3^4 



NATURE 



[September 8, 1923 



Gaseous Combustion at High Pressures.' 



By Prof. W. A. Bone, F.R.S. 



Introduction. 



IN the course of the researches upon gaseous com- 

 bustion which for many years past have been 

 carried out in my laboratories, it became necessary to 

 study the subject under much higher pressures than 

 those heretofore employed. As this aspect of the work 

 has recently assumed greater importance from the 

 point of view of the mechanism of combustion than 

 was at one time foreseen, an outline of it may be of 

 interest. Before, however, explaining what our new 

 observations have been, something should be said 

 about the apparatus and methods employed for such 

 work. For they must obviously differ from those used 

 for experiments at atmospheric pressure, where the 

 conditions are much less severe. 



In the first place, the experiments must be carried 

 out in specially designed bombs of forged steel capable 

 of withstanding the sudden development of very high 

 explosion pressures. Thus, in our recent experiments, 

 the initial pressure at which the combustible mixtures 

 were fired ranged up to loo atmospheres ; and the 

 resulting pressures, which were developed in a small 

 fraction of a second, were anything up to ten times as 

 great. Hence the method of measuring and recording 

 the pressures must be capable of following accurately, 

 and with the least possible lag, a rise of pressure of 

 from (say) loo to looo atmospheres occurring within 

 ., J^^th of a second. For this purpose we have employed 

 a recording manometer of the form designed by Sir J. E. 

 Petavel, which is a most efficient appliance for high- 

 pressure explosion work.^ 



The photographic pressure-time records obtained 

 in our experiments show (i) the rate at which the 

 potential energy of the explosive mixture fired is 

 transferred into kinetic (i.e. pressure or temperature) 

 energy of the products ; (2) the ratio of the maximum 

 pressure attained on explosion to the initial pressure 

 at which the mixture was fired — usually denoted as 

 P,„/P, ; and (3) the rate of the subsequent cooling. 

 From a study of these and other features of the 

 records we are able to draw conclusions as to cer- 

 tain fundamental aspects of the combustion process 

 itself. 



Some Features of the Combustion of Hydrogen 

 AND OF Carbon Monoxide in Air. 



As an example of the potentiality of high-pressure 

 explosion research to reveal and elucidate new factors 

 in gaseous combustion, I propose to deal mainly with 

 the cases of hydrogen and carbon monoxide. For 

 although at first they may seem to be of the simplest 

 type, yet they present features of extraordinary interest 

 and complexity which for many years past chemists 

 have vainly tried to explain. Even engineers, who 

 study internal combustion problems in their own 

 way, without troubling themselves overmuch with 

 the mechanism of the chemical changes involved, are 



' Fiom a discourse delivered at the Royal Institution on Friday, May ii. 

 • A full description of the bomb and^accessory appliances^will be found 

 in PhiL Trans. Roy. Soc, A 215 (I9i5),.pp. 275-318. 



NO. 2810, VOL. 112] 



seeking light upon what is termed the "suppretsum 

 of heat " in such explosions. Indeed our present 

 ignorance about these matters shows how far we are 

 from really understanding the elements of gaseous 

 combustion, and the need there is of much further 

 fundamental research thereon. 



From a chemical point of Nnew there has 

 been something enigmatical about the ver)' different 

 behaviours of the two simplest combustible gases, 

 hydrogen and carbon monoxide, when burning in air. 

 For although their volumetric heats of combustion 

 (assuming the initial and final temperatures being both 

 about 15° C.) and the proportion by volume in which 

 each of them combines with oxygen are the same, 

 namely : 



2H, +0,=2H-0 

 2CO + 0,=2CO, 



68'0 } ^•^•^' P*' gram-molecule, 



yet in many respects their modes of combustion in air 

 present a striking contrast. 



Thus, for example, (i) the appearance of a flame of 

 hydrogen in air is ver\' different from the lambent blue 

 flame of carbon monoxide burning at the same orifice 

 and under the same pressure ; (2) hydrogen-air mix- 

 tures have lower ignition temperatures, and, under 

 similar physical conditions, propagate flame much 

 faster than the corresponding carbon monoxide-air 

 mixtures ; (3) the presence of even a minute quantity 

 of steam greatly assists, if it is not absolutely essential 

 to, the oxidation of carbon monoxide in flames, even 

 when detonation is set up — thus a flame of the dry 

 gas is easily extinguished on being introduced into a 

 jar of air that has been previously dried over strong 

 sulphuric acid ; (4) a flame of carbon monoxide burn- 

 ing in air loses by radiation nearly 2*4 times as much 

 energy as a hydrogen flame of the same size ; also 

 (5) the two radiations have their own characteristic 

 wave-lengths — namely, 2*8 /x from a carbon monoxide- 

 air flame and 4*4 \i from a hydrogen-air flame — which 

 have been attributed to vibrational conditions in in- 

 cipiently formed COj and HOj molecules respectively, 

 or, as I prefer to say. to the formation at the moment 

 oj combustion of intensely vibrating carbon monoxide- 

 oxygen and hydrogen-oxygen complexes, which ulti- 

 mately give rise to carbon dioxide and steam molecules 

 respectively. 



To summarise : carbon monoxide bums in air more 

 slowly and with a more highly radiating flame than 

 does hydrogen ; also apparently the presence of some 

 steam or other hydrogen-containing substance is neces- 

 sary for its combustion. Precisely how steam acceler- 

 ates or determines the combustion of carbon monoxide 

 (and only a minute quantity suffices) has up to now 

 never been completely explained ; but chemists are 

 generally agreed that carbon monoxide molecules are 

 particularly inert towards oxygen molecules in flames. 

 Indeed I think there are grounds for believing that in 

 ordinary flames carbon monoxide cannot react with 

 undissociated oxygen molecules, but that it requires the 

 presence of either : O atoms or " activated steam " 

 : 0H« molecules. 



