LMS locomotive design features analysed

[Compound2632]

Moving this discussion away from the preservation topic, to give parties free rein:

 

LMS2689 wrote:

 

The subject of Midland axleboxes is complex, it must be admitted, but to give an indication of the problem, here goes.

The Midland boxes were undersized. This was deliberate policy as it allowed some frame flexibility to deal with tight curvature; they were on the small size as regards diameter, but also as to length. This was fine in Midland days when loads, and therefore piston thrusts, were severely limited; excess loads were dealt with by adding an assistant loco. The boxes therefore were not stressed to any great extent, and hot boxes were not common.

While this was fine before 1923, from then on the 4Fs were spread over the enlarged system, where the strict loading restrictions did not apply. Hot boxes then began to appear in sufficiently large numbers to cause concern. Remedial action wasn't possible on any economic basis: enlarging the boxes would require major frame modifications around the horn gaps, and the inside eccentrics prevented increasing their length.

To be fair to the 4Fs, they did a lot of good work over many decades; their boxes were at worst marginal, and the locos were not so much bad as not particularly good. They had several other issues besides the boxes: poor steaming and sheared crankpins being on the list.

The problem was that these self same boxes were applied to bigger engines with much higher piston thrusts: the Garratts and Austin Sevens, and Derby must be held responsible for this. Here, their size made them far less than marginal but inadequate. The Austin Sevens appeared in 1929. Goods engines generally have long lives, but in 1949 61 of them, over a third of the class, were withdrawn, with the same number disappearing over the next two years. The remainder were transferred to former L&YR lines, where shorter runs allowed time for the boxes to cool.

If you want contemporary accounts, try a memo from E Stuart Cox to S J. Symes, P.A. to the CME, 09-06-31:

"0-8-0 Standard Type 7

Hot Boxes. The right driving boxes on these engines have been very troublesome during 1930. In some cases it was found that side thrust had caused the heating and this may have been due to inaccuracies in the machining of the box. Some improvement has been obtained by fitting end thrust pads as is done with the No.4 class Goods, which are also fitted with these boxes. Figures for 1930 out of 120 engines are: Driving Boxes 45; Others 8. The driving journals, which are only 8½" wide as against 9" on the G2 class, appear to be somewhat overloaded."

Some of the engines were less than two years old at this time, and in fact some were yet to enter traffic.

 

Compound2632 wrote:

 

The story can be supplemented from Adrian Tester's well-researched book on the subject - call it revisionist if you will. He identifies changes in early LMS days in the composition of the whitemetal used for the bearings and in the type of oil being used as lubricant, compared to the materials in use in Midland days and for which the axleboxes were designed, as significant contributory factors to the poor performance experienced. He also points to the improvements in both the finishing of the bearings and the lubricant resulting in a much lower incidence of hot boxes by the post-War period, for engines being worked just as hard, if not harder, than in the 20s and 30s. Clearly there was a major problem, as Cox reports, which was in fact successfully addressed; it is indeed probably the case that, for the 4F, the design was marginal making it susceptible to the changes in whitemetal and oil. 

 

John_Miles wrote:

 

Hot boxes only started to be a problem on 4Fs from around 1930 onwards when the LMS tried to save money by using a cheaper oil. Even then the incidence of hot boxes on 4Fs and other Midland locos was much better than on many other LMS engines. For example Super Ds were noticeably worse but the myth of the Derby axlebox (apparently started by E.S. Cox - ex-L&Y) is so well established that it has become  the accepted history. Fortunately the LMS kept fairly good record (Stamp liked statistics) and so the facts are available for those who don't like fake news.

 

LMS2968 wrote:

 

...and the Garratts and Austin 7s?

I find the cheap oil story difficult to digest. A hot box took an engine out of traffic and potentially blocked the running line to following traffic. The engine would be stopped - so not earning revenue - for some time, and the damage might involve a journal's being turned - a works job - or an axle replaced if the scoring took it below scrapping size. The potential costs were very high, probably more than any savings against quality oil. I doubt Josiah Stamp would condone that.

What evidence do you have to support this theory? 

 

John_Miles wrote:

 

The Austin 7s like the Super Ds suffered from being powerful engines with inside cylinders which limited the size of the axlebox which could be used. The Garratts - put that down to stupidity. Incidentally, when Stanier came to the LMS some 4Fs were fitted with GWR style axleboxes - which were found to be no better than the Midland boxes.

 

LMS2968 wrote:

 

No, we're back to size again. The Stanier boxes were no bigger than Midland 4F ones; they couldn't be without major modifications. But these could have been built into the Austin Sevens, at least as far as diameter goes. This use of the same box could also be put down to stupidity.

