[ 84 ] 



III. On the Measurement of Gas-Engine Temperatures. 

 By Bektram Hopkinson, Professor of Mechanism and 

 Applied Mechanics in the University of Cambridge*. 



[Plate VI.] 



A K important point in the experimental study of the gas* 

 .iTjL engine is the determination of the temperature of the 

 gases at each point of the cycle. It is obvious that if the mean 

 temperature at any one point is known, that at any other can 

 be calculated from the indicator diagram, assuming the relation 



L--— constant to hold throughout. This relation is probably 



true as a practical approximation for the gas-engine mixture 

 since it consists, as to 80 per cent, of its volume, of nitrogen 

 and oxygen, which pass unchanged through the combustion.' 

 For the starting point of the temperature determinations it is 

 most convenient to choose the temperature at the end of the 

 suction-stroke, when the charge of gas has just been drawn in 

 and its compression is about to commence. The measurement 

 of this temperature, or " suction temperature " as it is generally 

 called, is best effected by determining the quantity of air and 

 gas drawn into the engine at each stroke, from which, 

 knowing the pressure and volume of the mixed gases when 

 the inlet-valves have closed, their temperature can be 

 calculated. The accurate measurement of the large volumes 

 of air used by a gas-engine of considerable size is, however, 

 a difficult operation ; and it is desirable, if possible, to use 

 some means of directly measuring the temperature of the 

 charge at the end of the suction-stroke. Moreover, the sub- 

 sequent calculations from the indicator diagram of tho tem- 

 perature at other points of the cycle, lead to mean values for 

 the whole volume, and give us no information as to the differ- 

 ence of temperature between one part of the cylinder and 

 another. Such differences may amount to several hundred 

 degrees centigrade during the first half of the expansion or 

 working stroke. They arise partly from the cooling effect 

 of the walls, and partly from the manner in which the in- 

 flammable mixture is ignited. As I have shown elsewhere f, 

 the temperature round about the point of ignition is neces- 

 sarily some 500° higher than at a distance, because of the- 

 relative slowness of the propagation of flame as compared 

 with the time of reaction of the gases when the flame has 



* Communicated by the Author. 



+ Proc. Ivov. Soc. A. vol. lxxvii. p. 387 ; Engineering-, vol. IxxxL 

 p. 777. 



