Imaginary Locomotives

[Gilwell Park]

Hi. I was recently reading an old Railway Magazine article which compared the use of suburban tank engines by the LSWR, LBSC & SECR. The LSWR & SECR mostly used 0-4-4t of several classes while the LBSCR, although having the D3class of  0-4-4t primarily used 0-6-2t. I therefore decided to build a LSWR 0-6-2t to supplement my G6 0-6-0t which is not very healthy these days. It is basically an old Triang M7 body, with new smokebox & splashers fitted to a Hornby modern 0-6-0 chassis. I think it looks feasible & actually runs very well. I also think that I have found a number not used by the LSWR/SR.

 

Roger

[tythatguy1312]

alright so idea that I'm too inexperienced to try and visualise

LNER A1 4-6-0. 
nobody mention Henry

[Corbs]

Take your pick, 4-6-0 is B

https://www.lner.info/locos/B/b.php

 

The B17 is probably closest as it's a Gresley design.

https://www.lner.info/locos/B/b17.php

Edited July 25, 2021 by Corbs

[Andy Kirkham]

The LNER Encclopedia https://www.lner.info/locos/B/b1thompson.php has this to say about the B1s.

 

The B1s did not require any major developments in the design. This was mainly due to the simple, robust nature of the design; but also due to impending conversion to diesel power. The firebox plates did tend to fracture, though. This was sufficiently severe that in 1955 there was a plan to replace the boilers with BR Type 3 boilers, as fitted to the BR Standard Class 5MT locomotives. 

 

Would anyone with Photoshop skills care to tackle that?

[rockershovel]

21 hours ago, Compound2632 said:

 

F.W. Webb pioneered the 8-coupled goods engine in Britain. After an initial inside-cylindered prototype, he built three-cylinder and four-cylinder compounds, all 0-8-0s. One of the earliest acts of George Whale on acceding to the throne at Crewe was to stick a carrying axle under the front end of the four-cylinder engines - the cylinder block forward of the leading pair of coupled wheels had but an excessive weight on that leading axle. The majority of the compound engines were eventually rebuilt to resemble the inside-cylinder prototype, and subsequently (or directly) gained larger boilers. It was not until early LMS days that they gained superheaters. But they kept that front overhang.

 

... which demonstrates my point, really. Webb was a highly original designer, and few (if any) chose to follow where he led. Look at the NER or early LMS (Midland -designed) 0-8-0 for a typical type.

 

Robinson at the GCR, Churchward at the GWR and subsequently Stanier at the LMS demonstrated the superiority of the 2-cylinder, simple expansion 2-8-0. Whale had demonstrated the problems of trying to fit leading trucks where they weren't designed to go and the 0-8-0 types were left to soldier on, updated intermittently as technology improved - a common process when it comes to high-value, long-lived major assets. But nobody designed any more 0-8-0 and they were largely gone by the mid-1930s. Had it not been for two decades of war and depression, they would have been gone a decade earlier. 

Edited July 25, 2021 by rockershovel

[Compound2632]

30 minutes ago, rockershovel said:

 

... which demonstrates my point, really. Webb was a highly original designer, and few (if any) chose to follow where he led. Look at the NER or early LMS (Midland -designed) 0-8-0 for a typical type.

 

 

The LMS Standard 7F 0-8-0 (the Austin Seven) was closely based on the final version of the LNWR 0-8-0, the G2. All those LNWR 8-coupled engines had stuck to the 5'9" + 5'9" + 5'9" wheelbase of Webb's engines (from which of course many of them had been rebuilt) with the leading axle sufficiently far back to clear the outside cylinders of the compound engines. On the 7F, the wheelbase was 6'9" + 5'9" + 5'9", bringing the front axle forward to a much more usual position relative to the inside cylinders - much more like the layout of the 4F 0-6-0.

 

The North Eastern engines, also having outside cylinders, were of similar proportions to the LNWR 0-8-0s, with an even more pronounced front overhang, with the footplating extending well forward of the smokebox front, on account of the piston tail-rods: 8'6" from buffer beam to leading axle. The Worsdell brothers were Crewe by upbringing...

 

There was, I should think, a desire to keep the wheelbase of the early 8-coupled engines as short as possible, since they had to go round the same curves as the 6-coupled goods engines.

