THE BRITISH OVERSEAS RAILWAYS HISTORICAL TRUST   A UK Registered Educational Charity

Kevin Jones' Steam Index

Journal of the Institution of Locomotive Engineers

Volume 12 (1922)

The IMechE virtual library is accessible (full papers, all diagrams, photographs, extensive tables, etc) via SAGE organization

Journal No. 52 (January 1922)

Great Northern Railway: new three-cylinder 4-6-2 express passenger engine. 4-6. illus., table
Table compares the leading dimensions of the A1 Pacific with the large Atlantics.

New "Baltic" tank locomotives, Glasgow and South Western Railway. 6-8.

Grassick, J.P. (Paper 114)
The locomotive from a footplate point of view. 51-67. Disc. 67-104; 311-28.
Presented in Glasgow on 25 November 1921: chaired R.H. Whitelegg.
Experience on NBR: spark arresters had to be removed. Problems with steel fireboxes. Notes on boiler fittings and the risk of developing leaks whilst in service. The fitment of injectors was considered to be a particular source of difficulty. Criticism of tail rods. Notes on problem of carbonisation in cylinders of superheated locomotives. Noted that comfortable cabs had been introduced by Adams on the LSWR and these had been followed by comfortable cabs on the NER and GER. Results from tests on GWR 28XX, T3 NER 3-cylinder 0-8-0 and NBR 0-6-0 on Glenfarg Bank in which the NER 0-8-0 put up a superb performance. Northumberland coal appeared to be superior to Scotch coal as former led to consumption of 0.099 lb/mile as against 0.118 lb/ ton mile. The water evapouration achieved also favoured Northumberlanbd coal (7.88) as against 6.49. On page 65 Grassick quoted lubricant consumption for NBR classes ranging from the Atlantics to the 0-4-0Ts and where fitted with superheaters the extra oil for cylinder lubrication..
Discussion: R.H. Whitelegg (67-71) noted Grassick's comments on spark arresters and agreed; furthermore he considered that it was vital to maintain a vacuum within the firebox and ensure that the door was tight-fitting.M. Blacklock (Beardmore 71-2) suggested propietory jointing materials such as Walkerite and copper joints to ease the problem of boiler mountings.Like Author queried the resaon tail rods. Cited the advantages of three and four cylinder designs: more even torque and better balancing. Questioned the most suitable location for snifting valves: in Glasgow the steam chest appeared to be the favoured location; critical of Grassick's calculation of tractive effort on the basis of 85% boiler pressure, suggested 80% or even 75% as being more realistic. Advocated smaller diameter driving wheels for 0-6-0. J.W. Tyler (GSWR 72-3) considered that locomotive trials should be conducted in winter under severe conditions.
R. Guest (NBL 73-9) contrasted American practice where steel fireboxes were the norm and where locomotives were worked to the absolute limit of their capacity aided by such devices as automatic firedoors. He considered spark arrestors and stated that Drummond's fireboxes as fitted on the LSWR with cross tubes acted as spark arrestors. He claimed that tail rods were useless. Noted that he had served as a fireman on the LSWR "many years ago", but that fuel consumption did not appear to have increased. On page 81 Grassick responded: spark arrestors: "I don't want them and don't believe in them at all" Considered that there was little to choose mechanical and sight feed lubricators. Very damning about steel fireboxes: at one NBR shed where there had been 19 firebox failures, 17 of these had been steel ones. Tail rods were no use.
"Surely if there is one item more than another which calls for revision of design it is the driving and coupled axleboxes of most of the locomotives of the United Kingdom. How is it that it is almost universal practice to use solid gunmetal boxes for this purpose, weighing, in some instances, not less than 5cwt. each? The London and North-Western and the North-Eastern are striking examples of a step in the right direction, both these companies using cast-steel boxes, the horn faces of the boxes in the case of the former being lined with white metal, and in the case of the latter with gunmetal liners pinned in; the horns of the North-Western engines are not fitted