Imaginary Locomotives

[Flying Pig]

5 hours ago, JimC said:

Is that a general opinion? If the occasional what if of standard parts I post is unwanted then I'll go away.

 

No, discuss what you like if you find it interesting or think others might.  There's very little on this thread that is new anyway.  As was noted above.

[JimC]

5 hours ago, BernardTPM said:

And the coupled wheelbase is ridiculously long.

 

4 hours ago, Compound2632 said:

Really? 8' 6" + 8' 6" - only 6" more than the standard Midland goods engine and less than a good few 0-8-0s.

It's the boiler that's ridiculously long, 

Maybe one aspect to this topic is that it can highlight design problems that exercised the minds of designers back in the day. Clearly this extension of a reasonable 4-4-0 has produced an unreasonable 4-6-0, but should we ask why? I'm of the opinion that the turn of the century transition from eight wheeled to ten wheeled locomotives went badly on some lines, and this illustrates to me the issues.

An eight wheeler, be it a large wheel 4-4-0 to a small wheeled 0-8-0, was a fairly straightforward proposition. There was room for the firebox between the last pairs of wheels, a reasonable sized boiler fitted on top, the wheelbase was under control, and we see this in all sorts of reasonably successful locomotives from different lines. 

Enlarging this into a ten wheeled locomotive most especially with driving wheels north of 6'8 or so, was quite a different proposition. I think we can see that here. If we have the same firebox position and drive to the leading wheels, as was customary, then what we get is something like this Johnson pastiche. I got a very similar result when I tried to enlarge a Dean Atbara. The whole locomotive gets so long that the boiler is ridiculously long, it won't steam, the weight distribution is up the chute, all sorts of evils. And we can see that when designers tried it previously competent design teams produced some very mediocre locomotives. 

So I think you have to look at the photo edit and say, well that's how it could have gone even wronger, but what would have been needed to make it right? How do you get 3 sets of 6'9ish driving wheels and a sensible boiler? What did the draughtsmen in the drawing offices have to consider? And if you do that, maybe you can do some imaginaries around how some of the designers that failed to get to grips with the ten wheel era should have gone about it? (Even if the answer is "copy Churchward"!).

 

Edited October 1, 2023 by JimC

[The Johnster]

Indeed.  The big issue facing CMEs in Edwardian days was that trains were getting longer and heavier and restaurant cars meant no long refreshment stops where your inadequate 4-4-0 could be replaced or at least you could sort the fire out a bit.  Boilers needed to be fatter and longer to proved steam for bigger cylinders, which led to Atlantics and 4-6-0s, not, as you rightly say, always successfully.  There is more to a successful 10-wheeler than enlarging your last and biggest 4-4-0; a whole raft of problems arose.  As well as 'where to put the bigger firebox' and 'how to distribute the weight', the new fat boilers obscured the drivers' view ahead and steam hung around them, obscuring it further, because the stubby fat little chimneys were too close to the loading gauge limit to be made taller and had trouble clearing the exhaust. 

 

Churchward neatly avoided this with his taper boilers, which allowed for taller and more efficient-at-smoke-lifting chimneys.  No GW taper boiler design, even the Kings and 47xx, needed smoke deflectors, ever.

[Compound2632]

4 hours ago, The Johnster said:

Boilers needed to be fatter and longer to proved steam for bigger cylinders, 

 

And also more work for the fireman, if you weren't Churchward. Labour unrest was general in the later Edwardian period but one particular issue raised by the LNWR men was the increased labour for the same rate of pay consequent on Whale's coal-eating Precursors and Experiments. Bowen-Cooke, who was popular with the men, acted as internal arbitrator and of course in the new reign gave them the superheated Georges and Princes, which improved matters a good bit but inevitably still ate more than a Webb compound.

 

There was more to big engines than engineering.

Edited October 2, 2023 by Compound2632

[The Lurker]

8 hours ago, Compound2632 said:

 

And also more work for the fireman, if you weren't Churchward.

The implication of this is that GWR 10 wheelers were not more coal-hungry than their predecessors. Is that right?  What was it about his designs that drove this? Was it the taper boilers, or superheat or something else? I guess superheat had a role to play - after all the LNWR were impressed by the coal consumption of a Marsh Atlantic tank (IIRC an I3) on the Sunny South Express, when it ran through to Rugby.