Edited July 25, 2021 by Compound2632

[Compound2632]

I've looked up the proposed Midland 0-8-0 of 1903. This too had outside cylinders, 20 in diameter x 26 in stroke, with wheelbase 6'0" + 5'6" + 6'0" - so probably had the leading axle in a similar position relative to the cylinders as the North Eastern and LNWR engines. So the question must be, why was there a preference for outside cylinders for these 0-8-0s? A 0-8-0 is only more powerful than a 0-6-0 if it has larger cylinders (for the same boiler pressure, etc.); that means larger diameter, which with just 4'1" or so between the frames, becomes a problem. But the LNWR 0-8-0s - at least Class G onwards - managed to have 20.5 in diameter cylinders between the frames.

 

With superheating and higher boiler pressure (200 psi for the LMS 7F vs 160 psi for the LNWR G1), it was possible to use smaller cylinders - 19.5 in diameter x 26 in stroke for the 7F.

[Compound2632]

I should perhaps have also noted that when the Derby LDO did produce an outside-cylindered 8-coupled goods engine, it did have a leading truck - the S&DJR 2-8-0s of course.

[AlfaZagato]

Were there any three-cylinder simple 0-8-0 in the UK?  I could see that being an interesting compromise design.   No pilot truck to interfere with maintenance of the inside cylinder, while pushing extra power with outside cylinders to the limits of the loading gauge.

 

Having typed this out, I think a four-cylinder simple would work better in this regard.   Effectively, strapping two cylinders to the outside of a 7F or the like.

[DavidB-AU]

4 minutes ago, AlfaZagato said:

Were there any three-cylinder simple 0-8-0 in the UK? I could see that being an interesting compromise design.   No pilot truck to interfere with maintenance of the inside cylinder, while pushing extra power with outside cylinders to the limits of the loading gauge.

 

NER Raven T3 (LNER Q7). Despite being extremely powerful (hauling a 1400t test train from Newcastle to Carlisie) they were hated by the crews because it was difficult to maintain and lubricate the inside cylinder. 

 

Cheers

David

[rodent279]

3 hours ago, DavidB-AU said:

 

NER Raven T3 (LNER Q7). Despite being extremely powerful (hauling a 1400t test train from Newcastle to Carlisie) they were hated by the crews because it was difficult to maintain and lubricate the inside cylinder. 

 

Cheers

David

Impressive machines as well, no.901, currently out of boiler ticket, hardly seemed to notice the 1:49 climb out of Grosmont to Goathland. Shame they didn't try one banking the Lickey.

[Johnson044]

On 24/07/2021 at 22:15, Gilwell Park said:

Hi. I was recently reading an old Railway Magazine article which compared the use of suburban tank engines by the LSWR, LBSC & SECR. The LSWR & SECR mostly used 0-4-4t of several classes while the LBSCR, although having the D3class of  0-4-4t primarily used 0-6-2t. I therefore decided to build a LSWR 0-6-2t to supplement my G6 0-6-0t which is not very healthy these days. It is basically an old Triang M7 body, with new smokebox & splashers fitted to a Hornby modern 0-6-0 chassis. I think it looks feasible & actually runs very well. I also think that I have found a number not used by the LSWR/SR.

 

Roger

This convinces me!

[Flying Pig]

21 hours ago, rockershovel said:

Robinson at the GCR, Churchward at the GWR and subsequently Stanier at the LMS demonstrated the superiority of the 2-cylinder, simple expansion 2-8-0. Whale had demonstrated the problems of trying to fit leading trucks where they weren't designed to go and the 0-8-0 types were left to soldier on, updated intermittently as technology improved - a common process when it comes to high-value, long-lived major assets. But nobody designed any more 0-8-0 and they were largely gone by the mid-1930s. Had it not been for two decades of war and depression, they would have been gone a decade earlier. 

 

This gives a very misleading impression of unsuccessful designs.  Actually both the LNWR and NER types were well suited to the low speed freight work they were built for and served in significant numbers well into nationalisation, with the last Q6 not being withdrawn until 1967.  

[DenysW]

What if ...

 

To reduce ridicule of the consequences of their Light Engine Policy (formalised in 1907), the Midland Railway's marketing group decided shortly afterwards to re-brand it as a Lightly-Loaded Engines Policy (which it already was) and therefore had to introduce classes of locomotive that were equivalent to double-heading two 2Fs, two 4Fs, and two 4Ps. As I crunch their classification system into horsepower (and I wish I'd made this up) you get:

--   2F + 2F = 8F

--   4F + 4F = 11F

--   4P + 4P = 11P

 

11F is just-about achievable with manual stoking, but 11P (2100 hp) may not be.