Did you read E. Stuart Cox's memo? He mentions the Ds only as to the bearing size, he makes no mention of their suffering hot boxes, even in comparison. 

 

Comment:

I think those are the substantive posts on the axlebox question. Here's to the continuation of informed debate!

 

 

 

[LMS2968]

Thanks, mate, for moving it. I was a teacher in a former life, and so very prone to being sidetracked!

 

I'm actually interested in this but need some evidence, and a simple statement that the oil formula was changed or that records exist doesn't provide it. I hope it's obvious that I do look at railways in general and the LMS in particular, and if anyone has information - supported by evidence - which increases my knowledge, I really do want to see it.

[Dava]

The SDJR 2-8-0 class were Derby designed and built engines though much more advanced than any freight or mixed traffic locos built by the Midland for its own use. As I understand it they had the same axleboxes as the 4Fs. In their working lives they do not seem to have been prone to hot axleboxes operating on the severely graded S&D. The two preserved locos have clocked up many more miles over nearly 40 years than the remaining 4Fs, yet again have not suffered undue axlebox problems. Does this reinforce the view that the axleboxes design was not the problem?

[rodent279]

The SDJR 2-8-0 class were Derby designed and built engines though much more advanced than any freight or mixed traffic locos built by the Midland for its own use. As I understand it they had the same axleboxes as the 4Fs. In their working lives they do not seem to have been prone to hot axleboxes operating on the severely graded S&D. The two preserved locos have clocked up many more miles over nearly 40 years than the remaining 4Fs, yet again have not suffered undue axlebox problems. Does this reinforce the view that the axleboxes design was not the problem?

I think it was the 2-8-0's I was thinking of when I referred to the Crabs in the earlier thread.

Apologies to any Crabs reading this!

[Compound2632]

Apologies to any Crabs reading this!

 

I have assumed that LMS2968 has some connection with the Stanier, not Hughes, mogul currently on the Severn Valley Railway.

[LMS2968]

Yes, I was one of four founding members of the Stanier Mogul Fund way back in 1970, and am currently its Archivist and general dogsbody. I'm also Archivist to the Stanier 8F Locomotive Society, owners of the 8F on the SVR.

 

As to the S&D 2-8-0s record on hot boxes, I've no information, but I would query 'much more advanced' over the Midland based goods engines. They were bigger - a bonus in itself - and the works were on the outside, but the boiler was that from the Compound and the valve gear was the usual Derby short travel short lead affair, unlike the advanced gear of the Austin Sevens. It was their greater piston thrusts which knocked hell out of the boxes.

 

When considering a bigger goods engine for LMS use, trials were held over the Midland main line to Toton between a Suiper D and a S&D 2-8-0. The D won, which was not in the scrip. The 2-8-0 was whisked off to Derby to have its valves reset and the trial repeated - with the same result. This is why the Austin Seven was an 0-8-0 and not a 2-8-0. The S&D 2-8-0s were good machines ON THEIR OWN LINES, but they didn't travel well.

[Compound2632]

The S&D 2-8-0s were good machines ON THEIR OWN LINES, but they didn't travel well.

 

I think that sums up the situation with pre-Grouping design generally and is the fundamental reason why the LMS's early attempts at standardisation ran into some difficulties. Of course there were some lame ducks among the pre-Grouping engines and some places were the standard designs - especially the 2P and 4P 4-4-0s in SW Scotland - were successes, especially with the locomotive men (a prerequisite for success generally).

 

My chemistry teacher at school many, many years ago - early 80s - was a member of the 8F Soc.

[John_Miles]

Thanks, mate, for moving it. I was a teacher in a former life, and so very prone to being sidetracked!

 

I'm actually interested in this but need some evidence, and a simple statement that the oil formula was changed or that records exist doesn't provide it. I hope it's obvious that I do look at railways in general and the LMS in particular, and if anyone has information - supported by evidence - which increases my knowledge, I really do want to see it.

 

The evidence for the oil being changed is in Adrian Tester's book, which is a heavy read. You need to have a grounding in engineering to get very far with it and preferably Mechanical Engineering with a good knowledge of fracture mechanics (he deals with frame fractures as well) and tribology. Basically what happened was that the original oil used was of organic origin (IIRC it was an animal product) and this is "sticky" i.e. it smears and so sticks to the bearing. The new oil was a mineral oil which lacked this stickiness and so led to a significant increase in hot boxes, especially in engines with inside cylinders where the valve gear and motion restricted the room for axle boxes. The solution came from the LNER who had similar problems and devised a blend of oils which led to fewer hot boxes.

