mitting power, and that this varied _ directly as the moment of power ap- plied and the length of shaft over which the measurement was made, and inversely as the modulus of rigid- ity of the shaft and the diameter of the shaft to the fourth power, the author said that the torsionmeter provided a means of measuring the angular deflection on a length of shaft while rotating and transmitting power, and from this the value of the power could be easily calculated. Teles ~ @=Angle of deflection. O N= Number of.revolutions per minute. _, L=Length of shaft in feet. M= Modulus of. rigidity. D=Diameter of shaft in inches. 8 D*N Then H. P.= ML The law embodied in the above form- _ula was found to hold good for all shafts which were not stressed beyond the. elastic limit. The angle through which a shaft was deflected at normal full load was very small, the actual value at ordinary factor of safety being about 1 degree per 10 ft. in length. This meant a deflection of %-in. at the surface of a shaft 12 in. diameter and 10 ft. tong. . The measurement of this angle, therefore, must be very accurately carried out if reliable results were to be obtained. The error of a hairsbreadth, in fact, might mean several thousand horsepower. Marine Shafting. Referring to marine shafting partic- ularly, the author proceeded to say that in order to secure the best possible re- sults the shafts should be calibrated be- forte being put in position in the vessel. "This is. done by bolting up the coup- lings of the desired length of shaft and suitably supporting same. The one end is fixed rigidly, and to the other end a | lever. is attached, to the outer end of which various weights can be applied. Another method is by using two levers at opposite sides of the shaft, one be- ing pulled downwards and the other up- wards by means of spring balances. This gives a more equal: turning moment and reduces the chance of error due to friction at the point of support. Two pointers are attached to the shaft, one clost to each end, and these move over graduated scales. If the pointers are made 57.3 in. long from center of shaft the reading will be exactly. 1 degree per inch of deflection... The shaft is now subjected to a known twisting moment by means of the weights or spring bal- ances and the deflections of the two points noted. The difference between the readings will give the true deflection on the part of shaft between the point- ers. With reference to couplings which may intervene within the test length it is found sufficiently accurate im practice to deduct the axial length of these en- tirely. The net rated length of shaft will therefore be equal to the length be- tween the two pointers, less the sum of the couplings which intervene." The author had been fortunate in having access to the results of tests of a large number of shafts, constructed by differ- ent firms in this country, and it was surprising to find how closely these re- sults agreed with each other. This was specially the case for solid shafts, and the mean value got from these tests could be accepted as correct within a very small percentage. Indeed, the act- ual calibration 6f many shafts could be dispensed with entirely. For hollow shafts the results were not so consist- ent, but where extreme accuracy was not required the torsional rigidity of a hollow shaft could be taken as equal to that of a solid shaft of the same ex- ternal diameter, minus that of a solid shaft of the same diameter as the bore of the hollow shaft. Compression Stresses In addition to being subjected to a twisting moment propeller. shafts were at the same time subject to a compress- ing stress equal to the thrust transmitted from the propeller to the thrust block. Some engineers held that the thrust did not affect the accuracy of the torsional readings, and one eminent marine engineer on the Clyde had in- formed the author that he had looked very carfully into the matter and was convinced that it only affected the tor- sional readings to the extent that the shaft was made shorter under compres- sion, In opposition to this view, how- evr, Mr. J. J. Hamilton Gibson in a pa- per read before the North East Coast In- stitution of Engineers and Shipbuilders in January, 1908, stated that compression augmented the torsional deflection about 1 per cent for: solid shafts, and as much as 3 per cent in the case of hollow shafts at full power, and he recommend- ed that while the shaft was under static calibration it ought also to be subjected to a compression stress such as it would sustain under working conditions.