 

I assume the 8F would be a bigger 2-8-0 on the same logic that gave the Somerset & Dorset Joint 7Fs, but I think they'd have to buy a 2-6-6-2 North British Mallet for the 11F. Some were supplied by NB to the South African Railways in 1911, and this is before Beyer Peacock had any realistic reference standard gauge reference builds.

 

Would the whole ideal of an 11P have pushed them into overhead electrification of Leeds-Carlisle?

 

I'm also assuming that, it being Company Policy to have 11Fs, the cost of a dozen extra 85-ft turntables would have bean absorbed by cutting some of the non-policy-makers pet projects.

Edited July 26, 2021 by DenysW
Correct typo

[tythatguy1312]

10 minutes ago, DenysW said:

What if ...

 

To reduce ridicule of the consequences of their Light Engine Policy (formalised in 1907), the Midland Railway's marketing group decided shortly afterwards to re-brand it as a Lightly-Loaded Engines Policy (which it already was) and therefore to introduce classes of locomotive equivalent to double-heading two 2Fs, two 4Fs, and two 4Ps. As I crunch their classification system you get:

-

that's basically what Stanier did, but I'm having an... amusing time picturing a 4-6-0 version of the SDJR 7F's. Can someone with better editing skills and a less crash-prone editing software get on that?

Edited July 26, 2021 by tythatguy1312

[Compound2632]

52 minutes ago, DenysW said:

What if ...

 

To reduce ridicule of the consequences of their Light Engine Policy (formalised in 1907), the Midland Railway's marketing group decided shortly afterwards to re-brand it as a Lightly-Loaded Engines Policy (which it already was) and therefore had to introduce classes of locomotive that were equivalent to double-heading two 2Fs, two 4Fs, and two 4Ps. As I crunch their classification system into horsepower (and I wish I'd made this up) you get:

--   2F + 2F = 8F

--   4F + 4F = 11F

--   4P + 4P = 11P

 

11F is just-about achievable with manual stoking, but 11P (2100 hp) may not be.

 

 

I'm interested to see how you arrive at these classifications, since the LMS system of power classification did not define anything beyond 8P or 8F.

[Corbs]

IIRC there's no upper limit to power class 9 in BR's updated system.

[AlfaZagato]

1 hour ago, Corbs said:

IIRC there's no upper limit to power class 9 in BR's updated system.

I wasn't aware of that at all.   Puts greater paid to some of the fantasy Standards posted here before.

[MikeOxon]

Since 0-6-2Ts have recently been under discussion, I wonder what would have happened in the Welsh valleys at the Grouping, if someone more imaginative than Collett had been in charge at the GWR.  A fairly frequent ‘what if’ is to assume that Stanier stayed with the GWR, in which case, it is said, we simply look at his LMS designs to see what would have happened on the GWR.  There were, however, some very significant differences between operating conditions on the GWR and the LMS, which make such simplistic assumptions unsatisfactory.

 

If the story given by John Gibson in his book “GWR Loco Design – a critical appreciation” is true, then Collett made a serious mistake in the layout of the 56xx valve gear and, as a result, it was his “last attempt at producing anything original”. Thus, he kept to all Churchward’s design features as ‘sacred cows’, including nonsenses such as inside Walshaerts motion, bar frame bogies, extension frames, and, arguably, unnecessarily complicated 4-cylinder designs and low superheat.

 

To return to the 0-6-2T in particular:

 

According to Wikipedia “… the 5600s were very unpopular with footplate crews at the time. They were beset by numerous failures, the most common of which was hot axle boxes. ... They also had the tendency to derail, so those driving them preferred them in reverse, where the pony truck was able to guide them around tight curves. ...  the Welsh railwaymen discovered that the new GWR 5700 Class 0-6-0 pannier tank (introduced 1929) was even more suitable for the same work – being shorter and lighter, with roughly the same (slightly lower) tractive effort “

 

So, to come back to the Welsh valleys, what should have been provided?