with adjustable wedges, and should a box he found to be knocking in the horns the engine is merely placed on the drop pit, the offending box or boxes removed, relined with white metal and replaced, simplicity itself, and an operation which can be carried through in the very shortest possible time, American practice is to use cast-steel axle boxes entirely, not even lined on the horn faces with either white or gun metal and with gun metal bearings forced in. by hydraulic pressure to save the expense of fitting; there may be some hidden virtue in a gunmetal box, but I would ,venture to say that to use such expensive luxuries, for luxuries they are, is more to the final advantage of the scrap metal. merchant than to the pockets of the railway shareholders, The same could be said regarding the use of gunmetal eccentric straps; what is the matter with cast-iron or caststeel straps? If it be thought by some that gun metal must be introduced in some shape or form at all costs, why not then line the cast steel or wrought iron strap with gunmetal, which gives all the advantages, if any, of the solid gunmetal strap without needlessly throwing away shareholders' money, Mr. Grassick says, and I am quite in agreemcnt with him, that too little encouragement is given to drivers and firemcn in Great Britain to take an intelligent interest in their work; while in the United States some years since I found, almost without exception, that every railway company there had its classes for locomotive employees, which had, perforce, to be regularly attended, and men only became engineers and firemen by attending these dasses and passing the examinatibns set for this purpose. I believe I am ,correct in saying that the late Mr. Dugald Drummond, while Chief Mechanical Engineer of the London and South Western Railway, introduced such classes and examinations with the very best results, and so great has been the success attending them that they are still continued by the present engineer.
Pp. 84 et seq: Manchester Meeting on 8 December 1921: J.W. Smith (Chairman) spoke about lubricants; W. Sinclair (NSR p. 88) described a staff suggestion scheme (document for which reproduced on pp. 97-8); S. Clayton (LYR, Darwen) pp 88-9 noted the need for simplicity and also commented on lubrication; R. Opie (GCR, Gorton) pp 89-90 considered that six-coupled locomotives were more versatile than eight-coupled as they could also be used for passenger work; D. Corbett Fletcher (GNR, Manchester) pp. 90-1 noted the performance of Foden steam wagons used between Manchester and Southampton and of petrol lorries between Manchester and Leicester; J.N. Gresham (LYR, Horwich) pp. 91-3 commented on injectors; F.W. Attock (LYR) pp 93-5: Leeds Meeting on 5 December 1921 pp 99 et seq: A.C. Stamer chaired Leeds meeting on 5 December 1921: comments 99-102: engine failures were very rare, structure limits were more significant than gauge; advovcated 20 ton capacity wagons; the use of assisting engines on severe gradients was an effective means; considered that the NER T3 had been superior to the Churchward 28XX due to the better turning momemnt provided by three cylinders and the total adheion available on the 0-8-0, also considered that the exhaust steam injector was very useful; G.A. Musgrave (GNR, Doncaster) was against the pooling of locomotives; A. Hird (Hunslet) requested information on carbonization in cylinders on superheated engines, especially where the Detroit displacement lubricator was used.;
Walter Chalmers chaired a second meeting in Glasgow and returned to the topic of the Glenfarg trials; Whitelegg noted the accessability provided by outside cylinders, the need to avoid cranked axles, and the advantages of the traffic control system introduced by the Midland Railway and adopted by the NBR. Kermack (318-19) observed the use of feed water heating on the LBSCR and noted the potential of exhaust steam injectors; McFarlane (319-21) considered that compressed air should be used to power auxiliaries; J. Aitken (NBR) 321-2 advocated improved training for footplate staff and noted that this was far more advanced in Canada; M. Blacklock showed the relationship between coal consumtion and tractive resistance.

