20 



NATURE 



[November 3, 1892 



The number of revolutions were also low, 14 "8 without jackets 

 and 1578 with jackets. This was evidently an engine which 

 should pay for jacketing. We next come to an experiment of 

 a different nature, carried out by Mr. Davey and Mr. W. B. 

 Bryan. The engine is triple expansion surface-condensing 

 engine of the inverted direct-acting marine type, and is placed 

 in ihe Waltham Abbey pumping station of the East London 

 Water Works. The cylinders were 18", 30 "5", and 51" in 

 diameter, by 36 inches stroke. There is a Meyer expansion 

 valve to the high pressure cylinder, by means of which the speed 

 of the engine was regulated during the experiment. The bodies 

 and both ends of all three cylinders are steam -jacketed. The 

 jacket steam of the high pressure cylinder is at full 

 boiler pressure, but the other two cylinders have the 

 pressure reduced to a little above that "of their steam-chests by 

 means of reducing valves. Each cylinder is therefore jacketed 

 with steam a little above its own initial pressure. Without the 

 jackets in use the amount of feed water per I. H. P. per hour 

 was I7'22 lbs, and with the jackets in use 15*45 lbs, showing a 

 percentage of less steam used owing to the jacket of io*3. The 

 total jacket water was 172 lbs. per I.H.P. per hour. The coal 

 consumption is given in these experiments, being 2-09 lbs. per 

 I.H.P. per hour without the jackets, and 1 79 lbs. with. The 

 amount of coal burnt is not, of course, necessarily a measure of 

 economy of the engine, but possibly the steam-generating plant 

 — which included an economizer — was practically constant in its 

 duty during both trials, and if so the commercial gain by the 

 use of the jacket is quite an appreciable quantity. The boiler 

 pressure here was 130 lbs. above atmosphere, the number of 

 expansions without the jacket 22, and with the jacket 30. The 

 revolutions were 23 per minute, so that the jacket had again a 

 favourable chance. 



The next series of experiments were carried out by Colonel 

 English, Mr. Davey, and Mr. Bryan Donkin, and in these we 

 reach a much higher piston speed, so that the results stand on a 

 somewhat different footing in this respect to those before quoted. 

 We have no positive knowledge of this engine beyond that given 

 in the report, but it would be desirable to know something 

 more of its working before accepting the very high percentage of 

 gain in steam used — 19-0 per cent. — as that due to a steam 

 jacket used on a good engine. The feed used per I.H.P. per 

 hour was 24*68 lbs. without the jacket, and with the jacket in 

 use the quantity was 20 lbs. The following are the particulars 

 of this trial : — Horizontal surface condensing compound engine, 

 with intermediate receiver cylinders, 18 and 32 ins. by 48 ins. 

 stroke. The ends of the cylinders are not jacketed, and the 

 receiver jacket was not in use during the experiment. The 

 boiler pressure was 50 lbs., the revolutions 57 '06 without jackets, 

 and 63 '62 with, the feed water supply as stated, and the jacket 

 water condensed per I.H.P. per hour i "13 lbs. The coal used 

 without jackets was 3*26 lbs. per I.H.P. per hour, and with 

 jackets 2"66 lbs. 



The last set of experiments we shall quote were made by Prof. 

 Unwin, upon the experimental engine ^ at the City and Guilds 

 of London Central Institution, South Kensington. It is a two- 

 cylinder horizontal-surface condensing engine, and can be worked 

 either simple or compound. The cylinders are 873 inch and 

 15-76 in diameter, by 22" stroke. The high pressure cylinder 

 is fitted with Hartnell expansion gear, and the low pressure with 

 Meyer expansion gear. Only the bodies and the back ends of 

 the cylinders are covered. We will first give results of trials 

 working the engine with the low pressure cylinder only. The 

 pressure was 60 lbs. above atmosphere, the jacket pressure being 

 taken direct from the boiler. The revolutions without the 

 jacket were II2'40, and with the jacket 10173. The feed 

 water per I.H.P. per hour without the jacket was 32"I4 

 lbs., and with the jacket 26*69 lbs. This gives a saving 

 of 17 per cent, working simple. It will be seen presently that 

 when the engine was working compound, the saving was 7*3 per 

 cent. The jacket-water per I.H.P. per hour was i'88 lbs. 