According to John Gibson, the axlebox problems were down to the use of inside cylinders on such powerful engines. There was insufficient space for adequate bearings on the driving axles. On later engines the inside and outside cranks were arranged in-phase (Stroudley arrangement), which reduced the thrust on the bearings at the expense of needing very large balance weights, which increased axle loading.

 

So, what are we looking for? An outside cylinder engine, perhaps with outside valve gear and with high adhesion weight and short length (to fit local turntables).

 

Broadening the field, We could expect a different designer than Collett to have come up with something more like the later modified Halls, rather than simply providing a smaller-wheeled ‘Saint’. And what about the call from the board for more tractive effort, which Collett met with the ‘King’, restricted to very few routes. Harry Holcroft present a paper on 3-cylinder engines to the Swindon Engineering Society in 1923. Could this have been a better way of achieving more power, with less weight and complexity?

 

Perhaps, some of the designers on here would like to exercise their imagination.

[Gibbo675]

13 minutes ago, MikeOxon said:

Since 0-6-2Ts have recently been under discussion, I wonder what would have happened in the Welsh valleys at the Grouping, if someone more imaginative than Collett had been in charge at the GWR.  A fairly frequent ‘what if’ is to assume that Stanier stayed with the GWR, in which case, it is said, we simply look at his LMS designs to see what would have happened on the GWR.  There were, however, some very significant differences between operating conditions on the GWR and the LMS, which make such simplistic assumptions unsatisfactory.

 

If the story given by John Gibson in his book “GWR Loco Design – a critical appreciation” is true, then Collett made a serious mistake in the layout of the 56xx valve gear and, as a result, it was his “last attempt at producing anything original”. Thus, he kept to all Churchward’s design features as ‘sacred cows’, including nonsenses such as inside Walshaerts motion, bar frame bogies, extension frames, and, arguably, unnecessarily complicated 4-cylinder designs and low superheat.

 

To return to the 0-6-2T in particular:

 

According to Wikipedia “… the 5600s were very unpopular with footplate crews at the time. They were beset by numerous failures, the most common of which was hot axle boxes. ... They also had the tendency to derail, so those driving them preferred them in reverse, where the pony truck was able to guide them around tight curves. ...  the Welsh railwaymen discovered that the new GWR 5700 Class 0-6-0 pannier tank (introduced 1929) was even more suitable for the same work – being shorter and lighter, with roughly the same (slightly lower) tractive effort “

 

So, to come back to the Welsh valleys, what should have been provided?

According to John Gibson, the axlebox problems were down to the use of inside cylinders on such powerful engines. There was insufficient space for adequate bearings on the driving axles. On later engines the inside and outside cranks were arranged in-phase (Stroudley arrangement), which reduced the thrust on the bearings at the expense of needing very large balance weights, which increased axle loading.

 

So, what are we looking for? An outside cylinder engine, perhaps with outside valve gear and with high adhesion weight and short length (to fit local turntables).

 

Broadening the field, We could expect a different designer than Collett to have come up with something more like the later modified Halls, rather than simply providing a smaller-wheeled ‘Saint’. And what about the call from the board for more tractive effort, which Collett met with the ‘King’, restricted to very few routes. Harry Holcroft present a paper on 3-cylinder engines to the Swindon Engineering Society in 1923. Could this have been a better way of achieving more power, with less weight and complexity?

 

Perhaps, some of the designers on here would like to exercise their imagination.

Hi Mike,

 

Mr Bullied's Q1 class of locomotive had cylinders of the same stroke, yet with an inch extra on diameter and an extra 30lb of boiler pressure, so the inside cylinder layout causing hot boxes does seem to me be be a nonsense.

 

As for being simple the Q1's were certainly that !

 

Gibbo.

[DenysW]

Basics -

 

Physics: Power = force times velocity. In imperial units 1 hp = 550 ft-lb-wt/sec, the MR classifications are at 25 mph (Freight, giving 25/15 times 22 ft/sec) or 50 mph (Passenger, giving 50/15 times 22 ft/sec), and they quote the force range available in lb-wt at these speeds. During "A" level maths many years ago one of the numbers you memorised was 15 mph = 22 ft/sec.

 

Calcs:

 

So if you take the mid-point of the MR classifications,  and multiply the listed force by the speed in ft/sec and divide by 550, then 1F = 450hp +/- 50 hp, and the classes go up in 100 hp increments (all roughly) . 1P = 500 hp +/- 75 hp and the classes go up in 150 hp increments, (again roughly). Hence 4P+4P does not equal 8P.