Fowler, Sir Henry (Paper 115)
Steels for locomotive purposes. 106-27. Disc.: 127-33; 198-209; 259-64.
Presented in Glasgow on 16 December 1921: chaired R.H. Whitelegg.
The first steel purchase made by the Midland Railway was for spring steel in January 1861. This was followed in 1874 by purchases of steel for tyres and for straight and crank axles. In 1884 steel was adopted for frame plates, boiler plates and wheel centres. There were no specifications for steel until 1887. Sir Henry describes how testing and analysis were introduced. Note running title given as "steels for locomotive fireboxes" which might have led to some false citations. Page 110: mentions machine designed by Deeley for the torsion testing of crank axles. Cited contemporary work on steel and testing: Charpy, Izod, Fremont and the the work of Sir Charles Hadfield. The Midland Railway operated 3000 locomotives and the following table shows how the life of crank axles improved.

Discussion: D. Irvine Kempt (CR) (127-8) asked about the mileages achieved by crank axles; Walter Chalmers (NBR) (128) asked about built up crank axles; M. Blacklock (Beardsmore) (129-30); W. Chalmers (130-1) on steel fireboxes; Thos Lunt (NBL) (131) cracked frames; Wm Hain (CR) (131) the Izod impact test; H.A. Treadgold (131-2) the need for closer co-operation with steel manufacturers. In Manchester A.E. Kyffin (Gorton Foundry) (260) discussed steels for connecting rods and J.R. Billington (Horwich) mentioned low carbon steels for cranks. Fowler response on pp. 262-4.

Turner, F. (Paper No. 116)
Management difficulties. 135-50. Disc. 150-70.
Presented in London on 17 December 1921: chaired J.A. Hookham. The difficulties of costing, especially to make allowance for increasing wage rates and the intervention of shop stewards. Discussion: A.E. Howell (Armstrong, Whitworth, 153-4) shared the same problems and introduced time and motion study...
J. Clayton (pp 154-5): shop stewards should be educated men. H.A. Tonkin (SECR 168): "Locomotive engineering is in essence a problem of £ s. d., and of producing things at a price, and the function of costiong is the translation of technical results into monetary terms of measurement."

Journal No. 53 (February 1922)

