September 8, 1923] 



NA TURE 



;65 



|»<>o 



High-pressure Experiments. 



Bearing the foregoing considerations in mind, let us 

 now see what new light has been shed on the problem 

 as the result of high-pressure combustion research. 

 Here it should be pointed out that, inasmuch as the 

 chief difference between the condition of high- and low- 

 pressure experiments lies in the absolute concentration 

 of the interacting molecules, it may be expected that 

 factors the operation of which chiefly depends on such 

 concentration will become more dominant as the pres- 

 sure arises. Indeed, the value of high-pressure work 

 lies in the fact that it tends to show up and accentuate 

 the operation of factors the influence of which may be 

 either masked or overlooked at ordinary pressures. 



One of the first things disclosed by our experiments 

 was the absence of any direct relation between the 

 rate at which the potential energy of an explosive 

 mixture is transferred on explosion to its products as 

 sensible heat (pressure) and the magnitude of the 

 chemical affinity between its combining constituents. 

 Thus, for example, the time required for the attain- 

 ment of maximum pressure on exploding 

 at 50 atmospheres a methane-air mixture 

 (CII4 + O2 + 4N2), in which the combus- 

 tible gas and oxygen are present in equi- 

 molecular proportions {i.e. corresponding 

 to the primary chemical interaction in the 

 flame), was many times longer than that 

 required in the case of the corresponding 

 hydrogen-air mixture (2H2 + O2 -I- 4N2), 

 notwithstanding the fact that the affinity 

 of methane is at least twenty, and pos- 

 sibly as many as thirty, times as great as 

 that of hydrogen for oxygen in flames. In 

 other words, the avidity with which a 

 combustible gas seizes upon oxygen in 

 flame combustion is not necessarily the 

 factor which mainly determines the rate 

 at which the potential energy of the mix- 

 ture is transferred into kinetic energy of 

 its products. 



Later experiments have chiefly dealt with the ex- 

 plosion usually at an initial pressure of 50 atmospheres 

 of what may be termed isothermic mixtures of either 

 carbon monoxide or hydrogen with sufficient oxygen 

 for complete combustion plus some variable diluent 

 developing as nearly as may be the same amount of 

 energy on combustion. I will now endeavour to 

 explain their significance. 



Tin: ("oMKAST between Carbon Monoxide-air 

 AM) 11vi)Rogen-air Pressure (Curves. 



W I Ilia', .ippropriately begin with n 1 niisidcration 

 of two t\j)iral pressure-time records (I'l-. 1) ohtained 

 when normal carbon monoxide-air .iiid hydio^en-air 

 mixtures {2{'()+02 + 4N2 and 2!!^ tO^ t 4N,,) were 

 respectively tired in the boml) at an initial ])ressure of 

 50 atmospheres. 



\'(i\\ altli'iii li tli(~,( tuM mixtures dexcloped as 

 III 'ital amount ol < m r.' \ on 



c\|Mw,r -;,. " ■ '-M-t-) ,t lu'twciii the 



charaitir ni ib uhtained. For 



whereas in the iNpnai iiydin-m aii' cnrve tile J)ressure 



NO. 2810, VOL. I I2j 



rose with extreme rapidity (actually in 0*005 second) 

 to its maximum (about 400 atmospheres), and almost 

 immediately thereafter began to fall and assume the 

 character of a simple cooling curve, in the correspond- 

 ing carbon monoxide curve the pressure rose much 

 more slowly and only attained a maximum (about 

 410 atmospheres) after o"i8 second, after which it 

 was maintained almost at its maximum for a con- 

 siderable time interval. The comparative slowness 

 with which pressure energy is developed in such a 

 carbon monoxide-air explosion, together with a con- 

 siderable exothermic effect after the maximum pres- 

 sure had been reached, were indeed very remarkable 

 and significant features of our experiments. At first 

 we were inclined to attribute them to the supposed 

 " slow-burning " property of carbon monoxide as com- 

 pared with the " quick-burning " of hydrogen ; but 

 further experiments revealed the operation of another 

 totally unexpected factor — namely, the presence of 

 nitrogen, which, as we discovered later, is not inert but 

 acts as an " energy-absorber " in the combustion of 

 carbon monoxide at such pressures. 



TiHt in '/loo SrCoffOS 



Fig. I.— 



TiAfc in 'lioo €teonos 



Pressure-time records for the explosion of carbon monoxide-air and 

 hydrogen-air mixtures. 



Effect of Addition of Hydrogen upon the Carbon 

 Monoxide-air Curve and upon a Carbon 

 Monoxide Flame burning in Air. 



It was next discovered that the replacement, even 

 in very small proportions, of carbon monoxide by its 

 equivalent of hydrogen in the aforesaid normal carbon 

 monoxide-air mixture had a disproportionately large 

 influence in accelerating the rise of pressure on ex- 

 plosion. This remarkable result, which is of con- 

 siderable theoretical import, was dealt with at length 

 in a paper published two years ago by the late W. A. 

 Haward and myself in the Proceedings of the Royal 

 Society.^ Indeed at first sight it seemed as if the 

 hydrogen had imposed its own character upon the 

 whole course of the carbon monoxide combustion, even 

 when the combustible part of the mixture exploded 

 contained only one part of hydrogen to twenty-three 

 parts of carbon monoxide by volume. 



In this connexion it may be mentioned that the 

 addition of a moderate amount of hydrogen to carbon 

 monoxide burning in air at ordinary pressure has a 

 consideral)le effect 1 ' t'rumof 



> Proc. Koy. Sor, 

 the cm 11 



,.il I't the Clicmi<Ml .Sdcicty. 



paper in 