 

This also means you can extend either/both series up as far as you like, even if the MR didn't, have nothing to put into the classes. Or if it's stupid: why put the Lickey Banker or (later) the LNER Garratt into classes that assume they can deliver that power at a particular speed, when they were not designed to achieve it.

 

You could extend them down as well if you fancy explaining negative numbers to forthright, hard working folks who started their railway life as locomotive cleaners.

 

 

 

 

[Compound2632]

48 minutes ago, DenysW said:

Basics -

 

Physics: Power = force times velocity. In imperial units 1 hp = 550 ft-lb-wt/sec, the MR classifications are at 25 mph (Freight, giving 25/15 times 22 ft/sec) or 50 mph (Passenger, giving 50/15 times 22 ft/sec), and they quote the force range available in lb-wt at these speeds. During "A" level maths many years ago one of the numbers you memorised was 15 mph = 22 ft/sec.

 

Calcs:

 

So if you take the mid-point of the MR classifications,  and multiply the listed force by the speed in ft/sec and divide by 550, then 1F = 450hp +/- 50 hp, and the classes go up in 100 hp increments (all roughly) . 1P = 500 hp +/- 75 hp and the classes go up in 150 hp increments, (again roughly). Hence 4P+4P does not equal 8P.

 

This also means you can extend either/both series up as far as you like, even if the MR didn't, have nothing to put into the classes. Or if it's stupid: why put the Lickey Banker or (later) the LNER Garratt into classes that assume they can deliver that power at a particular speed, when they were not designed to achieve it.

 

You could extend them down as well if you fancy explaining negative numbers to forthright, hard working folks who started their railway life as locomotive cleaners.

 

 

Excellent. There was a thread (it might even have been further up this one) where the formula was discussed. I was looking for it, but failed to track it down! The key factors are total cylinder volume, pressure (cylinder inlet and exhaust), and driving wheel diameter (i.e. distance travelled per cylinder cycle). So if you built an engine that had the same 5'3" drivers as a 4F, gave it four 20" x 26" cylinders and a boiler pressed to 175 psi, it would, notionally, have twice the power of a 4F. But that would be unlikely to work out in practice. The LMS system included a second criterion, as I recall, which tried to allow for the capability of the boiler to produce a sufficient volume of steam to maintain the nominal power - grate area came into it.

 

As you say, the power classification scheme was only useful in the circumstances for which it was intended - such as the suitability of a goods engine for plodding along at 25 mph with a 60-wagon mineral train. 

Edited July 26, 2021 by Compound2632

[The Johnster]

The genesis of the Collett 56xx class seems to have been a desire to provide a general purpose locomotive built of standard Swindon components to standardise the variety of types, some of which were museum pieces with very short life expectancies, that the GW had acquired at the grouping.  The GW had considerable experience in South Wales, using heavy panniers and 2-8-0 tank locos for most of their lines in the area, such as the Monmouthshire valleys, Vale of Neath, and the Tondu valleys.  The absorbed and constituent railways had slightly different needs, however, including intense and tightly timed commuter work around Cardiff, and collieries in the Taff, Rhonddas, and Rhymney (period spelling) valleys that were not ideal for 2-8-0s.

 

Recieved wisdom on railways like the Taff Vale, Rhymney, Barry, and so on was that the best loco for the work was an 0-6-2T, with different driving wheel sizes for passenger (about 5'3"), 'goods' (about 4'7"), and mineral (about 4') work; the trailing wheels helped with the ride on the faster downhill passenger runs and tipped the locos forward so that the weight was carried on the driving wheels in a way that might not have occured in the same way had a leading pony truck been provided.  Speed was not the primary concern, nor (perhaps surprisingly to those who don't know the area) was braking, as all these lines employed 'incline working', manually pinning down handbrakes on the loose coupled coal wagons until the driver called a stop to the game, which was when the train was braked enough for him to descend the bank under light steam, actually pulling the coal trains down the steeper hills to control them (mostly, this needed nerves of steel and did not always go according to plan...).  The extended frame over the radial rear axle allowed big water tanks and enough coal for the day's work, and the locos needed to be powerful enough to haul the (fairly weighty and long) trains of empties uphill to the collieries.