 

EDIT - or does that reflect that the LNWR engines were excessively coal-hungry as per the rest of your post?

Edited October 2, 2023 by The Lurker

[Compound2632]

19 minutes ago, The Lurker said:

EDIT - or does that reflect that the LNWR engines were excessively coal-hungry as per the rest of your post?

 

Yes, one gains the impression that their coal consumption was greater than the other big engines of their day.

 

20 minutes ago, The Lurker said:

The implication of this is that GWR 10 wheelers were not more coal-hungry than their predecessors. Is that right? 

 

 

[Embedded link to Wikipedia Commons.]

 

That is, I understand, a 3,000 gal tender, the same as fitted to Churchward's 4-4-0s and essentially a late Dean design: http://www.gwr.org.uk/no-tenders.html. Churchward did start production of a 3,500 gal tender. Now that's water not coal capacity but surely indicative?

 

26 minutes ago, The Lurker said:

What was it about his designs that drove this? Was it the taper boilers, or superheat or something else? I guess superheat had a role to play - after all the LNWR were impressed by the coal consumption of a Marsh Atlantic tank (IIRC an I3) on the Sunny South Express, when it ran through to Rugby.

 

I think - wait to stand corrected - front end design was important. The engine could make the efficient use of the steam supplied, notwithstanding the Stephenson motion.

[rodent279]

2 hours ago, Compound2632 said:

 

Yes, one gains the impression that their coal consumption was greater than the other big engines of their day.

 

 

 

[Embedded link to Wikipedia Commons.]

 

That is, I understand, a 3,000 gal tender, the same as fitted to Churchward's 4-4-0s and essentially a late Dean design: http://www.gwr.org.uk/no-tenders.html. Churchward did start production of a 3,500 gal tender. Now that's water not coal capacity but surely indicative?

 

 

I think - wait to stand corrected - front end design was important. The engine could make the efficient use of the steam supplied, notwithstanding the Stephenson motion.

GJC understood the importance of long travel long lap valves, and that, married to good boiler design, was the secret.

[Miss Prism]

That is actually a Dean 4000g.

 

 

[Compound2632]

2 minutes ago, Miss Prism said:

That is actually a Dean 4000g.

 

I stand corrected. But some of the early 4-6-0s and 4-4-2s had 3,000 gal tenders at first, didn't they? Per Jim's gwr.org.uk article.

Edited October 2, 2023 by Compound2632

[Miss Prism]

8 minutes ago, Compound2632 said:

I stand corrected. But some of the early 4-6-0s and 4-4-2s had 3,000 gal tenders at first, didn't they? Per Jim's gwr.org.uk article.

 

More or less, yes. The 4000g Deans were only 20 in number, and built initially to go non-stop to Plymouth. but the post-1900 water troughs dispensed with the need for further 4000g units.

 

I have to credit @wenrash as the expert Dean 4000g spotter!

 

[The Johnster]

10 hours ago, Compound2632 said:

That is, I understand, a 3,000 gal tender, the same as fitted to Churchward's 4-4-0s and essentially a late Dean design: http://www.gwr.org.uk/no-tenders.html. Churchward did start production of a 3,500 gal tender. Now that's water not coal capacity but surely indicative?

 

Water consumption is indicative to an extent of how much steam the boiler has produced in response to the route, load, and timing's demands for it.  But it would be very simplistic to use the size of the tender as an indicator of how much steam can be produced, as post-1900, water troughs were provided and the tenders could be replenished on the move. 

 

1900 is pretty close to the major change in operational methods that resulted from the introduction of gangwayed bogie coaches with heavy steel underframes, which allowed catering and toilet facilities on the train.  This was important as the previous refreshment/comfort stops could be done without, so longer non-stop runs became more common, at about the same time as working-class wages rose to give them a measure of disposable income, some of which was spent on long-distance rail travel which simultaneously increased the loads; an express on a trunk route was now typically 13 or 14 bogies and the best part of 500tons, well beyond the capacity of a Victorian inside-cylindered 4-4-0, which meant that the Midland and the LNW resorted to double-heading almost by default, hardly an efficient use of motive power or crews.  The upshot was that bigger and more powerful engines were needed quickly and the CMEs who were being asked to provide them were perhaps in some cases not as up to the task as might have been hoped for...

