GWR four-cylinder arrangement?

I did some diesel factory and shipyards(now closed) time and found it unfit for a living.

 

Be careful of criticising four-cylinder designs! You are liable to upset not only GWR fans but also fans of the numerous French four-cylinder compounds.

Your own design suggestions are certainly original. I think you would have to cross the Atlantic Ocean to find a King-size 2-cylinder compound loco. The largest in Europe appear to have been around 70 tons, or Manor-sized in GWR-speak. British builders in the 1920s supplied some large 2-cylinder compounds to Argentina, where they remained in favour later than elsewhere. The last such engine for domestic use was a 3-ft gauge 2-4-2T, built in 1920 for the Midland Railway's outpost in Northern Ireland (The "Northern Counties Committee" or NCC).

 

Four cylinder locomotives had been better made three cylindered.


A Bulgarian 0-12-0 class two-cylinder compound from 1922 is still on wheels.
It was superheated /simpled after WW2.
Mass 101 ton


https://preview.redd.it/bulgarian-s...bp&s=49603e8bf6de506a686cd90cd1e4a54e2fb88379

https://www.steamlocomotive.com/locobase.php?country=Bulgaria&wheel=0-12-0&railroad=bs

 

It's not always as simple as that. The French didn't have suitable supplies of steam coal, so imported huge quantities from South Wales. They even had their own ships for such. The GWR, on the other hand, had Welsh steam coal on their doorstep and therefore had less need for compounding.

 

Deleted

 

Less boiler maintenance for locomotives that uses less steam for same job?

 

Boiler maintenance and repair is dictated by time, and not even time in steam. Extensions are possible, but the amount of steam generated isn't likely to be a factor. Steam cycles have a greater effect.

 

Churchward experimented with compounding and found that simple engines were better. He did adopt the way Chapelon (have I spelt that right) used split drive for four cylinder locomotives because of the balancing improvements it offered. This was used on the Stars, Castles and Kings with success but Stanier's use of the principle on the LMS was less successful because of stress on the longer Pacific frames
Compounding in the British context was a failure. It was tried often enough and always found wanting; simple engines are better.

 

I think that you will find that the match between Churchward's locos and the French Atlantics was fairly equal but only because the former had advanced large, long-travel valves whereas the latter had narrow steam passages and slide valves for the LP cylinders. The split drive did not originate with Chapelon but with Alfred de Glehn.

Compounding was never seriously tried in the UK, especially after 1920. This was unfortunate given the advances in thermodynamics. The British loading gauge was also not favourable to it. If you think that compounding was a failure, perhaps you should study Chapelon's 160A1, 242A1 and the N&W Y6s?

 

Worlds most frugal normal steam locomotive, according to Cox, was SNCF 141P fourcylinder compound with 16 inch diameter outside HP cylinders.
Such small cylinders can easily pass even the narrow LMS mainline loading gauge

posted Feb 6, 2018 at 10:13 AM

The real UK steam devellopment problem was structural.
Locomotives were only tested by its own design team .
After ca 1920 the smartest young engineers made cars and aircrafts.

It would be maddingly interesting to see what a Webb Class A three-cylinder 0-8-0 Compound could do today with a little invisible up to date improvement.
And quite usefull on a heritage railway.

 

Well, the Midland compounds seemed to do OK, but yes, the UK loading gauge was/is pretty restrictive.

I do wonder how Fowler's proposed compound Pacific would have turned out - hopefully with decent axleboxes, but that's another discussion...

Mark

 

Thank you for the correction re de Glehn. The proponents of compounding are never going to give up, not ever. It didn't work in the UK (or North America) because of the maintenance overhead. The US and Canada have a very generous loading gauge but the compound Mallets weren't as successful in North America as simple Mallets. Just too expensive to run.
Best to keep it simple. Three cylinders are enough for the British loading gauge, four is one too many.

 

As far as compounds are concerned, you won't find anything to beat Chapelon's 242a1 in terms of power per unit of grate area. Compounding didn't work in the UK because we didn't have the engineers to develop it or to be spoon-fed by Chapelon even (Gresley is an example of that).

Mallets are always compounds as compared with simple articulated locomotives. Mallets were only useful for slow speed slogging because the two units are not mechanically linked and therefore the steam pressure in the receiver was always unpredictable. But for this purpose, they were much more economical than the simple articulates. The N&W Y6s lasted until the early 1960s, long after most US lines had given up steam. The simple articulates were built for heavy, higher speed hauling on flatter terrain.

 

It is nonsense to say without qualification that compounding is good or compounding is bad. The choice between simple and compound expansion is like many choices in engineering, with each having pros and cons. Two factors in France, the high cost of coal and (as I understand) the higher level of training for drivers, tipped the balance more towards compounding then in most other countries. Then, given the building of large numbers of compound locos, there was perhaps more motivation than elsewhere to work on optimising their thermal efficiency (to save coal) even if that made them a bit worse in other respects such as the costs of initial building and maintenance. The results can be seen in the designs from Chapelon and de Caso.

 

Flatter terrain? I wouldn’t call Sherman Hill flat and the UP articulated locos were built to tackle that.

 

A very good reason for making articulates as compounds,was that flexible steam pipe to articulated front truck only had to handle receiver steam.
Max 4 bar and 175 Celcius instead of 18/400 .