 

If Collett considered that he had an opportunity to do something different to this in South Wales, he didn't consider it for long, nor did he consider using the existing GW designs that had proven themselves in South Wales for further new builds for work on the TVR, Barry Rhymney, etc.  A policy emerged of replacing the older locos with new types using standard GW components, which turned out to be the 56xx and 57xx, of which the 56xx was first on the scene.  It was basically a Rhymney R class built of standard Swindon bits and provided with a vacuum brake to work passenger trains, and proved fast enough to do this on the Taff Vale and Barry.   One might have expected a 'passenger' version with 5'2" wheels, and a pure mineral hauler with 4'1" wheels, but the 56xx was apparently considered adequate. 

 

It ran into trouble with the locomen of the absorbed and constituent railways, and company loyalty was probably at least as responsible for this as any other factor; locomen are by nature small c conservative and anything new from a different company that has assumed command is going have to be exceptionally good before the old  timers have a good word to say for it.  The cabs were too exposed to the South Wales rain, and sliding shutters were fitted, eventually appearing on other GW tank designs, the axleboxes weren't up to it, the cabs were small and awkward to swing a shovel in, why can't we have more Rhymney Rs/Taff As/insert type according to pre-grouping company.  Another complaint was from the railways' neighbours'; the grinding of brakes was deafening.  Notwithstanding, 200 were ordered and most of the objections eventually overcome.  GW power class was D.

 

Then along came the next replacement loco, the 57xx.  This had less of a South Wales pedigree and was a development on the 1854s and 2721s already common on the area's GW lines and across the railway.  These were ideal for the shorter haul colliery trips and became the go-to for yard pilot work, and they were undoubtedly pocket rockets much appreciated by the crews that worked on them, but to say that they were capable of 56xx work and that 200 less 56xx and 200 more 57xx should have been built is pushing it a bit.  They did not have sufficient coal capacity for 56xx jobs and were nowhere near as powerful, GW power class C. 

 

As part of the general policy, the more modern absorbed/constituent locos were Swindonised, and most lasted into the BR period.  The final part of the plan to replace them featured the 94xx heavy pannier, over half of which were allocated to South Wales depots.  These were power class D, one up from the 57xx, but also apparently considered no better than a 57xx by some locomen.  At Tondu for some reason they occupied a niche as passenger locos.  The 57xx, in the 8750 form, were still being built to replace older locos in South Wales up to 1950, but the 200 56xx, about 150 of which were allox to the area, lasted until the late 50s before withdrawals began, by which time some 94xx that were barely run in were being withdrawn as well.  Rhymney P and Taff Vale A passenger tanks, still new at the grouping and Swindonised, were replaced by the 41xx series of 5101 large prairies, and eventually by the first build of BR standard class 3MT prairies.  Auto working in the Cardiff Valleys was extended by a new timetable in 1953, and a number of 4575 small prairies auto fitted to work 3 trailer trains up the area's steep banks. 

 

It was only a few more years before the dmus, class 116, appeared and there was no need for passenger steam locos for local services, with the exception of some auto work.  The full replacement program was completed with the  delivery of the 94xx and the later 8750s, 41xx series 5101s, and BR 82xxx, but the writing was on the wall for steam by then.  The 56xx handled most of the coal traffic on the absorbed/constituent lines until the flood of D68xx began to oust them in 1963; all were gone from South Wales within 2 years, as was steam in general.

[PenrithBeacon]

The MR, LMS and BR calculated the tractive effort for freight at 25mph and for passenger at 50mph. There were separate ranges for the two classifications. BR's method of working it out differed from that of the MR & LMS in that it included a variable for free gas area which could have a negative affect for older designs

Edited July 26, 2021 by PenrithBeacon

[rockershovel]

12 hours ago, Flying Pig said:

 

This gives a very misleading impression of unsuccessful designs.  Actually both the LNWR and NER types were well suited to the low speed freight work they were built for and served in significant numbers well into nationalisation, with the last Q6 not being withdrawn until 1967.  

I didn't say the 0-8-0 types were unsuccessful, rather that the 2-8-0 was superior. They were faster, and (for reasons covered above by someone else) the 2-cylinder, outside cylinder, simple expansion configuration made better use of steam and wasn't constrained by excessive overhang or front axle weight. The 0-8-0  types did what they were designed to do, and thirty years of war, depression and technical stagnation (especially the persistence of the 4-wheel, loose-fitted coal wagon) kept that role in existence.