92 



NA TURE 



[May 23, 189 = 



cylinder wall without causing condensation ; he found it in super- 

 heating. He constructed, in 1855, a super-heating apparatus in 

 the flues of the boiler at Logelbach, which still exists. The 

 experiments with super-heated steam were carried out between 

 1855 and 1856, and showed clearly the eft'ecliveness of the method 

 in reducing comlcnsation. Superheating came largely into use 

 in the years 1S60-70 in this country in marine engineering, the 

 practice having been introduced here by John Penn. In every 

 case in which it was used an economy of coal was realised. 

 Generally the economy amounted to from 15 per cent, to 20 

 per cent. It was ascertained that this was due strictly to 

 economy of steam, and not to the utilisation in the boiler of heat 

 previously wasted. But the use of super-heated steam in this 

 country was gradually abandoned, partly no doubt from some 

 practical dilhcullies, but chiefly, I believe, because practical 

 engineers had no clear idea w hy super-heating should produce so 

 large an economy, and they were not indisposed to abandon a 

 complication, the action of which they could not satisfactorily 

 explain to themselves. 



No possil)le improvenient of the steam engine, of which we 

 have any knowledge at this moment, offers anything like so great 

 a chance of important economy as the reintroduction of super- 

 heating, anil especially of super-heating to at least 100° or more 

 above the saturation temperature of the steam. I obtained in 

 Alsace on a very good 500 H.l'. compound mill engine with 

 jackets, and ever)' appliance for economical w orking, an economy 

 of 15 per cent. .Mr. Mair Rumley has fitted a super-heater to a 

 Bahcock boiler supplying a triple engine, and has obtained an 

 economy of 10 per cent. In both cases the economy is economy 

 of steam, and therefore is not due to any increase of Ijniler surface 

 or increase of efticiency in generating the steam. Lately I'rof 

 Schriiter, of Munich, has been experimenting with a small special 

 comiHiund condensing engine of only 60 I. H.!'., running at the 

 moderate piston speed of 380 feet per minute, and with the not 

 excessive boiler ]iressure of 165 pounds per square inch. The 

 H. I', cylinder is not jacketed. The L. I', is jacketed with receiver 

 .steam. In this case in a lube super-heater of a rather special 

 construction in the uptake of the boiler, the steam is superheated 

 to 670° I'., or nearly 300' above the .saturation temperature corre- 

 sponding to the pressure. In two trials of .six and eight hours' 

 duration, perimls quite long enough for accurate determination of 

 results with so accomplished an observer as Prof Schroler, the 

 consumption of steam was only io'2 pounds per I. H.l'. hour, and 

 the consumptirm of (ierman coal of moderate quality only \\ 

 |>ounds per I.H.P. hour. The steam consumption is the lowest 

 on record for any engine of any type or size, and is very remark- 

 able for .so smallan engine. 



Conflict of the Katioiial and Experimental Theories. — Zeiiner, 

 Hirny and Hallaner. 



On the 'ap|)earance of Isherwood's researches in 1863, the 

 discrepancy between the rational theory and the results of 

 experiments was recognised by Kankine and others. But the 

 conditions of cylinder condensjitiun are so complex, that for a long 

 time the more the<jretical writers i>raclically ignored Imth Ilirn's 

 and Isherwood's results. Zeuner, perhaps, had pushed the rational 

 theory to the furthest limit of detail, and with the greatest insight 

 into practical conditions. But it wa-s not till 18S1 that he began 

 to cxplicity admit the largeness and im|)ortance of the condensing 

 action in the cylinder. Zeuner then was disposed to attribute 

 initial condensation to the presence of a permanent antl not 

 inconsiderable mass <»f water in the clearance space of the engine. 

 No doubt it is simpler analytically to deal with the thernml 

 changes of the .sleam plus a given mass of water than with the 

 thermal changes of steam, water, and a varying area of solid 

 cylinder wall. In opening a discussion with I lirn in 1881 , Zeuner 

 wrote that if the presence of water in the clearance space was 

 conceded, the .Alsatian calculations would be greatly shaken, and 

 "the enormous influence which they attributed to the cylinder 

 wall would in future be attributed in part, perhaps chiefly, to the 

 water in the clearance space. " He thought it conceivable thai in 

 certain cases the whole of the initial condensation was due to 

 water in the clearance smce. There thus arose a rather angry 

 controversy, which has been summed up in the (pieslion, " Is it 

 water or iron ? " I «lo not know that this controversy has been 

 as yet completely ilecided, or that perhaps an absolute decision 

 is |)ossil>le. I cannot help thinking that Hirn, with the clearness 

 of view due to his experimental work, had on the whole the best 

 of the controversy, and I do not know that anything better or 

 more instructive can lie said than the words in which he finally 



NO. 1334, VOL. 52] 



summed up his position. " We recognise," he said, " that the 

 interpretation of the Alsatians differs from that of M. Zeuner, 

 not at all in that it denies the possible presence of water in the 

 cylinder (we are not so hydrophobic), but in that it admits that 

 the water, varying in quantity, is jiresenteil only temjK>rarily, is 

 carried away and reneweil stroke by stroke, and acts chiefly as 

 the medium between the sleam and the cylinder wall. In the 

 Als;ilian explanation the action of the water raises the thermal 

 action of the sides. In Prof Zcuner's view, the water is per- 

 manently present and acts indejKndently of the cylinder sides." 