 

Two of them emerged from this period of enlarged engines as successful, though, Ivatt on the GNR whose Atlantics were fast, powerful, and efficient because of wide but fairly short fireboxes over the trailing wheels, and Churchward, whose 4-6-0s were fast, powerful, and efficient because of long narrow fireboxes and the GW 'haycock' fire, though this had to be discarded in favour of an all over incandescent bed for 'collar' work.  Note that everything is increasing in size but the key component in both of the successful men's cases is the firebox, combined with boilers that could raise steam at a rate higher than the engine could use it even under load at high speed. 

 

Previous to this, during the Victorian era when loads were much lighter, stops were more frequent and longer, and engines much smaller, it was almost a given that your loco, a 4-4-0 or even a reversion to 4-2-2 when steam sanding made it's appearance, would be easily capable of managing it's work without crippling the fireman.  Other givens were that the forward view from the cab through big spectacle plate windows over a low-set boiler was pretty good, the low set of the boiler made the ride pretty good, and the tall chimney that you had room for between the lower top of the smokebox and the loading gauge limit was easily capable of throwing exhaust smoke and steam upwards away from the cab. 

 

Within the first quarter of the twentieth century all those givens had been taken, at least for express passenger work.  The locos were now bigger to haul heavier trains, the work involved in firing them over the longer non-stop distances at the continuous rate needed to keep sufficient steam being raised was getting closer to what could reasonably be expected of human firemen (and, in some cases, beyond that reasonable expectation!), and the view ahead was restricted by big boilers with exhaust smoke and steam hanging around them.  The locos had higher centres of gravity because of the big boilers, and outside cylinders, so the ride at speed was not as good as it had been on the 4-4-0s, and hammer-blow became a 'thing'. 

 

Some of these issues were not resolved by the end of steam in normal commercial service.  By that time steam-raising from very efficient boilers and improved front-end designs had eased the firemens' burden a little, but poor coal eroded most of any gains in that direction. Ride on 4-6-0s was mostly diabolical, to the mechanical detriment of the engine and track, and the fatigue of the crews.  Forward visibility and hammer-blow were never really resolved, and were one of the main reasons for the replacement of steam locomotives. 

 

 

[Compound2632]

3 hours ago, The Johnster said:

Water consumption is indicative to an extent of how much steam the boiler has produced in response to the route, load, and timing's demands for it.  But it would be very simplistic to use the size of the tender as an indicator of how much steam can be produced, as post-1900, water troughs were provided and the tenders could be replenished on the move. 

 

Very true and I was acutely aware - it's why LNWR tenders were miniscule compared to those of other companies in the 19th century - 1,500 gal to 2,000 gal compared with, say, 2,300 gal to 3,250 gal on comparably-sized engines doing similar work on the Midland.

[DenysW]

10 hours ago, The Johnster said:

Ivatt on the GNR whose Atlantics were fast, powerful, and efficient

Fast: Yes

Efficient: No data

Powerful: No

Wide grates: Yes

 

The GNR Atlantics were 2P locomotives, continuing the long-standing but top-secret GNR 'small-engine' policy*, which made sense for their easy gradients, and is why they were able to keep Singles as their status engines as long as they did. Until they went London-Edinburgh non-stop, they handed over to the North Eastern for the worst gradients on the route, which were, not coincidentally, on the North British section.

 

*They waited until 1922 to go big-engine, but only managed to build one prototype loco before Grouping.

[Compound2632]

1 hour ago, DenysW said:

The GNR Atlantics were 2P locomotives, continuing the long-standing but top-secret GNR 'small-engine' policy*, which made sense for their easy gradients, and is why they were able to keep Singles as their status engines as long as they did.

 

Interesting insight into the Atlantics. There were some tough gradients in the West Riding, where the Great Northern was second or third-comer, and the Gresley 2-8-0s were 8F, whereas the Robinson 2-8-0s were 7F like the Derby ones, and also the LNWR Super Ds.

[DenysW]

44 minutes ago, Compound2632 said:

the Gresley 2-8-0s were 8F,

The GNR Atlantics (and, later, Gresley Pacifics) were also 20 tons/axle, and I suspect that means they had a less draconian Civil Engineer than more of the other pre-Grouping companies.