 

The lubrication of axle boxes is interesting. Because there is no way of containing the oil in the bearing it just leaks out so the idea of the box being fully lubricated is erroneous, unlike for example a car crank shaft where the oil is circulated under pressure. So the stickiness of the lubricant is important because this smears around the bearing and provides the lubrication.

[John_Miles]

I think that sums up the situation with pre-Grouping design generally and is the fundamental reason why the LMS's early attempts at standardisation ran into some difficulties. Of course there were some lame ducks among the pre-Grouping engines and some places were the standard designs - especially the 2P and 4P 4-4-0s in SW Scotland - were successes, especially with the locomotive men (a prerequisite for success generally).

 

My chemistry teacher at school many, many years ago - early 80s - was a member of the 8F Soc.

 

The Midland 4-4-0s seem to have had a problem with friction in the steam supply which meant that at high speed they lost more power than one would expect. The main steam supply pipe was IIRC only 5" diameter whereas other railways used larger diameters. Friction is proportional to velocity squared and velocity is flow divided by cross sectional area.

[Compound2632]

The Midland 4-4-0s seem to have had a problem with friction in the steam supply which meant that at high speed they lost more power than one would expect. The main steam supply pipe was IIRC only 5" diameter whereas other railways used larger diameters. Friction is proportional to velocity squared and velocity is flow divided by cross sectional area.

 

Which engines? The 483 Class / standard 2P or the engines of the Johnson era? The latter I don't believe as a 4-4-0, a 4-2-2, and a 2-4-0 (all with the same front end except that the 4-2-2 in question had piston valves) appear in the select group of locomotives recorded at over 80mph in the 19th century [O.S. Nock, Speed Records on Britain's Railways (David & Charles, 1971)]. The speed-limiting effect of the combined reverser for a all but the first two Compounds is well-known, so not those, I presume.

[LMS2968]

I don't know of the others, but the Midland / LMS 2Ps had a very restricted steam flow to the cylinders. This was due the valve chests being below the cylinders, so the steam in and out had to squeeze between narrow passages between the cylinders and frames. One effect was that steam could not enter or leave the cylinders at any appreciable speed, so at high speeds the engine would overrun the steam flow. A second was that it was impossible to work the engines hard; it just couldn't be done. So this gave the illusion of a free steaming boiler - the same boiler as used on the 4Fs - as it could provide steam faster than the steam could be used by the cylinders; and their repair costs were very low as it was impossible to wear them out!

[DCB]

I have read the oil type being changed in numerous books. I have been reading Working with LMS Steam by HCH Burgess, which gives some interesting views from the running shed as well as the ES Cox books on the design end. Shame the two didn't seem to communicate.

The Midland axlebox heating was obviously not a big issue as Midland sheds were well versed in repairing hot boxes, I have read they could do an axlebox in one shift where the GWR needed 3 days including sending the box to main works for machining.

Stanier took Colletts GWR Axlebox to the LMS.  Cook in his book Swindon Steam outlines the work Collett did in improving the older (Churchward) axlebox's lubrication.

I don't recall dates but early 20's sounds likely.   The GW had severe lubrication problems with small ends which showed up in the Collett era pre the London -Edinburgh non stops when the GW Penzance to Paddington run was the longest run for a loco in the UK (Exeter- Penzance, Penzance-Paddington,Paddington-Exeter)  After many years with Saints Star class were drafted to the turn and suffered seized small ends and broken rods, they stayed in the Alligator slide bars so no big drama.  Collett investigated and modified the oil boxes and later the rods.  Also at  this time the composition of felt pads etc of the axleboxes and the underkeep design were changed, and as GW locos had regular maintenance every 80K miles or so this was rolled out rapidly.

My point is the Stanier axlebox was 1930s state  of the art not 1900s GW technology.

I think the S&D 2-8-0s had mechanical lubrication to the axleboxes and they were kept off passenger work until the 1950s through fear of running hot according to the Ivo Peters books.

The old chestnut of Midland valve design being OK as spinners did 80 downhill is interesting, William Deans contemporary 4-2-2s did 80 on the flat!  However short travel short lap valve gear is no impediment to high speed as long as the driver drives on the regulator with long cut offs, it is the economy which suffers with short lap short travel valves as you can't use the steam expansively and still get a free exhaust, City of Truro did OK with short travel valves while LNWR Experiments were clocked at 90+ on 6ft 3 " wheels probably helped by nice wet saturated steam.  