 We will now take the records of the compound trial. The boiler 

 pressure was 6673 lbs. without the cylinders, and 67*80 lbs. 

 with the jackets. The revolutions were 93*66 without the 

 jacket, and 96*11 with. The feed water used per I.H.P. 

 per hour was 21*06 lbs. without the jackets in use, and 19*52 

 lbs. with. The saving, as stated, made by the use of the jacket 



I This engine is stated in the report to have been fully illustrated and 

 described in Engineering of November i6, 1888. The triple expansion 

 engine at Waltham Abbey is also said to have been illustrated and described 

 in the issue of August 8, tSgo, of the same publication. 



was therefore 7*3 per cent. The jacket water used per 

 I.H.P. per hour was 2*40 lbs. We regret we are not able to 

 give all the interesting details which Prof. Unwin includes in 

 his instructive report, but for these we must refer our readers to 

 the original paper. 



Probably Prof. Unwin's 7*3 per cent, saving in steam used is 

 a far better measure of the value of the jacket than the inflated 

 promise of 19 per cent, in Major English's trial. It should be 

 remembered that the jacket is more effective in small than in 

 large engines, the area of cylinder will be in a higher ratio to the 

 contained steam in the former than in the latter case. The number 

 of expansions in the South Kensington engine working without 

 jackets was 7*23, and with jackets 9*29. The corresponding 

 figures in the case of the Woolwich engine were 9*4 and I2*6. 

 The boiler pressure with the Woolwich engine was, however, 

 16 to 17 lbs. higher than in the other case. The revolutions 

 were 57*06 and 63*62 respectively in the two trials at Woolwich, 

 whilst at South Kensington they were 9366 and 96*11. It 

 would have been instructive if the committee had had the 

 courage to attempt some balance of these figures, and then have 

 endeavoured to account for the large difference which we believe 

 would have remained still to be accounted for. 



The next experiments quoted comprise a series made by Mr. 

 Bryan Donkin, junr., at the works of his firm at Bermondsey. 

 Mr. Donkin's labours in this field are well known, and engin- 

 eering science is largely indebted to him for the contributions he 

 has made to its lore. One most valuable feature in connection 

 with these investigations is the means he has used to ascertain 

 the temperature of the walls of the cylinder at various distances 

 from the surface. In this lies the essence of the problem. If 

 the Jackets Committee would give us minute and trustworthy 

 information on this point we could evolve the rest from existing 

 data. If we do not quote Donkin's figures in full it is 

 partly because his experiments are not yet complete and partly 

 because they have been dealt with more fully in " another 

 place," namely, the Proceedings of a Society other than tha 

 with which we are now dealing.^ We may state, however, 

 that in one case when the steam in the jacket space was 298° 

 Fahr. the cylinder walls averaged 290° Fahr., whilst at 006 in. 

 from the piston the temperature of the cylinder wall was 284° 

 Fahr. These temperatures were ascertained by thermometers 

 placed in holes drilled in the cylinder. Other instances are 

 given, but the matter is far too interesting to deal with in a 

 cursory manner, such as a report of this nature alone warrants. 

 The difficulty that suggests itself is the fact that a thermometer 

 itself has a very appreciable thickness, and the record will be 

 but a mean of the temperature due to that thickness. It is pos- 

 sible that Mr. Donkin gets over this difficulty in some way. 

 Perhaps the thermo-couple as used by Le Chatelier might 

 afford a solution, although this [apparatus is not so useful 

 for recording small differences at low temperatures, being 

 rather adapted for such work as- hot blast stoves and other 

 metallurgical purposes. Mr. Bryan Donkin's experiments are 

 the most suggestive in the report, as might be anticipated. 

 Trials were made with steam at various rates of expansion to 

 determine the effects of the steam-jacket on the speed of engine 

 and temperature of the cylinder walls, and on superheating. 

 The engine used was a small one (6" x 8"), but it was specially 

 constructed and arranged for the work. We again repeat Mr. 

 Donkin's investigations are well worthy of the study of all 

 interested in these matters. 



The report concludes with a valuable appendix in the shape of 

 suggestions for the use of those desirous of experimenting in 

 this field. 



The discussion on this paper was of a protracted nature, but 

 was ijot of a kind altogether worthy of the leading mechanical 

 institution of the country. Mr. Morrison, of Hartlepool, made 

 the most weighty contribution amongst the speakers. He pointed 

 out the difficulty of maintaining a good circulation of steam in 

 the jacket — one of the most important points to which the 

 designer of jacketed engines should turn his attention — and 

 illustrated a simple method by which he had secured this end. 

 His arrangement consisted of a series of diaphragms, by means 

 of which the steam was made to take a devious course through 

 the jacket. Mr. Schonheyder pointed out a mistake the com- 

 mittee had made in placing an air-cock on the top of the jacket, 

 when it was required to draw off air from the steam. Of course, 

 this is one of those little slips which the wisest are apt to make, 

 for it would be absurd to suppose such authorities as those en- 

 ' See Proceedings Inst. Civil Engineers. 



NO. i2or. VOL. 47] 



