288 Mr. Dugald Clerk [Jan. 29, 



take the time of maximum pressure as the measure of the time of 

 the explosion, but where differences are so ^reat as these experiments 

 show, it is more accurate to compare maximum temperature points. 



These experiments clearly showed that in the ordinary engine 

 cylinder the velocity of the flame through the mass depends very 

 largely on the residual turbulence due in the first instance to the 

 •charge velocity at entering. From this it would be expected that the 

 liigher the charge velocity became, the higher became the rate of 

 •ignition. Experiments show that this is the case. AVhere the 

 •charge velocity through the valve is high, the rate of ignition is 

 ■rapid ; and in petrol engines, where the charge velocity sometimes 

 rises as high as 150 feet per second, the flame velocity approaches 

 "90 feet per second. In the large gas engine Fig. 8, as has been 

 stated, the charge velocity is about 160 feet per second, but the rate 

 'of rotation of the engine is only IGO revolutions per minute. The 

 residual turbulence is accordingly less than it would be in the case 

 •of a petrol engine, where the rate of revolution varies from 1000 to 

 -2000 per minute. 



Prof. Hopkinson also made tests of turl^ulence by a series of 

 •closed vessel experiments. The vessel used was cylindrical, and about 

 'one foot diameter and one foot long. The cylindrical portion was 

 completely lined with a helix of copper strip, the turns being insulated 

 from one another. Means were provided for recording the resistance 

 of this strip during the progress of an explosion whence the rise 

 of temperature and therefore the rate of flow of heat into the copper 

 could be determined. A small fan like a screw-propeller was fitted 

 at the centre of the vessel, the axis of the fan being along the axis of 

 the cylinder. An optical indicator was fitted to the vessel. The 

 spindle of the fan passed through a gland in one of the end covers, 

 and it could be driven at any speed up to 5000 revolutions per 

 .minute, thus setting the gas in the vessel into motion. The general 

 •character of the motion would be an eddy, the gas passing along 

 ■the axis of the cylinder on the inside and then outside in contact with 

 the cylinder walls. Cambridge coal gas and air were used at atmos- 

 pheric temperature before firing. The mixture was fired by a spark 

 near the centre of the vessel. The cylinder was filled with a given 

 mixture of gas and air, and the fan was run in all cases for a few 

 minutes, in order to ensure the mixture of the gas and the air. 

 Diagrams were taken with two mixtures, one mixture containing 15 

 per cent of gas, the other 10 per cent. 



Figs. 18 and 19 show the diagrams obtained with these two 

 mixtures. 



Diagrams Figs. 20 and 21 are deduced from Figs. 18 and 19, and 

 give the relation between the rate of heat loss and the temperature of 

 the gas. Comparing the pressure curve A in Fig. 19 with the curves 

 B, C, D, it will be noticed that the efl'ect of the agitation of the gas 

 jat the moment of firing is to greatly reduce the time of explosion. 



^ 