[Flying Pig]

2 hours ago, DenysW said:

Fast: Yes

Efficient: No data

Powerful: No

Wide grates: Yes

 

The GNR Atlantics were 2P locomotives, continuing the long-standing but top-secret GNR 'small-engine' policy*, which made sense for their easy gradients, and is why they were able to keep Singles as their status engines as long as they did. Until they went London-Edinburgh non-stop, they handed over to the North Eastern for the worst gradients on the route, which were, not coincidentally, on the North British section.

 

*They waited until 1922 to go big-engine, but only managed to build one prototype loco before Grouping.

 

I'm struggling to understand how a loco so big could only be a 2P.  Using the usual online sources, they were more than 25% heavier than an LMS 2P, had 50% more grate area and 70% more total heating surface.

[Compound2632]

11 minutes ago, Flying Pig said:

I'm struggling to understand how a loco so big could only be a 2P.  Using the usual online sources, they were more than 25% heavier than an LMS 2P, had 50% more grate area and 70% more total heating surface.

 

None of which (except grate area) counts in the calculation of LMS power class. Compare cylinder volume and boiler pressure (the bigger the better) and driving wheel diameter (larger reduces the tractive effort):

 

Energy transferred per cycle = cylinder volume swept out per cycle x steam pressure = tractive effort (force) x distance travelled per cycle

 

barring various fudge factors.

 

[Flying Pig]

13 minutes ago, Compound2632 said:

 

None of which (except grate area) counts in the calculation of LMS power class. Compare cylinder volume and boiler pressure (the bigger the better) and driving wheel diameter (larger reduces the tractive effort):

 

Energy transferred per cycle = cylinder volume swept out per cycle x steam pressure = tractive effort (force) x distance travelled per cycle

 

barring various fudge factors.

 

 

 

In which case I have to wonder how meaningful it is to call a C1 a 'small engine' based on that calculation.

[DenysW]

2 hours ago, Flying Pig said:

I'm struggling to understand how a loco so big could only be a 2P

I took BR appraisals for the Atlantics (as 2P) from Wikipedia. Extra to what @Compound2632 has written, the Midland/LMS/BR system calculated the tractive effort at 25 mph (goods) or 50 mph (passenger) first, then did a sanity check on grate area, resulting in the lower of the two possibilities being used. The Midland thread on p 40 has the Midland/LYR fiddle factors for effective pressure vs. piston speed to de-rate starting tractive effort to 25 mph and 50 mph; the LMS used 20% higher to reflect better valves giving less pressure loss by their time.

 

The Midland, I am told, used to under-boiler its goods engines (judged by this method) rather than perfectly-match: they wanted to start-up very heavy trains, and then just trundle them along slowly with the boiler at close to its maximum efficiency. Not a philosophy designed to appeal to speed-freaks, just to the accountants.

[The Johnster]

The Midland/LMS power classifications, as used by BR on locos that had never been subjected to it before, came up with some odd results.  According to it, the GW small prairies were 4MT, which sounds as if they should have been 'twice as powerful' as the similarly sized and similarly capable Ivatt and BR Standard 2MT small prairies; I don't think they were!  The 56xx 0-6-2 tank was a pocket rocket and could certainly pull, but BR classed it as 5MT, which is the same as a Hall or Black 5, and there was no way a 56xx could have done that sort of work, good as it was!  Stars, with 4-cylinders and no.1 boilers, were rated at 4P, not much for a big engine that had been at the cutting edge for efficiency when it was introduced and was known to be more powerful than the LNWR Claughton, nominally a 5P.

 

The LBSCR Atlantics, which were based on the large GNR Ivatt engines, were rated at 3P or 4P by BR, which makes the small Ivatt Atlantics at 2P seem more rational, but I suspect all of these engines were doing work that belied those ratings.  On the WR, the BR power classifications were ignored and the GWR alphabetic system continued in use, even in diesel days.  BR standard and LMS locos allox to the region were given the GW coloured route availablility spot and alphabetic power classifications as if they were GW engines.  I suspect the same thing happened on the Southern, which modified the BR system to it's own needs, and on former LNER territory as well.  On the WR, for example, a BR Standard Class 5MT was called a '73 thousand', a continuation of the GW vernacular of '49s', '28s', &c.  The TOPS classifications were very readily adopted when they appeared; prior to that there had been 'D 800s', 'D 1 thousands' and such.

