108 



GAS-ENGINE 



possessing much originality is Atkinson's, the dis- 

 tinctive features of which are shown in fig. 4. Here 

 the piston acts on the crank-shaft not directly but 

 through a toggle-joint, which has the effect of com- 

 pelling the piston to make four single strokes for 



Fig. 4. Atkinson's Gas-engine. 



one revolution of the shaft. The four strokes are 

 of different lengths. In the first forward stroke the 

 piston starts from the back end of the cylinder and 

 draws in gas and air. Returning it makes a shorter 

 stroke, compressing the mixture into a space not 

 swept through. Then the mixture is fired, and 

 work is done during another and considerably 

 longer forward stroke, and finally the cycle is com- 

 pleted by a return stroke, which is long enough to 

 completely expel the burned gases. The mixture is 

 ignited by means of a red-hot tube, but in tins case 

 there is no valve to control the time of firing ; it is 

 determined simply by the compression of the explo- 

 sive mixture against a cushion of waste gas in the 

 top of the tube. Fig. 5 is an indicator-diagram from 

 Atkinson's engine. AB is the admission stroke. 

 From B to C the explosive mixture is compressed ; 

 at C it is fired, and the effective working stroke, 

 CDE, begins. Its length is more than twice that of 



Fig. 5. Indicator-diagram of Atkinson's Engine. 



the compression stroke. In the long return stroke, 

 EA, the products of combustion are wholly expelled, 

 except for a small quantity contained in the clear- 

 ance space, which is no greater than the clearance 

 necessarily left behind any piston. This complete 

 (or, to be more exact, nearly complete) expulsion 

 of the burned gases is a good feature in Atkinson's 

 cycle, but the most distinctive merit is the relatively 

 long working stroke, which secures much expansion, 

 so that the gases do not escape until their pressure 

 falls to a value not greatly exceeding that of the 

 atmosphere, and at the same time makes the 

 expansion occur quickly, giving the hot gases com- 

 paratively little time to part with their heat to the 

 lining of the cylinder. 



Messrs Crossley have lately introduced a modified 

 form of Otto engine, with two equal cylinders, the 



pistons of which make their strokes simultaneously. 

 The mixture is compressed, exploded, and expanded 

 first behind one piston ; then the products of com- 

 bustion are allowed to pass to the front end of both 

 cylinders, driving back both pistons, and under- 

 going further expansion. Mean- 

 while the other cylinder has taken 

 in a fresh charge, which is now 

 compressed behind its piston, and 

 is exploded when the next forward 

 stroke begins. 



During the explosion in a gas- 

 engine cylinder the highest value of 

 the pressure is usually from 180 to 

 200 Ib. per square inch, and the 

 highest temperature is about 3000 

 F. The process of explosion is by 

 no means instantaneous. After 

 ignition the pressure and tempera- 

 ture rise with great rapidity, as the 

 indicator-diagrams (figs. 3 and 5) 

 show, but combustion is not com- 

 plete when the highest point in the 

 diagram has been reached. Only 

 about 60 per cent, of the whole heat 

 which the combustion of the gas 

 should yield is 'developed up to that 

 point. During the subsequent ex- 

 pansion a slow process of continued 

 combustion goes on, in which a 

 considerable part of the remaining 

 40 per cent, is set free ; but even when the con- 

 tents of the cylinder escape to the exhaust the 

 process is generally still incomplete. The after- 

 burning, as it is called, which occurs during 

 expansion, after the point of highest pressure 

 has been passed, has the effect of keeping the 

 pressure of the expanding gas from falling so 

 fast as it otherwise would fall. But for this the 

 expansion curve on the indicator-diagram would 

 fall very rapidly, owing to the cooling of the gases 

 through their contact with the cylinder walls. 

 During expansion the gases are parting with much 

 heal to the walls, but the after-burning supplies 

 nearly enough additional heat to make good this 

 loss sometimes, indeed, more than enough and 

 the result is that the form of the expansion curve 

 does not differ very materially from that of an 

 adiabatic line. The experiments of Mr Dugald 

 Clerk, who has taken much pains to investigate 

 this action, show that the time-rate of the explo- 

 sion depends greatly on the richness of the explosive 

 mixture. When the mixture is much diluted the 

 process is so slow that the point of highest pressure 

 is not reached until far on in the stroke. 



Though the maximum temperature within the 

 cylinder is materially reduced by this want of per- 

 fect suddenness in the combustion of the gas, it is 

 still so high that in engines of even very moderate 

 size a water-jacket is essential. The actual maxi- 

 mum temperature of the gases is in fact higher than 

 the melting- point of cast-iron, while the temperature 

 of the metal has to be kept low enough not to burn 

 oil. The water- j acket involves an immense waste of 

 heat. In the most favourable cases it absorbs 27 per 

 cent, of the whole heat which would be produced by 

 complete combustion of the gaseous mixture, and 

 more generally the amount it absorbs ranges from 

 40 to 50 per cent. The best existing gas-engines 

 succeed in converting into work about 22 per cent, 

 of the whole potential energy of the fuel ; of the 

 remaining 78 per cent, a half or more generally goes 

 to heat the water which circulates in the jacket, and 

 the remainder is rejected in the exhaust, partly 

 through incomplete combustion, but mainly in the 

 form of actual heat, on account of the high tem- 

 perature at which the waste gases escape. At- 

 tempts have been made to save a part of this loss 