[John_Miles]

Which engines? The 483 Class / standard 2P or the engines of the Johnson era? The latter I don't believe as a 4-4-0, a 4-2-2, and a 2-4-0 (all with the same front end except that the 4-2-2 in question had piston valves) appear in the select group of locomotives recorded at over 80mph in the 19th century [O.S. Nock, Speed Records on Britain's Railways (David & Charles, 1971)]. The speed-limiting effect of the combined reverser for a all but the first two Compounds is well-known, so not those, I presume.

 

This again comes from Adrian Tester but not in his book. I'm not clear how many engines it applied to but it certainly was the 483s. It wouldn't be as important with the smaller locos because the rate of flow of steam would be less. A 4-2-2 was recorded doing well over 90 near Sharnbrook which was possibly a speed record for a steam engine until City of Truro's feat. Also Tuplin (who hated Midland engines) records a 4F as doing over 80. It was on the Thames-Clyde express. The train engine failed at Appleby and the 4F was pressed into action to take the train to Carlisle. 

 

The compounds are a more difficult subject. Unlike a simple engine, they needed to be worked at much longer cut-offs (60 to 70%) as the steam expanded twice. As there was no training of engine crews, many would presumably have no clue how to drive them correctly so reports of them not doing well on, for instance, the Birmingham 2 hour trains have to be considered with the above in mind.

[John_Miles]

The LMS did some extensive testing of locos shortly after the grouping to see which types should be proliferated in the future. The main line express engines tested were Claughtons, Prince of Wales, L&Y 4-6-0, Caledonian 4-6-0 and MR compound. The Compound came out best which doesn't say much for the rest. One thing you have to know about the Compounds is that they had quite a large grate area (IIRC 28 sq feet) and so their power output was quite substantial. For large goods engines the test was between a Super D, S&D 2-8-0 and an L&Y 0-8-0. The Super D came out best so we got the Austin 7s. Fowler used to boast he knew nothing about locomotive design (he was a gas engineer). I'm not sure how the 4F was chosen but what was the competition? ​Sorry just seen post above, somebody has already mentioned the goods engine trials.

Edited April 7, 2018 by John_Miles

[Ohmisterporter]

My only experience of riding a 4F was a footplate ride from Lancaster Green Ayre to Castle station with my uncle Dennis who was at BR driver at the time. No hot boxes but I thought the footplate was a dirty place with ash blowing up from below the footplate into the cab. Something I have not noticed on any other engine to the same degree. Perhaps the cabside to footplate bolts needed tightening or something.

[John_Miles]

I don't know of the others, but the Midland / LMS 2Ps had a very restricted steam flow to the cylinders. This was due the valve chests being below the cylinders, so the steam in and out had to squeeze between narrow passages between the cylinders and frames. One effect was that steam could not enter or leave the cylinders at any appreciable speed, so at high speeds the engine would overrun the steam flow. A second was that it was impossible to work the engines hard; it just couldn't be done. So this gave the illusion of a free steaming boiler - the same boiler as used on the 4Fs - as it could provide steam faster than the steam could be used by the cylinders; and their repair costs were very low as it was impossible to wear them out!

 

As Mr Compound aka Stephen Lea, says above the 2Ps or 483 class did some outstanding work on demanding lines such as the S&D and the G&SWR Port Road. Both have long banks and the 2Ps did well so they can't have been a complete disaster.

[LMS2968]

Minor correction: Henry Fowler had trained at Horwich as a premium apprentice in the loco shops under John Aspinall. He was a loco engineer and one of his specialties was boilers (metallurgy being the other). Once out of his time he took charge of the gas department at Horwich, and later moved to the Midland in the same capacity. But he did know about locos, and was a far better all-round engineer than generally given credit.

 

The actual comment was reputed to be at a party, where he told a guest, "I've never designed a locomotive in my life." It might have been true, but to be fair, would also have been true of most other CMEs. Generally, these never sat behind a drawing board and draw out schemes; they had people to do that!

[rodent279]

Minor correction: Henry Fowler had trained at Horwich as a premium apprentice in the loco shops under John Aspinall. He was a loco engineer and one of his specialties was boilers (metallurgy being the other). Once out of his time he took charge of the gas department at Horwich, and later moved to the Midland in the same capacity. But he did know about locos, and was a far better all-round engineer than generally given credit.

 

The actual comment was reputed to be at a party, where he told a guest, "I've never designed a locomotive in my life." It might have been true, but to be fair, would also have been true of most other CMEs. Generally, these never sat behind a drawing board and draw out schemes; they had people to do that!