Allen, C.J.  (Paper No.117)
Notes on the influence of design on express locomotive performance. 175-85. Disc. 185-97; 248-58; Meeting at Leeds 516-22 (Journal No. 55)
The Chairman (J.W. Smith) was hostile: "The author seems to have formed the opinion that locomotive engineers do not possess data obtained from actual tests to guide them on matters of design. They have quite a lot of information..."
Discussuion:
W. Rowland (G.C. Rly., Gorton 186) : In the early part of the Paper the Author, in commenting on the assumed impossibility of designing a locomotive by other means, says " We therefore have recourse to ratios," and proceeds to set forth various ratios, such as tractive effort to heating surface, heating surface to grate area, and so on, which he quite gratuitously assumes are actually employed by the designer. Although these ratios are apparently of great interest to the readers Q popular literature on the subject, they happen, with perhaps one exception, to be those with which any designer who knows his business is totally unconcerned. In fact, the Author, perhaps not intentionally, accuses the designer of total ignorance of the principles of his business and reduces him to slavish copying of proportions which have been found more or less satisfactory as the result of a course of errors or happy accidents in the past. Let me a5s~11-hei m that the state of the art is hardly as bad as that, whatever it may have been once. Certainly ratios do enter into design, but they are not those quoted by the Author.
For example, the ratio of firebox surface to volume, u hich is really a measure of the water-cooling effect of the fii -box sides, has a very important bearing on the completeness of combustion possible at high rate of firing, particularly when bituminous coal or oil is used as fuel. The ratio of firebox volume to fuel consumed per unit of time is also important. Then again, the ratio of hydraulic mean depth of a tube, or set of tubes, to its length is of importance as a measure of the heat-absorbing capacity of the tube or set. The ratio of the total resistance to gas flow of a set of tubes, which is deduced from the last quoted ratio, hydraulic mean depth to length, in conjunction with the number, is a measure of the power required from the blast, and is therefore of senice in conjunction with other considerations in designing a boiler to give the required power and efficiency. For high speed work, in particular, the ratio of mean port opening, on both steam and exhaust edges, to cylinder volume is likewise important; so is the ratio of the internal surface of cylinder ports and piston head to their voIume, as this ratio is largely concerned in determining the cylinder losses, and therefore the engine, as distinct from the boiler, efficiency of the locomotive. ‘The ratio of tractive effort to adhesive weight is about the only one of the .4uthor’s ratios that is of real importance, but even this ratio is by no means a constant for all engines, since it is by no means necessary to base it on the normal or full gear tractive effort for all classes of locomotive. In fact, a single driving wheel engine, designed with a nominal tractive effort to adhesion ratio suitable for a slow-running mineral engine, would be badly under-cylindered and most vneconomical at the express speeds for which such engines were designed, and to some extent the same holds good for four wheels and even six wheels coupled engines. May I assure the Author that modern locomotive design had progressed far beyond the rule-of-thumb methods which one would infer from his remarks to he still in vogue. The Author, further on, refers to the four-cylinder Great Western engines having a lap of 15/8. Perhaps he can also tell us the full gear valve travel which must be very much greater than is usual in British practice to prevent these engines constantly refusing to start. Is it a fact that these engines have a full gear valve travel of 7in. or so? If not, they must have a starting notch in the steam lap as is sometimes done in American engines designed to give an early cut-off in full gear. I have not so far heard of this expedient having been adopted in this country. It would have been of far greater interest to members of the Institution had the Author given records of long sustained high speed upgrade work with heavy loads, as this is the true measure of locomotive efficiency
E.W. Seilby (L. & N.W. Rly., Horwich: 188) : The Author refers to the ordinary every-day performance of a locomotive as opposed to special test runs. I think he is quite right, as there is extremely little published information about this subject. Some railways haVe test runs with a dynamometer car, but as a rule the engine is specially tuned up and picked coal is used. The results obtained are doubtless valuable for comparison with other similar test runs, but they do not give much idea of what those locomotives as a class may be expected to do day in, day out, in all v eathers, throughout the year.
As regards tractive effort, the two chief factors limiting the power of a locomotive are :-
1 . Adhesion (when starting and running slowly).
2.  Steam producing capacity of boiler (when running fast).
As the tractive effort formula does not take either of these factors into account, it seems rather futile and is often misleading.
Some reference has already been made to Great Western locomotives. These engines have quite exceptionally large steam and exhaust ports, and this point, coupled with their high boiler pressure, and with the fact that they have more lag on their valves than most locomotives, is largely responsible for the extraordinary turn of speed which these engines undoubtedly possess. The long lap enables the engines to be notched up to a very considerable degree without undue compr'ession, thus, whilst cutting off early, the exhaust port remains open until quite late in the stroke.
( J. Parry (G.C. Rly., Gorton 190) : The Author expressed the opinion that piston valves allowed a freer escape for exhaust steam than did slide valves. This is not necessarily so. With piston valves the steam leaving the cylinder has to break up into a number of streams in order to pass to the inside of the liner, where these streams converge and then suddenly have their direction changed through an angle of go degrees. Sudden changes of direction and velocity are as bad as throttling. Probably a balanced slide valve with the exhaust passage through the back offers the least resistance.