Recent Systematic Experiments. — Willans. 



It has been quite impossible in this lecture to do more than 

 select one or two of the most important of the experimental in- 

 vestigations during the last fifty years. But I .should not like to 

 omit all reference to the two series of experiments ol the late Mr. 

 P. \V. Willans. Mr. Willans' work is no doubt well known to 

 all sleam engineers, and needs no delailed description. How- 

 ever purely pr.iclical the object Mr. Willans h.ad in view, his ex- 

 periments were made in the true spirit of scientific research. Nt> 

 trouble was loo much to secure accuracy to the last decimal, 

 no possible cause of error was .so trivial that its investigation 

 was reckoned unnecessary. .\ few ex|)erimeiUers, Isherwood, 

 (lately and Kletseh and others had m.ide experiments on a 

 methodical system, varying a single factor at a time. Willans 

 carried out the method of experiments in series on a scale which, 

 till he proved that it could be done, no one would have supposed 

 ])ossible. There is a series of non-conilensing and a series of 

 condensing trials ; in each there are trials of simple, compound, 

 and triple engines: and for each of these, again, trials with 

 initial ])re.ssure varied, with expansion varied, and with s|ieed 

 varied. The results, tabidated in the clearest way, form a quarry 

 of scientific data, but at ])resent, in the main, an unworkeil 

 quarry. Perhaps that statement will .seem surprising, and of 

 course I am expressing only my own view, for which I claim no 

 infallibility. What .Mr. Willans might have done had he lieen 

 .spared, it is impos.sible to s,ay. He hail the most active mind 

 and the wi<lesl experience devoted, perhaps, at any time to the 

 study of steam problems. 



Let me ])rolest as strongly as possible, .again willi the reserva- 

 tion that I am stating my personal view , .against the tendency to 

 suppose that the great work of Willans can be summed up in a 

 so-called Willans' law, or that that law, handy as it may be for 

 l)ractical steam engineers, is more than a quite suborilinate part 

 of Willans" work. The Willai\s" law is nothing more than the 

 empirical descriptive statement that the relation of total steam 

 [ consumption and indicated or effective horse-power can be very 

 approximately expressed by a linear equation, for the case of .an 

 unjacketed engine working with a fixed cut-oflT. I'urlher, 

 nothing is done in Willans papers to fix what is the linear 

 equation for any given engine. So far as those papers go, and 

 until some kind of theory taking account of initial condensation 

 is discovered, we can only finil the relation t>f steam consumption 

 and horse-power for any given engine by making two accurate 

 trials of the engine itself Willans" law leaves us in regard to 

 any given engine in the same position as an astronomer with a 

 new comet. When the comet has been observed for a sullicient 

 period, and some of its positions fixed, a probable orbit can be 

 calculale<l. The straight-line law leaves the steam consumption 

 of a new engine as unknown as the elli|)tic law the orbit of a new 

 comet. 



Willans himself says nothing whatever as to any po.ssible 

 rational basis for the Willans' law. He put it forward purely as 

 the result of plotting his experiments. Later, Captain .Sankey 

 showed that the total steam ctmsumiHion of an engine working 

 adiabatically with fixed ratio of expansion would also follow 

 nearly but not exactly a straight-line law if all clearance los.ses, 

 radiation, and exhaust waste and back ))ressure loss were 

 neglected. 



If we a.ssume isothermal expansion (and really so far as the 

 area of the diiigram is concerned, it matters little what law of ex- 

 pansion is assumed), it is easy to find a formula for the total 

 steam consumption of an engine working without clearance loss 

 or exhaust waste. I have found such a f^ormula, and plotted the 

 results both for a condensing and a nonrondensing engine in the 

 di.agram. It is found that the liiu-s plotted are not exactly, 

 but very nearly, straight lines. Thai carries us a certain way, 

 but it is an enormous jump to a.ssume wilhoiU examination that 

 the steam wastes in the engine, anuninting to from 20 to 50 per 

 i cent, of the steam used, and arising fri'iu causes of the most com- 



