Carpenter is independently pursuing the study of the internal friction of engines. It is hoped that we may, in time, secure a very large body of valuable data, and especially that we may, after a time, be able to indicate the laws of its distribution with some accuracy for all types and styles of engine in common use. [NOTE.-This paper was presented and received discussion jointly with the Author's other paper entitled " On Variable Load, Internal Friction and Engine Speed and Work," published as No. CCCXVII. of the Transactions at the Scranton Meeting of the Society, Vol. X., page 138.] CCCXVII. ON VARIABLE LOAD, INTERNAL FRICTION, AND ENGINE SPEED AND WORK. BY ROBERT H. THURSTON, ITHACA, NEW YORK. IN a recent paper on the "Internal Friction of Non-condensing Engines," the writer gave the result of a series of experiments made to determine by further research the character of the internal friction of engines without condensers, and the method of its variation with variation of the usual conditions of operation. It had been shown previously that, in some classes of engine at least, this friction is constant with all loads whatever, up to and beyond the rated power of the machine. It was, in the last paper,* incidentally stated that the experiments kindly directed by Professor R. C. Carpenter, in the laboratory of the Sibley College, of Cornell University, have indicated a slight increase of internal resistances with increase of engine speed. For the small engine then tried, this increase amounted to eight per cent. of the number of revolutions made by the engine per minute. It is thus found, for that case, that the percentage of power lost by friction was a constant fraction of a given total power of the engine at all speeds. This investigation has been continued during the past college year, and further data obtained from a number of engines of vari ous types and sizes, the results of which investigation are now for the first time herewith presented. A summary of the work on the first engine used, and already reported upon, is here again given for comparison with the later results of trials made with other engines. It will be seen that the method of variation of this friction with change of speed is apparently very largely dependent upon the method of lubrication and its efficiency. * Vol. IX., Transactions Am. Soc. M. E., No. CCLXV. A very similar variation in this respect is to be traced in almost all cases of experiment made to determine the friction of journals or other rubbing surfaces, a very large number of examples of which will be found recorded in the recent work of the writer, on "Friction and Lost Work in Machinery and Mill Work." * One of the engines employed in this later investigation was built by the Lansing Iron Works, of Lansing, Michigan, an engine having a steam cylinder eight inches in diameter, and a stroke of piston of twelve inches. It was fitted with an "automatic gear," and was of the same class and very similar in many respects to the well-known Buckeye Engine, built at Salem, Ohio. The valve is balanced, and has an unusually quick and wide opening, giving steam in a very satisfactory manner. The alteration of speed was effected by changing the position of the balls of its governor. The brake worked well throughout the tests, and the counting, done by a hand counter, was thoroughly reliable. The following table exhibits the results obtained, both from this engine, and from a 12 by 18 inch engine also employed in the same research. The smaller engine had been in use about a year; the larger was new and had not left the shop. Another series of trials was made by the same observers, using a "Tandem Compound Engine," also built by the Lansing Iron Works, having cylinders 14 and 21 inches in diameter, and 20 inches stroke of piston. The two pistons were secured to one rod, and the cylinders were thus placed the one behind the other. Its usual speed was about 200 revolutions per minute, and it was non-condensing. The machine was new, and its friction therefore probably greater than it would have been later, by a considerable amount. The data obtained gave the following results: TABLE II. FRICTION WITH CHANGE OF SPEED. Tandem Compound Engine. Cylinders 14 x 20 and 21 x 20. NOTE. 3 The above test shows an increase of friction nearly proportional to the increase of speed, the equation of the curve would be The most extended range of speed was obtained with a small engine recently brought out by the Lansing Iron Works and called, from its inventor, the "Jarvis Engine." (Fig. 58.) It was one of the first of its class and a new engine. In this machine FIG. 58. the piston is fitted with a globe joint by means of which it is attached to the rod, which latter thus vibrates as it drives the crank by a direct connection with its pin. To permit this vibration, instead of having a fixed trunk secured on the piston as is usual, forming a "half trunk" engine, the front head is designed with a globe joint also, in which works a coned sleeve, within which the rod slides, and the whole swings backward and forward as the crank turns and the piston traverses the cylinder. This makes a novel but very practical arrangement, and, so far as experience yet indicates, a perfectly successful one. It has been |