In the same vane, Collet was a production engineer really, not a locomotive engineer.

[LMS2968]

As Mr Compound aka Stephen Lea, says above the 2Ps or 483 class did some outstanding work on demanding lines such as the S&D and the G&SWR Port Road. Both have long banks and the 2Ps did well so they can't have been a complete disaster.

Neither, as far as I know, were high speed lines, though. And outstanding is open to interpretation. As others have pointed out, the similar looking Southern L1s would leave them standing.

 

Yes, they worked. And that's about as much as can be said for them.

[Compound2632]

Neither, as far as I know, were high speed lines, though. And outstanding is open to interpretation. As others have pointed out, the similar looking Southern L1s would leave them standing.

 

Yes, they worked. And that's about as much as can be said for them.

 

They satisfactorily performed the tasks for which they were designed, as did the L1s. Different tasks.

[brack]

To my eyes the issue is more that they were good Edwardian designs, but built and used 2 or 3 decades too late. Plenty of other CMEs struggled with the step up from 4-4-0 to 4-6-0 as train loadings increased after 1910 or so. Just making longer versions of a successful 4-4-0 and putting in extra coupled wheels did not work (Drummond). On other lines this was either overcome by new CMEs willing to experiment with new design features or by new locos being built to replace their sluggish attempts post grouping. The choice of CME for each of the big 4 is key - churchward had cracked how to design a modern loco and collet continued in his footsteps, maunsell had shown he could do it with the n and u class, Gresley could do big modern locos (the LNER were lucky Robinson turned them down as his 4-6-0s were a mixed bag). Hughes, if left to his own devices might have got there, the crab was a good loco, the 4-6-0s were just about ok post rebuilding and he looked elsewhere to find and bring in new ideas. He was probably the correct choice for the LMS as pickersgill definitely couldn't make a good 4-6-0 and Fowler was happy to ignore what the rest of the country was doing and carry on in the same way, doing what the midland had been doing for years irrespective of changing circumstances.

The later Highland 4-6-0s, including the rivers, were good locos, probably because there was significant input from private builders. They might've offered a better starting point for development than those trialled by the LMS, but I can't see Crewe or derby taking directions from Lochgorm.

[LMS2968]

Yes, but those tasks were for the Midland, not LMS. Big difference!

[Compound2632]

Yes, but those tasks were for the Midland, not LMS. Big difference!

 

My impression is that at the Grouping, all the major LMS constituents - LNW/L&Y, Midland, Caledonian - were struggling to produce express passenger engines that met their Traffic Departments' demands. On the English lines at least, part of the problem was with the Civil Engineering side. On the LNW, this was down to a lack of understanding of the dynamic effects of a locomotive at speed - the George the Fifth class were more destructive of the permanent way than the four-cylindered Claughtons would have been if Bowen Cooke had been allowed to give them the boiler he wanted. On the Midland, the issue was more simply one of weak underbridges. Deeley had wanted to introduce compound 4-6-0s, and even on the goods side, Johnson's parting shot was a 0-8-0 that was cancelled as soon as he retired. 

 

The LNW engines were notorious for their high coal consumption. 

[LMS2968]

High coal consumption had never been a problem under Frank Webb!

 

The issue started with the Precursors and Experiments - quite big engines for the day and designed exclusively to move the traffic. They were saturated engines, still overloaded, so high coal consumption was inevitable as they were thrashed along in true LNWR fashion! The superheated George the Fifths and Princes were better. The Claughtons suffered from the Schmidt single wide piston ring. Once this was replaced, coal consumption came down dramatically to acceptable / good.

[Compound2632]

High coal consumption had never been a problem under Frank Webb!

 

The issue started with the Precursors and Experiments - quite big engines for the day and designed exclusively to move the traffic. They were saturated engines, still overloaded, so high coal consumption was inevitable as they were thrashed along in true LNWR fashion! The superheated George the Fifths and Princes were better. The Claughtons suffered from the Schmidt single wide piston ring. Once this was replaced, coal consumption came down dramatically to acceptable / good.

 

The coal consumption of Whale's engines was certainly high on the agenda for the locomotive men at the Conciliation Boards in 1909 - the first Government attempt at industrial arbitration and broadly a good thing from the employees' point of view - though key members of LNWR management such as Herbert Walker and C.J. Bowen-Cooke come out of it well too [R. Preston Hendry and R. Powell Hendry, The North Western at Work (Patrick Stephens, 1990)].

Edited April 7, 2018 by Compound2632