Parry did not think that the quoted opinion that harder rails and tyres reduced the undulatory motion of the road can be maintained. The coefficient of elasticity, Young’s modulus, of the harder steel is very little different to that of the softer, and therefore the rails, unless at the same time deeper, will not be stiffer. The rolling friction of hard materials is less than with soft; but, within the range of hardness permissible in railway practice, the difference is too small to be considered. The object of using harder steel is, of course, durability. The One of the comparisons of engines made in the Paper is between P. Stirling’s 8ft. single driver and Ivatt’s Atlantic, as regards their relative heating surface. In Stirling’s time coal was comparatively cheap, therefore fuel economy was not so much in the forefront as in later times. To get an engine low in first cost and maintenance, simple and reliable, it is quite permissible to sacrifice something in the way of fuel economy. Later, as in Ivatt’s time, large heating surfaces were in vogue, this being then considered the way to fuel economy. History often explains things. The Paper will serve at least one useful purpose, that of emphasising the fact that the design of locomotives, like the design of most other things, requires a knowledge of many matters other than the arithmetical rule of simple proportion.
J. Parry (G.C. Rly., Gorton, 189) : The Author expressed the opinion that piston valves allowed a freer escape for exhaust steam than did slide valves. This is not necessarily so. With piston valves the steam leaving the cylinder has to break up into a number of streams in order to pass to the inside of the liner, where these streams converge and then suddenly have their direction changed through an angle of go degrees. Sudden changes of direction and velocity are as bad as throttling. Probably a balanced slide valve with the exhaust passage through the back offers the least resistance.
I do not think that the quoted opinion that harder rails and tyres reduced the undulatory motion of the road can be maintained. The coefficient of elasticity, Young’s modulus, of the harder steel is very little different to that of the softer, and therefore the rails, unless at the same time deeper, will not be stiffer. The rolling friction of hard materials is less than with soft; but, within the range of hardness permissible in railway practice, the difference is too small to be considered. The object of using harder steel is, of course, durability.
The Author complains that railxvay companies do not publish reports of locomotive performances when apparently unfavourable. It would be very imprudent of them to do so, for these results getting into the hands of non-technical pe+ople might easily be the cause of much mischief. One of the comparisons of engines made in the Paper is beween . P. Stirling’s 8ft. single driver and Ivatt’s Atlantic, as regards their relative heating surface. Irr Stirling’s time coal was comparatively cheap, therefore fuel economy was not so much in the forefront as in later times. To get an engine low in first cost and maintenance, simple and reliable, it is quite permissible to sacrifice something in the way of fuel economy. Later, as in Ivatt’s time, large heating surfaces were in vogue, this being then considered the way to fuel economy. History often explains things.
The Paper will serve at least one useful purpose, that of emphasising the fact that the design of locomotives, like the design of most other things, requires a knowledge of many matters other than the arithmetical rule of simple proportion.
E.M. Gass (L. & N.W. Rly., Horwich 190-): The Author made no comments in respect to the loads hauled by the Stirling singles and the Ivatt Atlantics. There is little doubt that the running performances of the former have hardly ever been surpassed when hauling loads equal to their capacity, but the constantly increasing demands of the running department for heavier trains taxed these engines beyond their limit, the result being that a more pomerful machine became necessary.
I believe the success of the Great Northern singles was due, in a large measure, to the low ratio grate surface to heating surface, being 1 to 58 as compared with 1 to 80 in the case of the Atlantics. This low ratio results in a higher evaporation value; for instance, at a speed of 30 miles per hour the evaporation values are 8.31bs. and 6.25Ibs. respectively per pound of coal with a feed water temperature of 100°F. I have endeavoured to figure out the loads that the two designs are capable of hauling on the assumption that they are running at a speed of 30 miles per hour on,an up-grade of I in 200, this being about the ruling grade over the section of the line of the Great Northern Railway, Leeds to King's Cross and vice versa. The detailed statements A, B, C show how the figures are arrived at. The loads behind the drawbar in the case of the single driver is 283 tons as compared with 393. tons for the Atlantic. I have also estimated the rate of coal consumption. Assuming a speed of 70 miles per hour with a steam cut-off of 15 per cent., and cylinder clearance of 8 per cent., the Stirling engine, on this assumption, would be burning coal at th'c rate of 157 lbs. per square foot of grate surface per hour, whereas Ivatt's Atlantics would be consuming at the rate of 115 lbs. The rate of the former is considerably above economical limit, which is about 130 lbs.
With respect to high speed, I should like to mention a record of 94 miles per hour attained by one of Sir John A. F. Aspinall's Atlantics when hauling a train of four bogie coaches, practically 100 tons behind drawbar, between Rainford Junction and Kirby, on the Manchester and Liverpod line. This section of the road is on falling grades, the ruling being 1 in 235. The time occupied in passing the two points was 3¼ minutes, the distance being five miles 179 yards. This engine was fitted with Geo. Hughes' patent compression ball release balanced slide valves at the time.
The bad starting qualities of the single driver are well known. In the case of the Great Northern engine the weight available for adhesion was 18 tons, and only 25 per cent. of this weight could he utilised, or 10,0801bs. tractive force. This figure gives a mean effective pressure in the cylinders of 108lbs. or 61.7 per cent. of the boiler pressure. The corresponding piston speed at this pressure is 310ft. per minute," which in the case of the Great Northern single is equal to a running speed of nearly 20 miles per hour.