 

My personal view is that the Midland system was useful as a comparitve measure of power for passenger or freight work on that railway, and continued to be useful as applied to new LMS classes, as the LMS introduced the MT and XP variants, but I doubt anyone took much notice of it for pre-grouping LMS locos except ex-Midland types.  Anyone involved with running Dunalastairs of various types, or LNW George Vs, knew what they could do and allocated them to jobs accordingly, and probably ignored the ex-Midland system as irrelevant and uneccessary.  Same probably happened with the BR standards, except perhaps on the LMR or ex-LMS parts of the Scottish.

[Compound2632]

2 hours ago, DenysW said:

I took BR appraisals for the Atlantics (as 2P) from Wikipedia.

 

Compare the vital statistics:

 

GNR large-boilered Atlantic:

Cylinders 2 x 19"/20" diameter x 24" stroke (swept volume = 27,200 cu in / 30,100 cu in)

Drivers 6' 8"

Pressure 170 psi (saturated) / 150 psi (superheated)

 

LMS Standard 2P

Cylinders 2 x 19" diameter x 26" stroke (swept volume = 29,500 cu in)

Drivers 6' 9"

Pressure 180 psi (superheated)

[billbedford]

1 hour ago, The Johnster said:

The LBSCR Atlantics, which were based on the large GNR Ivatt engines, were rated at 3P or 4P by BR, which makes the small Ivatt Atlantics at 2P seem more rational, but I suspect all of these engines were doing work that belied those ratings.

 

Most GN small Atlantics were withdrawn before WW2, with only two being allocated their new Thompson numbers. The majority of the large Atlantics went during 1945-7, with 5 out of 94 lasting till 1950. I don't suppose that much effort went into getting power classification correct. 

[JimC]

It seems to me that the wide firebox Atlantic solves or at least deals with a number of the ten wheel design problems evident in the Johnson pastiche photo. The short wide box over the trailing wheels means that the boiler length is kept down, and trailing wheels under the firebox helps the overall length. So what the designer achieves is a locomotive with a considerable boiler capacity increase over the 4-4-0s. To my mind it's very analogous to the successful Pacifics, where the wide box, together with a combustion chamber to increase firebox heating surface and reduce tube length, also enlarges the boiler capacity without the length of the locomotive getting out of hand. An expansion of the Churchward 4-6-0 type into  4-8-0 as a twelve wheel type, on the other hand, would I think run into a raft of problems with firebox, ashpan and so on. I must take another look at Dusty Durrant's flights of fancy to see how a Swindon trained man handled it. 

 

 

[Later] A quick skim of his book and I haven't seen that Durrant sketched of a 4-8-0 with wheels in excess of 6'0, which means a vital few more inches for grate and ashpan. His long firebox 4-8-0 has two axles under a part sloping grate. And a very long firebox it is too - a good bit longer than a King. Not, I suspect one to find favour with the firemen's union.

 

 

Edited October 4, 2023 by JimC
[later] section

[Compound2632]

9 hours ago, The Johnster said:

The LBSCR Atlantics, which were based on the large GNR Ivatt engines, were rated at 3P or 4P by BR

 

8 hours ago, billbedford said:

Most GN small Atlantics were withdrawn before WW2, with only two being allocated their new Thompson numbers. The majority of the large Atlantics went during 1945-7, with 5 out of 94 lasting till 1950. I don't suppose that much effort went into getting power classification correct. 

 

The Brighton Atlantics had boilers pressed to 200 psi (according to Wikipedia) which makes a big difference in the tractive effort calculation, the formula being linear in boiler pressure. 

 

It's a ruthless calculation that pays no regard to the merits of the front end design, to the effectiveness of the boiler in raising steam, or to sentiment. 

[DenysW]

10 hours ago, The Johnster said:

According to it, the GW small prairies were 4MT, which sounds as if they should have been 'twice as powerful' as the similarly sized and similarly capable Ivatt and BR Standard 2MT small prairies;

No. Sorry, but No.  The scale doesn't start at zero, zero.

 

From Bond's autobiography, p48, although I give the midpoint of his ranges, TE in tons:

 

Class     Passenger Tractive Effort      Freight Tractive Effort

2                                2.25                                         3.95      

4                                3.25                                         5.47