C.J. Allen was not present at the Leeds Meeting: A. Hird (Hunslet) pp. 516-17 considered that fuel economy had been ignored by Allen, and he had failed to address the problems of design for narrow gauge lines; J.J. Laine (NER, York) was critical of Allen's enthusiasm for the performance of the LYR 2-4-2Ts: these were only suitable for short-distance work with light loads. Noted that the NBR "Glenfarg" tests with NER 3-cylinder 0-8-0 had been reported in Rly Gaz.
E.L. Ahrons (pp. 519-22) The Author of this Paper has raised a question of great interest upon which much might be said. Since he quotes my views upon the effect of harder and stronger tyres, and harder rails of heavier section upon the speed of express locomotives, perhaps I may be allowed to explain these more fully. During the periods 1884-1892 and 1898-1904, I made a large number of observations on the speeds of express engines on many of the English main lines. During the period 1898-1902 I frequently noted that the maximum speeds attained by the older engines showed a very marked increase. The same engines which, during the period 1884-1892, had not within my own experience exceeded speeds of 70 to 76 m.p.h., were in 1898-1902 attaining speeds of 80 to 82 m.p.h. with considerably heavier trains. This was very marked in the case of the Midland 2-4-0 and 4-4-0 engines, and also to some extent on the Great Northern with the celebrated 8ft. single engines. Speeds of 75 m.p.h. before 1892 were extremely rare, and on three occasions only did I ever find this maximum to be exceeded. On one run a Great Northern 8ft. single attained 76.6 m.p.h. and on another occasion a Midland 4-4-0 engine with 7ft. wheels reached 76 m.p.h. Both instances occurred on descending gradients of 1 in 200 with trains of 120-140 tons weight.
During 1899-1901 I timed several Midland coupled engines with 6ft. 8in. and 6ft. 9in. driving wheels at speeds of 80 to 82 m.p.h. with trains of 160 to 210 tons. The. engines were the same as those which I had timed at speeds not exceeding 736 m.p.h. during the 1884-1892 period. No alterations had been made in the cylinders, wheels, dimensions of boilers, etc., and the pressure had not been increased above the original 140lbs. per sq. inch, but other changes had taken place. The tensile strength of the new tyres had been increased from about 37 to 40 tons per sq. inch to about 45 tons per sq. inch and over, and the material was harder. The old 85 1b. rails had been replaced by 100 lb. rails and these also were of harder steel. The moment of inertia of the 100  lb. rail section is of course considerably greater than that of the older 851b. section. From these changes in both rails and tyres, it resulted that there was not only much less bending of the rails between the chairs, but also that there was less crushing and flattening of tyres and rails at the line of contact on the tread. These two causes combined would reduce the resistance due to rolling friction. Incidentally, this tends to show that the old formula for train resistance, such as that of D. K. Clarke, are of very little use under modern conditions. Probably the formula derived by Sir John Aspinall as a result of his experiments, published in the Proc. Inst. Civil Engrs. 147-155, 1902 are the only ones upon which reliance can be placed to suit present day British practice. It appears to me that the success of the modern 4-6-0 locomotive as a fast express engine may be largely ascribed to the same cause. The 4-6-0 express engine did not exist in this country before 1901. It seems likely that had this type been employed in the 1880-1890 period with the softer tyres and softer and lighter rails then in use, it would not have been a success as regards high speeds, for the reason that the softer tyres of the six-coupled wheels would have worn too unevenly, and the resulting different diameters of the coupled wheels would have caused excessive frictional coupling rod resistance after the engines had been in service a comparatively short time. The present day hard D class tyre of 56 to 62 tons per sq. inch tensile strength has made all the difference in favour of the modern six-coupled express engine. Another point of the greatest importance in the design of a high speed express engine is the obtaining of a very free exhaust, in other words the back pressure horse-power should be the least possible. The back pressure horsepower may be defined as the horse-power absorbed in producing the draught and depends not only upon the dimensions, length and straightness of the cylinder ports and passages, but also upon the boiler dimensions and proportions, tube and grate areas, smokebox and blast pipe design and proportions.
The Midland 4-2-2 single express engines designed by S. W. Johnson were excellent engines in this respect, the indicator diagrams taken on express trains showing an average back pressure during exhaust of 3 to gilbs. per sq. inch only at high speeds, and their capabilities for fast running were well known.
On the other hand there are many saturated steam express engines at work which show an average back pressure during exhaust of more than 10 lbs. per sq. inch at speeds above 45 m.p.h. In the case of superheater engines of good design with piston valves the exhaust appears to get away more freely, with a considerable reduction of back pressure. Apart from the use of piston valves,. the use of superheated steam appears to show a great advantage in this respect. Short straight ports of ample area are essential in a high speed locomotive, but on the other hand the clearance space must not be excessive. Some designers in their anxiety to reduce the clearance spaces appear to have gone somewhat too far in the wrong direction and have throttled the exhaust to such an extent that the back pressure losses exceed the gain obtained from reduced clearance spaces. The amount of power consumed in driving the locomotive itself at high speeds is enormous. The following figures are taken from those given by H.A.  Ivatt Proc. Inst. Mech. Engrs., 1904) and relate to one of the Great Northern " Atlantic " engines, No. 251, working on saturated steam.